My Heathkit SB-200 rebuild page, by: K5JXH  Ham call sign, mine.                                             
The is my Journal... 
WARNING the voltages inside are LETHAL. Safety first, means zero second chances to live.

My RF Linear class "A" R.F  Power Amplifier "PA" 1000 watt (cw) circa 1964(mine) to 1978 (a.k.a. old iron or boat anchors)    It's not , full legal limit!
A better PA is a Heath SB220 or 221, with twin 3-500z tubes. (or the Kenwood TL-922 ! is at limit )
At the end of this page see the ICom IC7200 cures for its nasty habit of  gating RF on, way too soon ! (the Transceiver must wait unit the PA gets into TX mode and then excite it.)
Many Modifications are not needed at all, if using an old tube exciter  (sure soft AC power on mod. is always good and ReCap'ing)
The reason all the other mod.'s matter is so you don't blow up your new solid state exciter now called a Transceiver. ( 1 to 100 watts out of it)
The most offending are 2 things, one is the TX key line runs high voltages,  120vdc, and even 50mA or more current that blows out the relay  in that $1000+ Transceiver, is that fun? no!
The 2nd thing is GAS , if the PA gets GAS that gas can cause flash over and blow to hell your transceivers (antenna out port)  Always have plan for GAS, and keep and eye on the tube for signs of gas.
Modifications: Mod's for short !  (remember this, you are mixing 1970 gear to 2020 gear, a 50 year gap in evolution.  (but really is not hard to do) (joking: Tesla meets Mr. Sony)
No PA this old would you skip a full inspection, for damage , and missing or wrong parts used inside.(see this great web post on Kenwood)
  • Safety mod.'s top the list as one expects,  Case grounds inside and to the AC 3rd wire (USA green) to the Wall AC power all with way to the out side of home service entrance, earthen ground strakes.
  • Make sure the PA does not have cloned tubes in it from China or Russia, that have low Mu (in lay terms, weak low gain) this tubes have a wrong bias spec, and conduct  Ip current key off. (Ip = Current plate/anode)
  • At the gate (day 1 box open) check the Grid caps for 3 things, they are good, and leads as short as possible and there grounds are perfect !
  • Soft power on, mod.  to reduce or prevent input AC surge currents huge, via any method. (I used thermistors! )  ( a slam dunk in the good idea box,, regardless of method.)
  • Soft key mod, do this so your new Solid State (SS)Transceiver is not blown up, The microphone key or CW key jack has high voltaged 100-180vdc present, (tube PA does this; SS not ), This is just SS buffer card, (DIY) or .(SK201 or SK922)
  • HV PLATE glitch suppressor, resistor. (protects system from tube ARC/GAS events. (lots of myths too) SORRY Nichrome wire makes it worse !
  • HV bleeders still good and not missing. (tested) ( watch the PA HV meter fall when you turn it off, if it lasts too long (over 1 minute to 100v or less,  the bleeders are bad)  A 15Gig Ohm is there, test it make sure  it works right. ( use  CHICKEN STICK first)
  • ALC mod.  just surge device, so no HV can ever land on this pin ever. (transorb or other devices seen below) -10vcc max, never 180vdc nor 3000vdc.  (My Kenwood only needs ALC clamps, unlike what I did to the SB220.)
  • On 3 jacks I install the DUET protector kit or my Triad protections devices. (TVS/GDT,Zener) (not RF OUT ) and no zero on RF-input
  • CAP mod.s  (a.k.a, RE-CAPING the Power supply,  Do this, do no  run 50 year old CAPs TODAY. OK (talking only electrolytic only and the below pair)
  • Get rid of the now illegal AC power jack, line caps.
  • IF you run QSK kit , fast silent TX vacuum relay and soft key both, there is board for that, on ebay.
  • 10meter band upgrade kit. (if needed)
  • I will not tell you about  non-stock plate cap PC1 (L+R) VHF suppressors, The stock ones worked for 50 years dang !. The only problem child is the 2w carbon comp. resistors  inside are not made today but non inductive film type do work.
  • And last the Meter diode protectors, two 1n4001 diodes back to back parallell across the meter, for  GAS ARCs do not blow up meter, technology. (most old PAs need this)
Keep in mind, if you do upgrade it, resale may fail no buyers, or be less value. (on cars good advice but RIGs, no way Jose')


A.k.a, The alligator PA box:

(see my project huge slide show here)

The PA is called the Alligator box, big mouth  and tiny ears.  (you can't hear them but they sure can hear you)

My jump index table:

SAFETY
HV Cap. UPGRADE
Day 1 checks
Soft key mod.
Port clamps !
Grid fix mod.
  The (my) Triad protector
De-GAS
Filament worry.
In Rush surge mod.
Glitch mod.
Matched 572?
Anode Suppressor
Fast Tuneup steps
 My hit list
Teflon tricks.
Ground horrors
ALC analysis(long)
ALC short story
Final results!
Jargon


I decided to overhaul mine, (rebuild):
One can tell at a glance if the PA was home kit made or made in the Heathkit factory, (soldering work tells you that)
I'm doing this AMP military style,  (I'm doing a near MIL classification "B" overhaul) 
We did this in the military to be sure it's battle ready, not full of 1/2 dead parts and rotting wire insulation. (I did this in Viet NAM war) The GOV flat refused to do the overhauls, so I did it my self (tired of daily failures, and I ENDED THAT)
A Class "A" overhaul is all new tubes and wires and electrolytic caps, and more (the latter if you read the full data sheets on caps, you learn that the electrolyte does not last forever, most have a design life of  near 20 years , see the "bath tub curve here" page 20, it's real !
I will call those capacitors Cap.s from now on.  The HV (highvoltage) caps this old, I buy new ones with low ESR from top maker "JAPAN"  Nichicon, from MOUSER.com USA.
There life can be short at high temperatures and surges/shock/vibration/ and I might point out the original caps had a very poor ripple current spec, (mines 3times better ) and huge ripple current causes internal heating and short life span.
CAPs made today are far battery if not buying from C4,  (China commy cloned  crap) They even tied the steal the  electrolyte formulas and got it wrong in early 2000s.  LEARN THAT BRANDS MATTER .  See
counterfeit parts
The life of those parts can be super long if the Amp spent 50 years parked in a basement at 60F,if true?, or not in a 150F attic .
Old PA's that sat decades,  we use a thing called a VARIAC the first time, to reform those old caps.  (new capacitors, or what is called a Re-CAP job is wise.) All caps shaped like tiny garbage cans with + plus on one end.
 My original  Cap is  "AEROVOX" B25-34, 125mfd, 450dcwv. 1964 date code.  (they were bad then and for sure  now)
There are no data sheets on them now, but you can be sure the ripple spec, is marginal then,as did leakage and life ratings (even the size shrunk)
It's easier to up grade the PS (PowerSupply) than to test all 6 HV cap's. (for excessive leakage, LOW uFarads and high ESR, past hidden damage, etc) You do know bumping them cracks the pin seals, right?

I use Nichicon ! There are other great brands, but this Japan maker tops my personal list.  
There are 2 kinds of old caps that fail not the ceramics just those seen in old paper or old wax dipped CAPs.
See this page on those funny caps not in this PA (leave them alone unless found to leak in circuit, leave then be.
This PA has 2800vdc inside, do respect that 100% ,it is lethal to touch. (no second chances allowed)


At the least,  have all safety issues corrected and make your $2000+ ICOM not blow up, read on.
This is not about shot gunning anything , but the last steps check the AMP out fully for performance, before putting it on the AIR.  
And amp that is 50 years old, or mine 56years old ,shot gunning can be a winning plan.  (caps first, and tubes never, due to the $150 cost of each)
But see my list of things to check:
  • Unsafe or missing or Bad grounds everywhere. (mine had 10 bad, most were just screws loose.)
  • Bad cap's. (electrolytic only) I tested them 3 ways,  uF meter, leakage and my ESR tool showed 2 of the huge caps at 13ohms (oddly the uF passed),the good ones showed 1ohms. ESR ( I have zero faith in any 50 year old cap like this)
  •  so...upgraded to real Japan NICHICON caps, the best made. (from mouser.com 647-LGN2W331MELC35"
  • Bad Carbon comp resistors. (any over 20% off are bad, the end band tells you tolerance spec.)   Most of not all Carbon Comps Resistors are not made now, we use metal film or the like now. 
  • The Plate cap CP1 L/C inductor pair and resistor is near 50ohms and MUST BE NON INDUCTIVE 50 ohms. (do not use any thing else or bad happens, VHF parasitic oscillations
  • When powered up, the amp after above cures,  we use a variac transformer to easy the line voltage up , and see  if the tubes have gas.  Violet colors, or light blue,(nitrogen/oxygen), if yes, new tubes are the only cure. (open wallet and buy top brands not clones)
We do a full inspections and  test all critical parts, for condition.  (parts are either  in-spec or out of spec.,  many parts way out of spec work great, but consider that gross measured errors, is  a top sign of the parts near END OF LIFE) Carbon Comp resistors love to fail or be bad .

This page covers 3  major topics,  (The MACRO views)  (ah 2016 meets, 1964 and no smoke?) Mine has run for 4 years and zero problem after doing all the work my self.
  1. Not blowing up your $2000 +Transceiver, connected to this 2000v+ volt amp.  (solid state, is near 100vdc or less transistors and friends, and will never like 2000vdc ever, or BOOM)
  2. Replacing things that are for sure bad, like all 8 electrolytic Capacitors and using top brand parts, not fleabay clones nor counterfeit parts, ever. (hint, mouser.com sells real parts, try Nichicon)
  3. Mod.'s for protection and weak grounds and mod.'s for 10x better RF grounds at the Tube V1 /V2 pin 3 grids. (short as possible fix) This lower the lead inductance so that you don't get parasitic oscillations, or cross talk (mutual inductances feed back issues)
Things that can be ok, or not. A quick list: (the rebuilders list)
  • Chassis is ok. (the fine aluminum can't rust but can corrode) I clean it with vacuum , brushs it out and last with electronic spray cleaner safe for paint and plastics. ( I make sure all grounds are cleaned of all corrosion, and tight)
  • All wire wound resistors. usually are all okay.(ok)
  • All ceramic and paper caps (not found here) are OK. (all disc caps ,  usually are ok)
  • All mica caps, ok. (some are square with color dots) My grid caps are maroon drops and are usually ok,  but I bought new grid caps, to get he best grid RF grounds possible.
  • All grounds bad. (take that to the bank)  The safety ground is not safe, nor is the weak 6.3vac CT ground ,I moved it to a  REAL failsafe ground stud, make it so it can NEVER FAIL. now.  all screws on mine (vast) were all loose for ground.
  • The main variable caps, LOAD and TUNE are OK, but look for bent fins....look hard. (you don't want them to ARC or short out out from obviously bent parts...)  You can fix most those VAR CAP fins, bend them straight and file sharp edges works ok.
  • The TUBES if original may be good, mine were. (Hams only use small hours on 10,000 hr. rated tubes, is why many still are good after 50 years)
  • Bad solder joints , you will find these, for sure if it was a kit so.... will be bad.  (my CR1 coax cable, had no solder on the end center wire, making it DEAD,  the panel meter now works like new.)
  • The 2 caps on the 2 grids may have 4 times 1 inch of wire leads there, that's not good.  I moved those C14/C15 grounds screws closer to the tube pins for lowest, parasitic inductance and a good cure for VHF parasitic oscillations. (my leads are 1/4"  now)
  • Bad wiring, old ,cracked or burned, Use Teflon wire, and end this little horror. (the HV wire needs to be upgraded) Some wires the insultion is so old it is like "Graham cracker" frail and cruble at the touch. Teflon those I DO.
  • Look for missing parts, who knows, what you bought? Look for missing or mod.s that make no sense or wrong parts.  (like using 12v rated cheap wire in the place of the 3000v rated blue wire.)
  • Test all critical parts. Some that look ok can be bad. Testing is easy.  (R, C and L, and leakage. are all easy basic electronic testing skills) Or ask an ELMER to do that ?
  • Look for damage, if not clear what damage looks like, then look at photos of good amps. Learn what good parts look like, what burned resistors look like or parts that crack, blister, or peal or leak.  (Like in the movie Gravity, NASA says, "science said, give it a VIS")
  • Make sure the HV Power supply inside 15meg ohm bleeder is not bad,  (if 100vdc attained or less after 1 full minute turned off it is bad) If found this bleeded to be open, the RIG IS a danger to use. with 2800v lasting for ever.... Fix this first.


Inspections are the order of the day,

Protection mod.'s,   are key to protecting the connected gear,  (2 or 3 cables, from blowing up, gear that hates -130vdc or more, as the bias supply will surely do to your ICOM (or like kind, most are like kind)
Each protector has value, (I have 3 major (at least) protection devices to save your ICOM EXCITER , in my case IC7200,  most modern gear can not handle over 16vdc on any control line, and you sure don't want 2700vdc slammed in to the ICOM final RF PA, do you..(gas caused)
What if  some dweeb spilled coke in the AMP.?
What if it was used full power with no load and the wafer switches contacts are blown to hell. (open lid, chicken sticked after 2min. wait and check those contacts with due care)
The what If's, are endless, and it's all about the looking first. then turn it on.
 
There are  modifications, called mod.'s for short. (I tend to the less mods the better, but there are some serious issues with any 50 year old amp and some bad practices, in the Amp. )
The makers never dreamed there would be solid-state gear, like we have, nor did they dream if their amp lasting 50 years, into the future !  (Imagine 2700 tubes on one END AND DSP on the other. hark)
The Amp has no computer processor, nor any ability to protect the output of the amp, with  no load, the folks that say the TUBE amp. is immune to this, are wrong. (see the proof here, nice work, read this first is my statement)
The Amp has no operator error protect circuits (or brains). (one can even design analog protectors, easy, I worked military gear that hand lots of this, even fan blowers that had drop vane switches that killed power, (and a war over-ride switch )
The Amp can not protect itself, that is your job 1. (running no load on the output, is not going to bode well, Mr. Sparky. )


Some mod.s come in 2 flavors, the HYPE mod and real mod , Plate cap, RF suppressor paranoia. (yes, the 47ohms Res. like to burn up, yes...)  (it can be tested for stability or even tuned,  if you read the pages of HAM K8JI . Kilo Eight Juliet India)
You do not need any Aliens have landed , U loop, Nichrome wire is worse not better, RF suppressors. (IMO) nor the idea of adding relays to anything...
Try to understand that if the GRID really is at RF ground, there will be no VHF Oscillations, ( looked for VHF spurs?) I use my SDR, in spectrum mode, if nothing seen out of band then you don't need alien coils from mars.
Do not wire ground the grids, use the top rated caps only if you hard wire the ground GRID BIOS is now dead and so is ALC, and you will get IMD transmitted.
There are problems of gassy tubes, old tubes can become gassy and Arc, (look in the dark at them , ignoring the filament glow)! (pink/blue/purple,red, are gas , look at a distance, hands in pockets, or with 1 hand on key , standing back lid off...with due care.)

IMO, IN MY OPINION:
90% of the time SOFT POWER ON is not needed but for a fact this happens, transformer flex at the surge issues, are real and it is best to address them now. (ever price a new transformer or even try to find them ,so rare now)
I will show 1 easy way, not using  any stinking rotten RELAYS. (if you repaired huge TVs in in the 60s like me , you'd know that they used thermistors to stop the surge , even then and is simple cure)
The best way (surge fix) bar none, is the zero crossing detector way (or tiny $1 CPU does this), there is no better way !  it makes the surge never  happen again. (and is a solid state fix, not dang relays) << nobody sells these, well affordable. that is.
The THERMISTOR FIX is a winner.

I also know most users of this powerful linear RF amp will try to use it on a 50 year old rig or a much newer modern solid state Transceivers .
Most  modern folks will go , what is an exciterreading this old manual?  "it's the base transmitter, (in the olden days Hams even made the VFO(or a single crystal) and all parts of the transmitter by hand, DIY, and a 100watt PA by hand.)
In my case, I have a working and backup SB401, that can also drive my SB200 no problems and zero mod.s.
My boat anchor SB200  below, is connected to my huge antenna tuner, "PalStar" AT2k on one end (RF out) and the  ICOM 7200 at  my SB200 ANT.InPut jack.  
This is and exciter, old and borning, but now you know that that word means (at the least).
Called Home brew, now called DIY.  note the old crystal !





OLD internal hookup wire?:    (my first horror, here)
First off it's cheap vinyl plastic horrid solid copper hookup wire, is not even tinned (some are , some are not, oddly).  I hate solid wire ! loathe.
Most Vinyl insulated wire (1960) don't last long on  hot gear, that  is because red hot tubes cause this cheap plastic to fail, due to heat and age. (the best cure is Teflon™ wire, Generic name PTFE) 
This cheap wire can be in very bad shape,  old, hard and cracked, and the solid wire breaks if bent, or vibrated or is nicked, (means when stripped the copper was nicked (cut or scored)
Really, no gear or rig should ever be made with solid wire, ever.  (but is dirt super cheap and hates vibation. ) (but is used in the RF cage as  silver plated cooper pipe, due to skin effects and is always OK there)
But lucky us, most of this  cheap  wire is on the bottom cool side.

You  bought this Amp. , because we are painfully nostalgic for Heathkit's, and avoided the 10x better, ALS-600 Amp. (yes, we need our heads examined)?  It is fun to rebuild it, yes.! (a walk down memory lane for me)
Day1 PHOTO: In the rough. (a diamond in the rough as most are  50 years old)
 Mine is carbon dated at 1964, (date codes checks) I got mine from an ex Viet Nam vet (like me), ham. (one nice guy too )
click all my photos to zoom them.
As they say, SAFETY FIRST:
This unit has 2700vdc inside, and is lethal, (yah) :  IF this scares you and it should, then stay the heck out of there, let and old TUBE guy do the work and live. Find a TUBE ELMER.
See how to work as  HV tech. safely here.
The PA goes safe after 2 minutes off and HV meters lands on zero. if not the PA needs work. (bad bleeder resistors inside make it 10x more lethal)
 
Use a chicken stick  and live.

Wash your hands before eating, as this AMP is full of LEAD (Pb)  (don't let kids or the dog  lick parts as it does have  old solder with lead, (Pb)  )

The jack marked, Antenna Relay on the rear has -130vdc on it and will shock you,  not badly but will not make your day. Do not touch its center pin with hands. (mine no longer has this voltage , due to my soft key upgrade)

The odd chance....?
If say your home was built back  in 1950 and has no safety grounds (2 prong wall jacks? oops , as I had as boy, my short wave radio shocked me many a time)
In this case,  you need to upgrade the house wiring,  with 3 wire 14 gage to the shack power outlets , and  install a 16ft ground stake outside, in the ground and run at least 14AWG. wire to the rear AMP wing nut ground lug. (to be safe) using NEC and NEMA rules.
If not sure?, call up any IBEW electrician and get it done fast and at a fair price. (and safe)

BTW there is no true  full lightening protection, but one can use those many protective devices for near misses and WIN, and are the most common.  (win means the reduction of cascade failures, not 100% protection ever)


Never turn on this AMP, with the case not grounded to the best ground bus in the shack. or to NEC (national electric code) grade power outlets 3 wire ,working right (radio shack yours, not Tandy) or risk the case floating to 2,400 Volts D.C.  (lethal)
 
Never run it case open . (due to RF emissions on 10m bands and HV present) It's ok,  for photos as seen here, or testing with wires connected to meters for short tests.  (hands in pocket hot) (all meters etc, rated for 2700v or more, and probe wires same)
Live testing, requires: (when dead and chicken sticked, connect up say a HV rated meter,, power up, take readings hands in pockets,, power off, disconnect, the only safe way to work HV gear)
Never trust the 50 year  old gear panel meters, ever, to tell you , it's  safe to play inside, no really, (think hard on that ?) If the panel meter is deflected on HV, it is to be believed ! stay out,  but if stuck on 0volts.  It can be a man killer.
and never just  turn it on , when first received. (inspect it first, see how below....)  Did the tubes fall out during shipment. (or break?)
That panel  meter BTW, is now Unobtainium status.

Never use plastic panels in place of  metal, this amp. has Faraday shields, so do not damage any of the shields or cut huge holes in them. (think about your safety and EMI, RFI and 10meter radiation effects on your body and others standing by you!)
My top shield has and added hole drilled to reach my new custom ALC pot. but is very small, hole and 100% adult safe. (1/4" in size)
Always use a large handle  ground rod (chicken-stick), AMP turned off, to ground out all capacitor banks huge charges there, never ever trust the internal 15gig-ohm bleeder to save you, in fact we test it first, and  OHM it OUT ? after chicken sticking it on both ends)
Consider the those relic carbon resistors can be cracked inside, and work one minute and not the next. (hint, they love to do that trick)
 Please be careful.  Make darn sure R3/ R4 and R12 are GOOD ! Look hard and long at them,test them, or replace them. (and R1 too)
(end safety rant)

General checks:
The carbon comp. color banded resistors are light brown Bakelite(tm) (phenolic?) and if you see dark spots, or worse they are bad, or cracks or bubbling of the surface or the colored bands now all discolored brown to black?.
Case in point, my amp had many CC carbon comp resistor all reading 100% over marked value tolerance, In my long experience as a tech, 1965 to now, this is very common.  (either from aging factors or overheated)  even if that error is ok, in THIS circuit, it's a sign of END OF LIFE.
Some resistors can work great now and be 100% off, due to there purpose in the circuit, do don't let say the color band tell you need of accuracy, in many cases the gold band resistor was cheaper, or the only ones sold at the time.( or 20% were hard to find, as is and was common)
It is best to check all resistors in the Amp. using and ohmmeter for valid readings. (on any DMM meter, turn the dial to resistance) {lesson 1 if a resistor reads too low, lift it on one side, and check again}
Make 100% sure the badly designed 6.3vac center tap ground terminations is corrected , this CT (center tapped) wire pin must have  a  perfect ground, or bad will happen. (I  modified mine to make it never fail again)  see my Day 1 section below....
Never ever float the grid or float the cathode with  power on. On this grounded grid RF amp. (< mods seen online, all wrong, all a danger to you and others)
BTW: our grid is RF grounded, but is DC biased. In fact the tube draws 90mA un-keyed. (it is a CLASS "A" amp) (class "A" means the tube runs in its center, linear conduction zone and no distortion or clipping)
I see no reason at all, to change this tube to self bias or cathode biasing schemes at all, ever. (but I am China tube free, so am blessed)
Make sure all resistors in the power-supply are all good, for sure the bottom 3 resistors on the PS PCB(printed circuit board),  that terminate to ground (R1 does that), 100% GOOD. (mine are new)



The Job list is first,  and by order of IMPORTANCE:  (real fears not paranoia)  I have no Axe to grind ever, no parts to sell, zero, I'm 100% retired...  (I am not affiliated to Mouser.com , be that directly or indirectly!)

Next, I will take my top 3 top of page macro listed, items, and expand what I did, and why and details.
 (Now my joke,  the old retired , Heathkit engineer never once talked to , that ICOM engineer,  (50 years apart)  (Imagine Mr, N.Tesla talking to Dr. Gordon Moore of Intel)
Joke 2,  if Moore had his way, this amp would have 2million transistors. (somewhere,,,,,)
This is not a "TODO" list , it is only a list of things ,you might want to consider, on a case by case basis, the  first 3 are mandatory in my opinion)
Keep in mind OHMITE resistor maker lists carbon comps as obsolete now, it's a dying breed,  END OF LIFE, NOW. (but they make them in China now for 1cent each or sell them to Audio Amp. dudes for $100. )
The real truth is Ohmite can make and sell 1 cent resistors as China can..  (only 3 cents)
Some of these mod's are from top makers of Tube AMPS still sold and  still in business today. (we try to do what they recommend, if 100% applicable, after all the problems are all the same or very closely related)

  (PLUG AND PLAY words you see some (no names) sellers use this phrase and , is silly,  here is  a DREAM or at worst a painful expensive to digital wiz kids...) This is an analog Amp. there is no digital PnP back in 1964, NADA !
Now my long running list:   INSPECTIONS WORK,  and for sure old HV gear of anykind made.
  1. All inside ground screws upgraded (or missing) star washers, for the best ground you can attain (all ground screws tight as can be, spec) On inspection day1, I found 10+ critical grounds "loose as a goose". including the green safety ground.! (So glad to find this...)
  2. Make my new ICOM 7200 safe from this nasty 2700vdc beast..   if nothing else done, do this next.  That is done many ways,  the below Keyline MOD and using voltage clamp technology on  the ANT.RELAY jack (right at it) (I used no less than 9 clamping devices)
  3. The keyline "Softkey" SK201 mod. {Ant. Relay. (makes the key line safe and allows new wimpy transceivers to drive the huge T/R relay inside, at minus -130vdc, (that too is painful  or worse touched)
  4. A New Power supply PCB  and all new electrolytic caps resistors and diodes.  (I got my PCB for $15 and used all  6 Nichicon caps. , wire wounds resistors and 1N5408 (x16)  diodes  from Mouser.com.  $65 for parts alone. The best diodes are 78-BYT78 (mouser #)
  5. All the silly "RCA phono Audio" jacks are now gone and BNC, and SO239 , for sure you will not like RF-in falling off to your ICOM would you ? That is what and audio jack would love to do.
  6. All 3 rear jacks ANT. RELAY & ALC are not protected with a paired set if TVS, and GDT#2/3 seen below:  (Transorbs and Gas Discharge tubes)
  7. In the same vein , 2 more GDT devices. GDT1 and 2 below in my schematic on 2 cathode  pins of the V1 and V2 main tubes.   This mod. protects our amp parts and  the ICOM RF out  FET transistors.(mostly on cloned 572b tubes)
  8. If the sockets are loose these might work gold sockets for the 572B tubes  ( later a pin broke and had to put in new ceramic shockets)
  9. 2 New 47 ohm plate cap,   2watt resistors , my are reading 100ohms, obsolete now in this size, so we use modern parts, carbon oxide and sold on mouser.com  3watts fits. best.) Mine were 100% off too high. My antenna analyzer at high Freq, shows the resistors too high.
  10. A selected a  set of Teflon (PTFE) wiring and coax. all mil spec.  Nice upgrade this, RG PTFE coax. Mil .spec. (on mine the internal insulation was melted bad and deformed)
  11. I changed 2 wire colors, one gray (too many gray in the box) to orange, and  one meter wire to Black (-minus) All wire changed by me is mil grade Teflon, 19strand , silver plated. the best of the best, I stock it in my shop.
  12. Inspect every inch of the AMP for damage or parts cracked or burned, replace all found bad. Do the web linked ohms checks seen here.
  13. Use Carbon oxide resistors (mouser has em') in all cases where you see larger  & bad carbon composition type, cracked, blistered, burned  or aged and iffy? (Just the large ones with color bands?), the old COMP's are now obsolete. (mostly 1watt size and up) 
  14. Clean it the best you can after using a brush. ( blow it with shop air, and use a soft brush to brush out dirt, lent and electrocuted dead spiders. ( coffee and cola with sugar damaged)
  15. I use  Dutch cleanser on outer main case(no green Comet with bleach) on the paint to get it 100% clean,? My cleaned out case looks like new now.
  16. The horror of SURGE? (hihi) If you just have to have it? , try mine $2 fix . Power ON SURGE fear.(sold at mouser for near nutt'n) See my live measured data too.
  17. Strategic cap changes, all 8 electrolytic , any cap that you see, that is 30years past the 20year design life of the cap,, needs new Nichicons.  (using the  Arrhenius formula) 15k hours is common spec. but stored in the attic, what? 2 to 16 years is the life, in a hot attic less.
  18. Stored caps can out gas from heat or poor seals, get lighter in weight, and can have high ESR, in some cases powering  them up , they can reform and self heal and last far longer.   1hr /day is 41 years, life, used every day. (but heat is the killer)
  19. ADD 2 Meter diodes,  connected back to back on rear of meter,  as seen on all other equipment mead with analog meters, 2x 1N4004 in parallel at the rear lugs of the panel meter.
  20. I made one more tiny mod that allows me to simply unbolt (screw) my 2 plate resistors, allowing me to do tests or run Getter gas cleaning mode. I can also run only one tube this way. (proving amp works with 1 tube and not 2, if true, or reverse them like that... we are not helpless..)
  21. The top lid (blank aluminum )  inside, has many screws, that I will upgrade to NUTSERTS  and 6x32 sst screws. (not using sheet metal screws there ,which I loathe and were stripped out.) Like the bottom feet screws were  presently nut serted and takes a special tool to install.
  22. I found the paint codes on line Apache Green,  and didn't like the shade, of green.  I do know how to match colors and paint it , any commercial shop has the paint and can color match this paint. (but won't be cheap, end to end.)

Apache codes: FOR ONE QUART (rumor says too dark from front panel)



The Caps issues are complex.   (electrolytic)
Luck, and heat top the list.
If the  leakage of the cap gets too bad , will send the voltage to the next cap that don't leak, and if over 450vdc, this cap, that is 100% good, shorts out,  bam.  (the irony of good caps short)
If running a stock relic, use a device called variac and bring up line voltage slowly and let those old relic caps, reform and pray , not leak or explode.


It is done, now (except ALC mod and tuneup),  and now so are labels..  It's even got full power, even with 1964 date coded tube, I am pinching myself. wow !
Lucky for me Nov.2016 they sell overlays for this panel, as see below. I later replaced the cracked meter for one that is NOS.

The Dial lamp is now a blue LED lamp, using matching socket as original.  I talked to Missouri (said on the river too and was mobile) last night.


Photo below is before new front overlay. seen above. DO NOT DO THIS AT HOME, I am a certified HV tech,...  and for sure not KEY UP ,flooding the room with RF radiation. This is before new front panel.

   
MATCHED TUBES,  ?
The need for matched tubes are mostly not necessary,  if not running at a full gallon you don't need matched tubes.  Matching tubes, might get you 10 more watts before it clips?  Why do you need 10 more watts.?
 If you monitor your signals and are not clipping (aka, flat topping) then your tubes are GREAT.
 It's not a push pull amp, nor is there a crossover, point from same, that can distort , if not matched or bias trimmed. (crossover distortion is not here, only clipping can happen) It is only a grounded grid AMP. CLASS "A".
The tubes are in parallel and with same DC bias and same RF grounded grid, and same cathode drive and same plate voltage, but if one tube conducts lower at a given bias, it's only weaker (or Chinese?).(why is this an issue? run it another 5000 hr's and be happy?)
 I did this trick using the PA as a tube tester here, SB410.
That means there is no risk of a weak tube, clipping the output, but we  are not taking gross weak tubes (as in end of life),  we are talking only not perfectly balanced, and is not needed at all.

THE AMP IS A REAL TUBE TESTER OF THE BEST SORT:
You can test the tubes 1 by 1, inside the amp.  see what your PEP is on each, and then find 2 that perform well, not OCD balanced. (I'd bet most tubes sold as balanced, are in fact NOT , not one seller can prove this, not one.)
Not one seller states, BALANCED to what DEGREE?   (SAME Ip at same Grid voltage?  with in what? say 10ma or what,??? what is the spec, on balanced,  answer is, ..... you guessed it..... Silence here speaks , volumes to me..
The drive curves are here,  "The Characteristic family of curves." This is the tubes, biased gains. Spec. Peak plate current is 625ma.  peak grid  is  125ma,  Peak PEP is 400 watts  per  tube or  800w both.
Testing, (after burning off Gases on old tubes)
You could , run the Amp with 1 tube and see what input drive, gives you 400watts and check for IMD a the full drive.  (3rd order, IMD spec. is -28 DB)
Then repeat this on 2nd tube by itself, and if they perform near the same (ball park) run them..
 All that matters is you get the power out you need , and  low IMD and no clipping (aka, flat-topping or early saturation)
I have no comments on how to deal with cloned tubes, from say, Shuguang.(production ended in 2013)

What remains sold new, is the joint RF parts and China parts(secret true maker , see my guess below in a photo of said tube)  At $80 bucks a pop.  (it's a nice monopoly , everyone knows what the down side of that is,  remember ATT charging 84cents /minute to call across the street?)
Keep in mind, there is no computer protecting your tubes, for crazy high VSWR or those arcing over hot switched at anytime, or  when you run this amp unloaded !  
The 2700v does sag to about 2400v, at full power.(naturally) Don't freak out when seeing this... ) <<< this is with my new 30amp AC SERVICE and filaments that do not dim, at transmit key up.


Warning many of these cloned tubes from china have low Mu. (and use lots of current in standby (key line not keyed) that be super bad, if you see that.
 572's from Shuguang (Dawn{engl.}) Minimum order 2, most direct. Here. (watch out for gas!)

FILAMENT WORRIES: (click here to skip this mostly is a non problem, worry)  But the bad ground is real and cured here.
(keep 120vac LINE into radio (not 100v) and the AMP, transformer secondary 6.3vac,  wires not cold soldered (etc) and all will be good and the known weak ground at transformer.
and
The filament secondary is limited to 8amps not 60. So the filament surge naturally limits, current to the filaments at start up, so IT'S NOT AN |ISSUE  (not to mention mutual inductance limits here)

I see folks acting like 572s  , are EIMAC 3cx---- or 4cx250s and up,  they read the EIMAC  white paper on filaments and then think 572s are the same, sorry NOT.
The rule is +  or - 5%  (.05 x 6.3vrms is , .315v variance ) at the tube, not the wires a the transformer, but on the actual pins of the tube.  fully hot. and normal line voltage.
So that means,  low volts is, 5.985v(rms), at actual pins. (no where else, just THERE) 
Do not buy internet sold fixes for this, it is not needed.
 If your A.C line voltages are low, fix that first, (new house wiring) or buy a 1kva UPS, (jocking the latter would blow the breakers) My house wiring sags 9v , transmitting,  4.5 for hot, 4.5v for neutral, I must  upgrade my house wiring, not the amp ever.
I had to upgrade my 120vac to  30amps service, due to line voltage drops to 108vdc.  My filaments are now with full power now, when transmitting.(mine were dimming)

RF power AMP.  (this not a mod, it is just getting full power to the filaments.)
W8JI 
W8ji covers all these topics , clear and the best read on the net, IMO,  on RF amps, for sure. (His topics are wide of scope and all good)

Every effort will be expended to keep filament voltage at spec. 6.3vac (rms) 

"572B tubes use 6.3 volt filaments, with a specified voltage tolerance of  +  - 0.3 volts. tolerance is roughly + - 5% of nominal filament voltage."
That means line voltage (19:1) must not fall below about 6vac.  120v - 6v = 114vac rms. That means 20amp mains for power.
You will not get this effect, but read how it can happen, on very old tubes run say 10,000 KEY ON, hours.  A good read.
Avoid this,
carburization properties or become getters  and damaging filaments...) In most cases you will never see low filament voltage damage or shorter filament life.
My rig, the filament transformers 6.3v center tap was not grounded properly, the Heathkit method of doing this ground is really really bad.. Using a weak phenolic stud. (I  now use the huge transformer mounting bolt, as this ground and a larger ring lug there)

If you run them at a very low voltage (say 4v), with a  bad solder joints (cold soldered)? , the tungsten acts as a GETTER, and eats up natural internal gases (good thing) but  , damaging the CARB inside, a bad thing, but most amateur will never see this, due to short duty cycles on 10,000 tubes.
The old parked in attic tubes, or NOS tubes, can leak air into the vacuum of this tube. (called gassy) They don't glow only orange, they make other wild colors, red/blue, purple are gases IONIZING in the poor vacuum in side.
There are ways to deal with this. I will make no claims on making tube last longer, as many of these tubes last for 10,000 hours.  (I can't  measure that)
The best way to avoid low filament voltages a full 120vac on the main lines into your radio, not using tiny gauge power strips, and the like & (not using 1950 house wiring with 16ga+ Romex wiring in the walls) and the bad grounds above. (my power cord is 12ga, 3foot.)
I've seen shacks running 100v weak lines with bad AC power feeds, with filaments at 5v.

See all words on GETTER here, the Getter eats up the gas.
The Getter eats gas up every day of its life then one day the getter is worn out and gone and now the tube, gains gas and fails.
W8Ji , nails all this, we own him a great debt for such great web pages,  they are priceless , IMO. ( no better truth exists on his topic parasitic oscillations and myths solved)
See here how to fix gas, if lucky.


Knob reduction drive mod:  (? meh ?)
I find this to be useless, totally. The "load" knob is the last knob to turn and It's just to back grid current off  and not power out.  (it's not touchy at all)


With fine Jackson 6:1 tuning vernier, (knob  shaft interface) Called Vernier drive.
 
I have no trouble tuning mine, lacking this MOD.



PROTECTION LEVEL I:   (the fix is called, by hams SOFT KEY RELAY mod)
This is needed for all transistor transceivers  , not for SB401s, (tube to tube rigs need non of this...)
SB200  (after 100 bids , I now have one) (your main radio is not a tube radio , is it?  If yes, then all this is just wasted fear...)
First off is the SK201,v3.1 mod.  that I modified the mod. with  a 12v ZENER clamp diode  D6,  but I also have the protect level 2 mod, that I use now...(like most of industry uses, in fact)

(there are better parts , like SMD parts, using  low voltage TransZorbs) like 1N6274A ( 60 cents at Mouser.com) with GDT on the jacks.
This whole card and my D6 are JOB 1,  do this and nothing else if you want.
Warning , one poster on Eham found issues on the 10m band the this card,  I hate to tell you it's NOT THIS CARD, you have a bad 6.3vac CT ground in your gear, fix your  ground first then do upgrades and win. (the green wire is at RF ground, a fact and yours is not.)

Questions asked, what changes will I see?, nothing, same as your cars AIR Bags, nothing until, BOOM happens.
This card is really just a 200v, key buffer card, it's prime directive, "dont blow up the new $2000 exiter", end story.
 Nothing will change, after you do this operationally (unless not smoking  the ICOM is cool?), this is only Protection,  and for  sure you don't want to send -130vdc backed into  any ICOM. (or any solid state transceiver)
The Icom only allows 16vdc max on this pin, the key part(no pun)  is a very small relay, RL501, and is not going to hold up long at -180vdc there.
The stock SB200 use a 100v key relay and 100ma load, both factors overload the ICOM , RL501 relay.
The maker of this ATN relay (yes, I have datasheet) , states max voltage is 25v. and ICOM guard bands that, to 16v. (wisely IMO)
The ICOM relay if forced to switch 180v will burns the contacts, every time you key and down, and the damage is progressive.  (the metal on said contacts burn , if really unlucky the 130v flashes over to the coil in said relay and blows the ICOM UP (this be unlimited damage).
 Other transceivers might work here,  if they have  huge noisy relay or use a HV FET, to switch this line, if they were smart they would have a FET, but many did not.
The ICOM app. guide clearly states that if using and external relay on this send line, you must use a diode clamp on that relay coil or the small back EMF from said coil will damage this receiver, in clear English it states that.
 It's just low voltage contacts below and low current, so don't burn them up.
One guy said, mine works ok, but gee, what if the damage progresses each day  and it cascades, in that the contacts are destroyed then the wiper bar, and this fails and allows HV to hit the RL510 coil.  (As a tech we see all this , all the time)
One could make an external box with a 12vdc relay in it, as the ICOM book shows, as does the SB200 manual (read it yet?). and all this worry ends,  as long as the box has  perfect grounds to the radio and amp at all times.
 The below SK201 , beats any external box in this regard.  A FET upgrade makes this PA AMP safe for all exciters made. Nothing beats this method, NOTHING ! That is all SK201 methods or like kind.
Note how the relay floats the line (RX) and ground it Keyed TX. 
Do not put 100v there or huge currents.  This photo is showing what not to blow up. $1200+ Radios, and more. solid state, have weak TR Relays. CALLED SEND below. (nor do you want arc, the contacts with 120vdc from PA.
 
Do not burn up the RL501 or get sad fast.

Theory on SK201 MOD board. PCB. ( a great produce for sure, and just 4 wires to use it)
Don't listen to rumors that it kills a band, as that is false information, that users had a bad ground in his rig.  (my guess, is the filament  ground is bad, mine had 10 ground bad, so.... there no reason to blame this board unless it shorts one of the 4 wires stated)
For old schoolers  (the Q1 device is a relay, a solid state relay to you young bucks)
Click any  photo to ZOOM  below photos.
This upgrade card for $30, lacks total  and full protection on the blue wire.   (  D6 is my fix as my TVS/GDT clamps added)
This upgrade can be made much better,  to protect the ICOM exciter from damage , in the event that this AMP, malfunctions that all 50year old tube amps, do and often.
The GREEN wire, is the problem here and captures the 6.3vrms filament wiring to produces 4.5v peak DC power  to run OC (4n33) and lets the exciter toggle the Ant. Relay line, and OC and then Q1.
The 6.3v connection is a smart way to get power, and will never tax the powerful filament source.
Cascade failure fix 1:
If the 6.3vac center tap main transformer  ground fails (helped by tube arc's), you blow up this board then it blows your ICOM to HELL via the blue wire.   ( D6 is my fix., rhymes)
MY MOD !
I added  D6 ,   and more protection at the  ANT.RELAY , JACK  using one GDT and one TVS device.  (costs about $2)
 For sure  the 2k ohm resistor below,  will explode (and desired "sacrificial")and save your ICOM,  all for the mighty cost of  D6 (for 40 cents) .

If I had designed this card, there would be 2 optical couplers (OC), one on the blue wire pin and same as below. see,  the OC (opto coupler) has 3000v barrier  protection, or more with OC upgrades. but the diode is dirt cheap and way smaller.
I love OC,  I've used them since the first ones were sold, in late 70s.  That 3000 volt protection is like magic, in applications such as this. but you really need 2 OCs here.
What we need is protection from HV tube shorts, even the filament can short the grid,  If HV lands on Green wire , if  lucky, C1 shorts at about 11vdc. and if lucky2, D2 explodes when C1 shorts, but I'd never trust  that to my ICOM.
If not, D3 conducts and blows the ICOM to H3LL, and you smoke a nice $1200+ radio.
 My mod, causes the 2k  R3 to explode first,in fact a 1/8w resistor here ,makes a far better  fuse (mod2 done) and what Killer voltages  passes this point?  my jack  15v and 70v  clamps kill the remainder! I now have 3 stages of protection on my BLUE WIRE.
My AMP had a loose and bad filament ground , this would  allow plate current to flow to here , to the green wire below and blow things up  it to kingdom come. 
I look at this device below and say to my self , what IF?  (the green wire is a very serious threat to my ICOM, then  at the blue wire seen clearly below...)
If you decide to use at ALC, jack, the same problems and solutions, must be done again. 
I am doing my best to prevent a cascade failure, with say the green 6.3vac center tap ground fails, as it love to do (not mine, it has  fail proof ground now) If it does fail,  it will vaporize, 2k below, and if not then my GDT/TVS pair  fire off and save the day.
What is sad is the AMP has no 5v,12, or 24vdc supplies to use, that would be free of RF and free of 2700vdc failures.  So to do this correctly we make  4.5vdc supply that has protection added, by me...  Those cathodes go to tubes that love to arc. (believe it)!
I have 5 layer protection here, C1 shorts, 2k blows, D6 fires. and the jack dual protectors in place.
My blue wire has about +2vdc on it , unkeyed, now.  no more 200v, ever again.
The U1 below LED inside is on by default until the Blue wire grounds this node.
That means node "A" below is 0v and Q1 is turned off by default, until  node "A" goes to about -10v (via that divider there ) The key line is -130v not keyed and -60vdc keyed
Q1 and U1 keep the 200v out of your Transceiver, all the time. So long as the black wire is connected to the best ground you can find.
SK201 PCB: (sold in many stores this card....) You can make your own key line buffer with 2 PnP transistors, easy. (in box outside the PA and Exciter, a box between ) or do like this and make the PC solid state friendly this way.
The design below is very good, it handles any PA ever made (Ham) and 3000volts of isolation !
 If you don't like modern Electronics, make a box with two 12vdc relays cascade to do the same thing. For sure low current 12vdc coils rated.  (in the old days we did not have relays like today)

found on the wide world web.
 
See how easy it is to see how it all works using real schematics?. above?  D6 is to protect from gas flash over to filaments ! The yellow wire drives the PA KEY TX relay. The blue wires goes to SS transceiver key out (send)
SK201 mod. was first.
R16 is lifted first (cut from your ICOM ANT.Relay jack, to do the SK201 mod)
The real AMP schematic  makes a job 10 times more easy to understand and to do the job?,  and no wall of  text, telling this blow by blow? (cut this  and jump that  and steps endless and confusing)
One guy said modifications cannot be shown in a schematic, well sure you can , use color,  the colors show added things and moved wires, and X means cut here. Sure you can!

The above is overkill IMO,  I just make up a perf board, and use a 2N2222 and MOSFET to drive the PA Key line.
The only tricks are finding power to run your transistors. (well the MOSFET 200v VDS or better) has 100vdc on it now, only the feed tranistor and gate need power.  12vdc wallwarts can do that.
You do not need to put  the protector in side the PA, you can use  project perf board and do this easy.

The same cure for kenwood TL922 (PCB # SK922 is sold too) or make your own. easy.


SUPER BAD STOCK GROUNDS:
MY GROUND MOD.'s:
I added G1, G2 and G3 grounds, so that never again will the case float, nor will the cathodes of V1/2 EVER float. The factory used, weak plastic studs to do hold grounds, (bad to the bone that) , 3 critical grounds, corrected.
The two bad ground are  are very unsafe, one allows chassis to float the other allows TUBE current to flow to chassis or float.  (no small horror here, make a real ground and do not use plastic studs for ground ever.
The amp is not UL rated and would never pass that... but,, it's and easy fix.
I changed just 2 wire colors. Orange and black (from gray)
My new PS board has 2 diode clamps across R1(a mod. too) , Far Circuits design (not seen below), this is to save the system if R1 blows open. I used these 2 diodes for this pair,  78-BYT78  (mouser) 
R1 has 0.5vdc at 500mA plate current meter drive and R1 drives the meters 200uA needle to 500 ma, on the scale. 

Be sure to Click the drawing to ZOOM it.
Click to ZOOM

PROTECT LEVEL II:  This is for ARC protection on all (non RF OUT) connectors on the rear of the AMP.
CLAMPERs they are called.!
I put protectors on both   (3 jacks,  ALC, RF_input, and ANT relay (KEY line), on the back side seen below. From inside jack tip to ground.
One gas discharge device, GDT and one  TVS transSorB  device in parallel on each jack.
I now have 4 GDT devices,  the above 2 and now 2 on the V1/2 cathodes.  (all non A.C power cables existing the amp are protected this way)  Gas protection IMO is 100% needed on tube amps.

Seen here GDT1/2  at mouser, my project  , Access ID:   is:   384614ba34
Pair them at the rear of the jack inside, center pin to ground. (cost is $4.20 total)  2 pairs. (your Insurance policy)
 
See those 3 red arrows, modify and upgrade this amp so that , at no time can those three JACKS, ever go to high voltage, and WIN.
I replace the 3 RCA audio (phono) jacks with RF jacks then add the clamp devices above. GDV/TVS
The GDV cathode (band) needs to be on the Jack center pin.
I put the above clamps on the 2 left jacks, and a new cap C18 inside, that protects RF input on the right.  Buy the best  cap you can find. (HV rated most high)
The wrong audio jack used, marked ANT. Relay (and ALC) was  replaced with a BNC, the R.F. INPUT will be SO-239 same as the OUTPUT. That green wire is the Softkey Capturing , filament power.
Below is  view, of this tricky mod (simple circuit, tedious to do in such a small space.)
 The new BNC jack was installed first. Then the 3 parts seen here and the blue wire is the softkey board, wire.  (this is ICOM insurance policy grade 1) The TVS is fast but  the GDT is more robust. (the are a good team)
Not seen is the ALC jack , done the same way.


Below:
All new Jacks, ANT RELAY. ALC both BNC and  On the right is  new SO239  jack.  SEE what that means here.
No more "RF-input" RCA silly audio jack plug falling out and insane SWR to your ICOM. BNC's lockup and are 10x more better.
See V1/V2 the horrid pin 3 grounds, it's now  moved, I drilled new holes(tubes out and safe) and make the C14/15 leads 1/8" long, super super short. 200pf caps leads.
Path to RF ground, this will end any possible VHF out of band oscillations. (do the W8JI Stability tests on his web site.)
The less inductance in those GRID  2 cap leads, the better !   (200pf)
My transformer crushed the base some what, so I corrected that with wood shims and a rubber hammer , gently. and now near perfect. That coax is now replaced with mil spec. Teflon. (grade M17)
The small blue/green wire is the softkey wire. (this wire is a power source to the SK201)
Using a washers I hide the old slots for RCA audio jacks. Same trick used on SO-239  (the Tee BNC "ALC" is just so I can find the correct jack in the dark or by blind feel., no mix ups, TEE = ALC)
RCA jacks are for Audio not RF and fall out super easy, not good at all that, ever.



Next, was "Nutserts «" , no more silly wood screws called sheet metal screws that like to work loose or strip, with new SST, pan head screws. (from boltdepot.com) That green spray paint to left is Walmart, best match. Krylon cover max. # 9164
The paper is me marking a resistor that reads 100% off.

New ceramic socketes, mine broke a filament pin , and no more glowing tubes , new sockets saved the day. See  my NUTSERTS here (long winter projects done)

Ceramic sockets with gold pins. and for sure the highest grade GRID caps and for SURE the shorted leads I can attain done here, nothing works as good as this, I even moded the ground hole as close to the pin as possible, it is not pretty but gee very low parasitic inductance WINS.
The caps are from MOUSER.com ! (Solder the pins with TUBES OUT)
That 200pF cap , see how short I made the leads for super grid RF ground?
Do not do as hacker do and use hard and direct wires to ground on the grids, and delete the CAP.s.  This kills dead  all ALC BIAS. (local and external go dead) A really BAD  thing to do IMO.


Next is to remove that silly 12 foot power cords and go with a short 3 foot cord.  (less voltage drop the more happy will be the Filaments and more WATTS to the ANTENNA) My line drop was only 2 to 3vdc at full power DRIVE.
I have new 30amp service power to rear of the amp (10AWG, to the breakers)
But the full drop on weak 20amp service, was at 108vac, and is not good,  my service load center is at 123vac, I now have 123vac 30 amp service to this amp, with 12awg cord to the wall jack. (my 15amp service was no good, for this amp , at all) CURED.
The transformer loves real 120vac, not 100vac and next is , on topic .  I also did the AC surge mod.


Here is the SK201 below. (the famous SOFT KEY CARD , and great deal )  "my name, the  key buffer card"  (one goal is to keep 100 to 180vdc out of your very expensive transceiver (solid state)
The tall thing there is a MOS FET, that is a solid state relay, to switch the -130vdc  key TX relay. (the ICOM killer voltage )
This line is -130v not keyed (real), and -60vdc keyed, as measured at R2 (supply bleed) resistor.  (like all voltages in this amp,  are effected by line voltage, and the turns ratio of the main transformer)
The 2 pin terminal strip (relic part in the kit) was added,  to the original . "Q" strip ground lug.   The 33ohm (moved) resistor goes to the new 2pin strip, insulated side.



Next, that HV blue wire to the right ,   , I use. Teflon sleeving on it now.  and a fuse and a 15ohm glitch resistor,  that topic covered later...
All wires subject to over 100volts , I use actual Mil-17 grade Teflon wire. 
I also  Teflon sleeved (3mm ID) the HV transformer 2 wires  800vac to PS board, and the blue 2kv line. (I don't trust the old HV wires and the possible jumping to new main capacitors  case super thin plastic covers)
 
See before,  looking back to the cardboard Cap. era,  SHOCK !
See my SOFTkey card , installed. and its 2 pin barriers strip with 33 ohm resistor moved to it.
See new 
coax   upgrade to Teflon RG400/316 coax. (2 sizes)
My slide show shows everything I have done all photos on this page are snippet samples. Below photo is before the Glitch mod. All old electrolytic caps also all changed for new. The 2 brown caps are now gone for new ones low ESR too.
all photos are IN progress, not done.

SURGE  worry, Inrush GRUNT sound? ....    
Some  don't think it's a problem, the amp ran for 50years and is still good, what is the issue here?  (well the transformers are OLD/RARE/and expensive)
It's only 600va transformer,sure the filaments suck huge amperes of current  dead cold. (as do that huge bank of huge HV CAPS, infect they KICKIN FIRST)
But in fact surge can damage the windings of any old transformer,  , so  protections added are good. (added) and super simple to do with thermistors.(NTC,= negative temp. cooefficent)
If when turning on the power switch you bounce said  switch (finger does that or the switch wants to) you can get this surge effect easy. (the fickle finger of fate?)
There are 3 cures: (besides a  scope and the fastest EYE and finger in the world, yes pun!)
  1. Thermistor. (simple to a fault and 1 part)
  2. More stinking relays, (not me, not ever) and banks of resistors to do that. stepped.
  3. CPU fix. take a ARM CPU ($1) and program it  to do soft power on, at the exact AC sinusoidal zero crossing point and have it fire a SSR, Solid state Relay. (nothing works better !) for sure on any full gallon PA.(or bigger, Commercial rigs do this)
The only trick with Thermistors (types picked correctly)  is not getting filament voltage sag when transmitting !

Ametherm (maker)  the classic,  NTC thermistor (NTC means it lowers resistance radically the hotter it gets) Sold at MOUSER.com
This NTC device is nearly a short hot and mine are  4 ohms cold.   It runs at 100║ f in standby, (not keyed) then gets real hot transmitting and has no drop.
This SB200  has 2 modes, Rx and Tx   Rx = 1 amp, Tx = 10amps rms.  The key is to keep the THERMISTER  hot, in standby. The factory maker to me, recommends this part # (by their engineer and he gave me samples)
Datasheet:  SL22 4R014
What could be more simple , as all OLD CRT TV's uses for 30 years. (as the filiments warm up this thing gets hot fast and goes to super low OHMS))



The AC feed with 30 amps service makes testing for sag (v drop)  super easy,  now that the wall jack does not sag , omg do not  let that happen)
With this part installed, it has  0.5v drop  across it in standby, and that jumps to 3v drop as you key up 500watts TX , warms it fast, the drop is near zero quickly. (0.2v)  in say 15seconds, With 30Amp service real, I get almost no line voltage drop.
I checked power out drop (none) and HV 2700v drop and filament drop. none,
Mine is cured for $2 (got free samples from factory)

Some complex electronic solutions to surge are hard to do , lacking an internal 12v or 24v power source inside the SB200.  (only the 6.3v (actually its 3.15v) is safe to use but way to low. ( I added 12vac transformer to my SB200 and can do what need)
Inrush is one of the most complex topics of all, depends on the core of the transformer, (all things same)
What matters is what it takes to saturate your E-I core, but it does. for 8mS, about every 10th power on, cycle. (  if you are unlucky and turn the amp on 10 times , 1 will case a huge surge ,even heard ! as grunting sound. )
MFJ makes and external power box. (it is just pre-charging device  , simple and effective)
That simply uses the delay of a relay, to miss all of the nasty  8mS  inrush moment, and some of the filament surge,  that is only 3 amps amps on the primary side (turns ratio makes that true at 19:1)
The minor filament surge  lasts for only 15 cycles. or  1/4th (.25) seconds as seen in photo 3.
The HV cap charge  surge in not large and shown below, in total. (scope views) (and gentle)

The big surge (called Inrush) everyone hates, the "caboom"  sound, is really FLUX overload surge, the line switched on at zero crossing !
This huge surge is the current from charging up the whole EI transformer core 2 times and hitting saturation, and violent,  this takes me about 1 in 10 tries at the power switch.   and BAM.

Now the experts talk (not me, I'm student , of this IRON core... )
"
Inrush current is directly proportional to the value of magnetizing flux and the flux is lagging at the angle  90║ with  the  supplied  voltage  at  the  time  of  energizing  the  transformer  the  value  of  Inrush
current can  go  as high as six to 10 times, the magnetizing current.  Due to saturation of the  core  of transformer in positive region,we see the high inrush current value
."

One more good quote:  From Mr. Warp speed. (Colors by me) Quote:
"The problem is that the magnetic core can go into hard saturation, leaving only the dc resistance of the primary winding   to limit the peak inrush current over the first few cycles. (mine never in-rush surgers,  over 8mS)
The way to reduce the problem is to use a magnetic core material with a much softer saturation characteristic, and design it for a lower operating flux density."

This happens at the zero volt crossing, and is counter logical, seems counter intuitive for sure to most hams , me or anyone not a big transformer guy, but the physics wins all arguments.
Then he said.
"Maximum peak magnetic flux is the point of zero change, and that corresponds to the zero voltage point.
If you switched it off there, the remnant flux remains as residual magnetism in the core.  Next time you turn it on, if you try to drive the flux further in the same direction, the core will saturate.
It is confusing, because transformer magnetizing current (and flux) are always exactly ninety degrees out of phase with the voltage.  When one reaches a peak, the other is always at zero. "
unquote...
The curve is here,  the dotted box seen here is the normal zone for the core, if you go into saturation bam, see that current go far to the right, BOOM "IN-RUSH". Graphs are so cool one can see clearly here the radical nature of core flux saturation.
Now we know what core saturation looks like,  SURGE INFLUX currents.

 The solution to inrush can be complex ,or simple "get a bigger fuse, or don't turn it off at all, and ham it up more... or buy an NTC and end it."
Last quote,  and  very good, it's not true todays gear, he is saying:
"The remanence in a modern transformer core is low and the state when disconnected usually doesn't have much influence on the inrush current.
It used to be a "popular" explanation several decades ago. It may have been valid then, but it is not now."  (he says trash the XFMR, "hi-hi")
The probem here is mostly legacy transformers.


The flux currents are not known or documented (endless internet scans... nada) , in the SB200, but I got photos of them now , lots, but is seen below ( takes me 10 to 20 tries to get it to land at zero crossing and bam !)
Now I know it is real for this PA.
Key point here, you must turn it off at the wrong time then back on at the wrong time to get the hit, ( it's that rare) In fact, I may have never seen the real worst one., but have  fun trying and learning how this surge works.
One great page on topic is here.

I cured my amp for $1. one NTC


Surge data: (tests)
My test R1  is located in free space .
Graphic 1:
 This is now working perfectly the in-rush surge is on the order of a transformer 8 times smaller, that is what the 4 ohms does.
The C1 and C2 caps need to be replaced with CLASS X1/Y2 rated UL safety cap.'s,  so of they can not short out,  sitting on the line power 24/7 .   The correct caps are sold at  Mouser, cheap.
Cap. maker  Vishay #
VY2103M63Y5US63L7  (SEE R1 on the right , bingo)
  


I tried this to measure surge and failed mostly. did ok with standby and the filament warm up photos.
Photo1, my test current loop probe.  CRUDE but fun to see the huge spikes!

 
 Standby current using above device. 1amp rms (photo 2)  ( not much, at all)




This is filament warm up,  my guess 5amps and dropping to 1amp. (keep in mind filament current is 19:1 ratio from them to line)  so  or 60amp startup current (I= E/R 6/.1) and then 19times less on the primary.  (not earth shattering that)
IMO filament current is water off a ducks back , zero worry (in fact messing with power feeds, invites full time low filament voltages) Not only that, but...
The filament secondary is limited to 8amps not 60.
 So the filament surge is naturally limited, the limits of the secondary winding max current and is now here near 60 amps. ever., so FILAMENT current IS NOT AN ISSUE   ( my power line is overloaded so this profile is slower,  1/4 of a second long...)
So avoid scams sellers selling complex HV  cap surge or complex filiament antisurge hacks,  the iron core is the enemy.
Photo3:
see the filaments get hot and lower current as temps rise?
 
 

 

Next ,is the SURGE  (influx) and is real, not the above minor filament currents above. Below  is something to consider,  not any of the above.

This fast rise time throws my detector for a loop (no pun), due to the laws of Mr. Faraday. (  Delta time is the divisor in his famous equations...  and I have not a clue as to how this sensor scales at this huge surge spike, fast rise time it is...)
Photo 4:
BUT !, with my R1 NTC installed , this no longer happens!.  (my wild guess , it is a  50 amp spike  !)? One in ten tries, and only if the core is already near saturated. (mad man at the switch I am)
this loop detector, will show excess levels, with crazy fast rise times, I've not a clue how much.
Last the tools.  I used these tools to prove the scope amp,   I do not have a $10,000. In-rush test system, nor want one. 


There are like 10 ways to cure surge , and all but  one, floats my boat and is the zero crossing detector , method that prevents power on , at the wrong time (at zero). Nothing beats this way!  Prevention beats protection.
 The $1.50 fix is what I used, R1.
What did I learn, I learned that to do this correctly one need an expensive, lab equipment that allows me to measure surge, (an switched) synchronized to the line crossing points (60Hz),  other wise the surge is rare and my current probe saturates too easy.
New modern transformers do not do this bad act.


end surge. (and cured because , I never hear that funny transformer odd sound now... ever..... at turn on... its like mouse now, silent..and fan whispering to me..)

My Power Supply PS  card,  UPGRADE, and refresh all electrolytic,  or how to make 2700vdc voltage doublers safe and last into the future,  (no electrolytic lasts for 50 years, we never run them this old ,ever....)
My old brown PCB (printed circuit board)
The old card is hard and brittle , brown color Phenolic (cheap), and is not FR4 flame retarded, not a good thing to have in any HV supply. I think its , (CEM) phenolic paper sandwich PCB. (dirt cheap and low performance) its not gone , and FR4 epoxy resin modern PCB.
The cap pin spacing is 0.40" (10 mm) Center-to-Center ,  and cap pin diameters are .080 (2mm) on PCB.   I matched that with Nichicon's,  (they will out live me)  I'd never use, China no name counterfeit CAP's in my gear, ever.
Nice how you can see back side traces.  Google FarCircuits or KR4HV , call sign searches you get him easy, and his spiffy lil board (PCB) Printed Circuit Board.  (the other makers of this board raw, want $20, some makers don't sell the raw card)  LR4HV ships FAST!
CAP.'s:
To be  loaded with Nichicon caps higher capacitance  (I used 47k ohm bleeders, plenty enough for those great caps)  R11a & b are 350 ohms, why ?,because  700 wire wounds,  are not made now. they be dodo bird like.
"Aluminum Electrolytic Capacitors - Snap In 450volts 330uF 105c 35x35x10L/S" (note high temp spec...)
The same is true on the old leaky , 1964 electrolytic caps, that leak (resistance leaks not boric acid) like sieve then and worse now., and the bleeder resistors prevents any one cap from seeing,  1130v, just because it had the bad fortune of being the only perfect non leaking cap. (grin)!
One  more point  the 1964 caps way back then, loved to deform (inside) while parked a long time  (no power) and LEAK. (we cured that, but many China cap cloners, are clueless to this., wiki that, they are infamous... for that and for exploding... or leaking boric acid out.)
I now learn the old caps had a weak max ripple spec. (and would be surge limited at that time)
The diodes also leaked then.
Really , its all just bad news here.

Diodes:
 We see others only use, 8 diodes total. now on there  PS board kits, (diodes 1N5408 are better after 50 years) and you can do that here, by skipping   pads.  (only possible with the glass packaged diodes)
This all glass diodes are 57cent each,  78-BYT78  (mouser)   and fit better bent. qty 16 +2 (he has 2 more protecting R1 and your meter and other things, if R1 blows open)
The total series PIV must exceed  1130 peak.  (800vrms x 1.414) (across them all)

Back in 1964 , 1n2070's these diode were 600 PIV rated, but were very leaky (reverse leak spec. and at temp.), so the 8 diodes per bank was used) "keep in mind when HV diodes are reverse biased, the diode that does not leak, gets all the voltage drop"  < no good deed; goes unpunished,! hihi
The non "A" suffixed 1N2070, had 4 times the reverse leakage , of the "A" at 200 uA (non "A"). (a lot)  the NEW 1N5408s are basically no leakage 1000 times less.(and over 3 times, more current spec)
Another good choice on diodes are classic 1N400x series, 1N4005 to 08 all work. here. (07s are prime bin one parts,. hint hint buy them , they are the best grade)

The PIV rating needs to match incoming A.C line spikes too, added. (like you see, as your air-conditioner cycles off and on line over and over, as they all  do,  and the inductive EMI noise spikes from it (QRM) "don't cheap out on diode, ok?"
When I see these old circuits,  I go, gee what would we use today,  answer low cost 2kV,   750ma to 1amp diodes, just 2. total. I see Microwave oven diodes like this..., but the data sheets are in Chinese. (hopeless) HVR-2X062H0A ($6, be fun to try it ?)
even 4 diodes in series is 25% here,  below 0.2uA leakage, 1000 times less than 1964.!  I say 2 diodes each side would be ok. (but really I used all 16, a knee jerk lazy upgrade)



Many after marketed boards sold , use  just 4 diodes 1N5408's is 1000 PIV spec, 78-BYT78 are better and are glass cases, and fit 10x better here.
Here is the BOM Project manager  from Mouser. I made.$59. (it's not cheap because I use prime parts not China dirt cheap counterfeits) "I use TOP MAJOR BRAND PARTS from JAPAN"
The diodes fit too tight on this cloned board, but I got them to fit bent at sharp 90degree's  (not good sometimes as that can crack the silicon. crystal inside)
The limiting factor are the 35mm cap bases, eats up all the space. One cure would  be 100k ohm smaller resistors, and lower watts, and use this space. I found narrow resistors above.
I tried to slide in just  4 diodes on pads left to right, and 4 more  D9 to 16, skipping every other diode..  but the hole spacing is too short.... Unless you use the smaller glass diodes.
 Ignore D17/18  I.D. call them D19 and D20.  he added those. These 2 diodes keep the meter from blowing up if R12 blows.
The diode spacing is all wrong for the  plastic diodes,  so use these 78-BYT78  (@mouser.com) Glass diode package. They are the best made.! and tiny more cost.

my last order. change 10s to 1., I never order just 1 cheap cap, I like extras...

As a rule , ceramic and mica caps don't go bad.  but I am a chicken , I replaced C16, V1,pin1, 0.02uf,  (1000v)   I take cascade effects seriously. (aka; the domino effect, and horrific)
 I don't want HV going it to my ICOMS RF out port ever..,  so sure I'm a chicken. and my be paranoid,,,  but it's my 30 cents. and 5min work to do.
I now have 4x  Gas Discharge Tubes - GDTs / Gas Plasma Arrestors... as seen in my above schematic.
The famous 2700v fuse and Glitch resistor mod: (Mr. W8ji , likes it as do others and Ameritron)
Not my invention, most what you see here is not my ideas, or my art, it's all from many others work. (ALC is MY WORK as seen only here)


Why use this, solution, to save parts from damage , when the gassy tubes arc, as most do.
Good tubes are not full of gas and will not ever arc unless something wrong happens, as seen here.
Last,  is  the famous  HV fuse mod  and  surge (glitch) resistor mod, same fuse line mod. photos. (It has some safety value, but its not panacea, a surge can happen that is faster than any fuse.)
Most web pages show no sources nor any part numbers for correct parts, I try here to correct this error.  (2700 volt parts are not trivial)
This glitch resistor is 175p pulse rated, 15 ohms,  10w? 1kv  handles 10x surge. As seen in a  Kenwood TL922 Amplifier  (Sourcing from..mouser  and OHMITE #40F15RE)

I'm using HV  1/2"  (12mm ceramic)  stand offs (old Drake parts, NOS new old stock parts)  to hold my new glitch-ister.  like this  holding  Glitch Res.
This mod also helps (just pull the fuse)  you work safer and when wanting to run the AMP with no B+ 2kv on the TUBE plate caps, and not wanting to pull the tubes risking them to handling damage.
Case in point checking live power on ALC, or KEY line -130v feeds and grid bias or filament voltage,  or other live tests (V1 and V2 dead cold)
We can pull the fuse and see if the HV board works. The Fuse only kills HV power to the V1 plates areas.
An Update here...., use the chicken stick only to the left side of this new fuse, or Chicken stick will  blow  the fuse. (power off)

This fuse is driven by  the above HV PS board.
 If a vacuum tube (aka; "Fleming Valves")   shorts out or arcs? from GAS, then  the fuse blows saving the board and the tube that arced.(in some cases, there are no guarantees with realities of HV)  The gassy tubes will show huge grid currents.(shut it down now)
W8JI , says.... and I link....
There are other causes of arc, seen here.

Problems with 572s  are here,  a really great study.(note the operators errors comments, bands switching hot, running no load, etc.

Amplifiers RF PA in general.

Using 250 to 600v fuse holders is not correct or safe, I use dirt cheap microwave oven fuses and cases, or even removed from one, I found the stock fuse blows easy, so now using 2Amp FB (fast blow, fast acting)  standard fuses.
The fuses go to the blue wire on its way to the RF dog house, main RF TUBE CHOKE.  
Those are 2 ceramic 1/2" tall stand offs,  threaded at 8-32 size. Black Jumper wire is me testing many NTCs to see which works best....


My OVEN mod:
The microwave oven fuse case (5kv rated !) is my idea, the glitch resistor is widely done, and recommended, by all, but  be ready for Mr. GAS then ARC !
For sure using this holder , is best. at 2700vdc,
The uwave fuses some  are sold with a too low amp fuse inside, so change it   out,  6x40mm is the HV class of fuse, used in the ovens.  The 1.0 amp works ok.  (0.9a maybe ok) I have not tried all sizes to see which is best....sorry!
The fuses are extra long, and regular 250v fuses are not a good choice,  they can blow and then arc. The spring inside, is a really good idea for HV, so the gap at the blow point, is wide. We want it to blow and NOT ARC after.
The 2011 ARRL hand book shows the convention here is 10,000v per inch , rule .  (clean, insulators)  so 1/4" for 2500v , so that looks pretty close to the gap on this fuse blown.  In fact the same book says there are no fuses like this sold but, below is my exception.
Do not chicken stick the HV supply to the right of the fuse seen above, or it will pop. (HV still high and floating lower , power off now)
As seen here. (no usa sources, sorry)

  The uWAVE holders are the bee's knee's.(super easy to find and buy but can take long time shipping from China) I think all these ovens are made there, or built in Mexico with China parts.

The glitch resistor is right next to my uWave fuse holder. (later, I put plastic shields over the big resistor,using tie warps.)


The HV fuse belongs on the hot side. (if at all)



ERRATA:  As seen on web sites: (what not to do)
The wrong place to put it FUSE, is in the cathode ground, in this case there is no real cathode, the filaments are the cathode, and the 6.3vac transformer is CT grounded,  but if you open  that line and the cathode fuse blows, you blow the Softkey , card to kingdom come, boom.  (and more !!!!)
Then goes your $1000+ ICOM (ant. relay line boom 2).
This ground is the HV Tube, DC current path, never open this path short of removing a vacuum tube. Bidda Bodda BOOM,  cascade failures.  (It will not make your day)
Then it may blow (short out ) those poor filament winding in the transformer, if the fuse was connected to CT and ground. boom. Those wires are NOT HV safe.
Do not cause or invite cascade failures. 
Same with grid fuses.  ( do not put fuses on grids. (you never want to float a grid, ever....) but is ok to float the TUBES plate caps, carefully, make sure they can not get near grounds. Tie them safe for  test like this with tie wraps.
Nor do you want the over stress the insulation qualities of the 6.3vac transformer windings (fuse added here), that can be a $250 mistake  and a fire hazard, HV does not belong on the filament windings ever.! dp not allow that to happen.
Never ever break the cathode ground on this amp.  (nor grid)
In fact, make it stronger, (it's weak by design I moved mine to a very strong, transformer bolt mount )   See my slide show, and how I cured mine.
There are 2 issues here, grid protection if the grid minus -180v supply goes dead, ( and AMP slams to max plate current)  I did not put in a bias loss detection system. (I think its super reliable as designed)
Keep in mind there is no computer  here, no ARM processor with code to protect the grid or plates from damage. (like on new high end, really nice gear)

Day1,  tests.  (HV means high voltage, and lethal) ( that ,means no second chances...)
  (I did  the overhaul first, but some folks are more brave than me..)
To be clear, I fully expected to find ALL THIS, it's PAR, (this old)
Clean it safely (dead) after you chicken sticking it.  (let HV drop below 100v and then use the ground rod, at the PS, HV terminal point.(or just  wait for 0vdc on the dial (needle) if you trust it, if zero trust with your life, then the chicken stick is the only way to live)
Inspect it for 1 hour, every inch for damage, burns /cracked parts, burned parts, any solder points touching other points, illegally or loose tubes, or broken tubes, do not turn it on, if tubes are damaged, many  are damaged when shipped.
Loose solder joints,  and more below.
Check all  wafer switches for  blow up (burned up contacts,  is first,  a real show stopper, mostly caused by users HOT band switching, running with no load (VSWR way over 2:1), or running full power with no load or under coupled. (tuning done wrong)
Many folks love to use cleaners or other horrors in spray cans on these switches, like DeOxit or WD40 or Acid dips (read the MSDS, it's acid, in all Silver dips)of any kind or form.  DON'T DO THAT, STOP and think of the facts first.
  1.  Doing nothing to the switch is best, if it works now leave it the heck alone. 
  2.  Phenolic (brown) or White Ceramic switches are porous,  any thing you spray on this sponge like material will soak into  it and wreck it, STOP avoid this act. Thank long and hard on how to avoid filling this sponge with CRAP. (the HV parts will arc over and bam, you wafer is wrecked)
  3.  The black scary silver oxides (sulphides) is in fact is harmless and the contacts self wipes the area, so just  exercise the switch and bingo, it's good as new. (won't look new, but is electrically good) (this is part of the design, a fact)
  4.  If when brushing the dust off the the switch (use dry air in a can  first) junk will not come off, a switch , then use pure alcohol, never rubbing alcohol as that contains both water and oil (Lanolin , sheep dip oil) DO NOT EVER USE OIL HERE.  (a fact)
  5.  Never ever use a cleaner containing oil of any kind or other snake oil (endless list) clean it using a safe solvent, like Pure Alcohol, do not use brake cleaner, as it eats plastic to hell, do not play with other solvents here, like , kerosene, lacquer  thinner, MEK or acetone or stoddard solvents. 
  6. if a contact is damaged then you need a new switch , no magic snake oil in a  can cure real damage. (no matter how many 20min ads on TV say otherwise or not sold in stores)
  7. Learn that  adding oil also acts as a dirt attractor (like used all oil soak air filters for engines, see?) so oil is bad 3 ways, it's and can be an insulator (like transformer oil is), it attracts dirt and then soaks into the porosity of the wafer and wrecks it. (oil is dead wrong here)
  8. Keep in mind that the switch is not sold now, so making it worse here, is a royal pain. Tread lightly here and win.  (and the band switches,  tread 10x more lightly)
  9. Do not use wire brushes on any ceramic parts , switches. etc. only organic or plastic.
Pure Alcohol is safe for plastics and ceramics. (let the RIG sit over night to let all this alcohol evaporate, or use a heat gun if in a super hurry, best is don't be)  (all paint stores sell pure Alcohol, even the big name big box store does)

Next is really #1 , number ONE !
Find the bad grounds, that I found in mine, all 10 of them,  2 are very very serious... (safety ground and that bad to the bone, 6.3vac center tap ground for the transformer is NO GOOD.)  I created what I call a NEC/UL ground screw, that is perfect, and moved the transformer center tap.
In order, I correct green earth ground first, then the transformer center tap ground is removed from the silly plastic stud (that can strip so easy) and moved to the transformer rear, main mounting nut/bolt, its huge and strong and will NEVER fail you.
You do not want to float the  two TUBES V1or V23 cathodes or there grids, ever. The filament CT ground must be THE BEST ground. Fix that now.
All #6 screw ground tabs (star-tabs) with that are horribly loose (many) and totally unsafe.  (and this mod to RF cap  grid grounds and make the Cap leads shorter)
 
See my slide show above, for how I cured the 3 biggest grounds (G1,2, 3)that were bad. (safety ground, line power, and 6.3vac center tap ground horrid. and last the grid caps were 4 inches long (1" time 4), combined length, crazy long)
Spend most your time on grounds, to be safer .

Next?
Check the stock R3 main bleeder resistor  ohms now , after a full chicken stick actions and see  15 giga-ohms. Do all dead volts,  ohms checks seen HERE:
Do not power it up , configured for 120v and then plug it in to 240vac.. boom city. (or forget to look first?)

Carbon fail. (all carbon comp. resistors failed in min, 9 of 10) this is not new to me.  ( A life time of this, I have) If those funny round cylinders have color bands they need replacing or checked super carefully.
New plate resistor/inductors, suppressors, are usually needed, with new tech.47 ohms 3 or 5 watt metal oxide, , no snake oil  magic suppressor coils "as seen on TV" , no U coils,  the stock coils are GOOD ! Just the old carbon comps, love to go bad. Just replace these  resistors.
I am told the 2 grid resistors love to fail if you run China TUBES, test them now.
The real fix for VHF suppression is here. (first bark up the grid tree ?)

Make real sure 100% sure  those 6 main 450v caps is the Power supply are new, not 50 years old, electrolytic caps almost never last 50 years.  (they will be dry, inside,) there are 2 more on the bottom that need to be checked, I use a  ESR meter...
After you are sure the AMP is safe to run.
Make sure the fan is working.
At first run check that the fan runs full speed, I oiled the 2 bearings first.  A larger fan doesn't   help much, as the tube is and IR emitters, and only painting the dog cage black (dull) helps (eats IR).  That is why the fan is so small that , and noise.
Run it with tubes installed and plate caps off for 1hour, this can help get rid of gas in the tubes, if you have the HV fuse mod, I can run it with no HV easy and do this. (chicken sticked first)
If you want to test it with no tubes at all you can, to be sure the HV is working good.
Run it with no tubes,  and HV caps  safely held back from hitting things, see if it can cook it for 1hr, and not arc over or catch fire.
Then run the tubes hot. with HV present and connected, caps (anode) back.
Make sure the HV is near 2700v, (the meter is not NIST instrument grade) when transmitting the HV drops to about 1900vdc, this is the normal loading effects.
It's good to cook this for 1 hour too. (at the least) The tube can remove its small amounts of its own gas, if you just let it cook in standby, not transmitting.
More checks, below:
Power off and HV at 0 volts.
Next, measure the R16 , 33ohm (org,org,blk) 2watt power resistor on the bottom for  minus -130vdc (typical and unkeyed)  
 This wire is the yellow  wire to the softkey option.  The stock unmodified   AMP , the "ANT RELAY" jack is at -130v but after the  softkey card  mod, its 1v or less, (they whole point of the softkey is this fact,)
This is grid bias, do not run the tubes with zero bias. or they will burn up. This -130v is standby bias to the TUBES grids. Check the 2 tubes grid resistors read 33 ohms.
Keep in mind the -130vdc supply (bias) is very very reliable with a new filter cap there,  if it shorts, all bets off.
Make sure the grid bias resistors are not burned up  33 ohms they are, 10% (plus or minus)spec.  the grounds for both grid 200pF caps must be tight. (see my MOD here)
If not trusting the 50 year old panel meter (wise), use a 3kv rated external meter. Using same rated test leads, I have  HV meter probe (Heathkit) made for TV CRT and use that. (its tells me hot or not hot, only)
Check high voltage (max is peak of 800v RMS x 2.)  1131vdc x2  = 2262vdc (line voltage can change this) (from voltage doubler) there is no load now, so, will be max of max.  the 50year old meter in panel is not very accurate. 10%? at best.   The PS end of my large blue wire is HV.
Do not use external meters , test leads, and scopes not rated way over 600vdc. most meters are limited to 600vdc only.  (checking the HV circuits)   You can buy them , so do that, if wanting to play with HV.
Once you trust the front meter to show 2700v, then you now know when it's at 2700v, it is , but  not just stuck there mechanically, be aware meters that get metal filings inside (steel wool?) it can stick at any time, even at 0v, and be a huge dangerous lie. A digital meter upgrade is more safe. IMO.
The rules for using external meters, scopes and doing advanced live testing:
Make sure the power cord end to end , that the green wire  is working first.  Check it very very carefully by eyeball and ohm meter to best  sure this path works.
If working on the older rigs, the green wire is missing and is 2 pin plug, time to upgrade this gear to 3wire or just add a ground wire from SHACK ground to the rear of said gear ground screw lug seen on all Heathkit gear of this type. (Ham gear)
  • Dead cold ,  (how to use external instruments safely)
  •  0 volts  on meter (Dial at HV), chicken stick it now.
  • Connect meter rated over 2700vdc and it's test  leads same rating or more. Try to learn now, that if the schematic shows, say  180vdc on a NODE , it might be 2700v NOW. (a failure of 1 ground and you are now dead)
  • Flip power switch , not touching metal, with one hand , other hand in a pocket.  Rubber shoe shoes are win here. (eg. not wet leather)
  • Look at the panel meter on HV select, if not at 2700v, now , we test and find out why now. work this failure first.
  • Look at bench test meter, or scope.  (write down what you saw or memorize it)
  • Key transmit if you want, take more readings.
  • Power off, let voltage drop below 100vdc on the panel dial. if voltage  fails to drop on panel meter , now, then chicken stick it and repair the system bleeder found on the bottom of the SB200 now. (this gear is unsafe lacking this safety device)
  • Chicken stick it again. 
  • Repeat as needed.   Best is not to touch meters hot or there wires while power is applied, working any HV gear or rigs.
  • Keep mind all this is true on most tube gear,  100v, 300v ,600v and 800v or more are found in most tube gear and all are a danger to your life.



Rule 2: no rings allowed on fingers and  no watches, on arm, HV safety first. (conductive tattoos? hihi)
 
If the panel meter ever acts funny, do not trust it ever again. (until you can prove otherwise) (funny means, sticks low, or sticks high, or sticks say at 1000v) Old schools meters love to stick , be a skeptic and live.

To use  external meters of a lessor grade, use voltage dividers, just like this AMP does, at its meter. SEE R4?   and R3?  if you measure both resistors ohms and are sure they are good, then R3 voltage will accurate as a test for HV. (measure it's voltage, the math is easy)
Let be clear here, if the panel meter shows 0v, or way low, I power  wait for full 2 minutes min., then chicken stick it and I check R3 first. R3/4 are correct ohms , and R3 is grounded then hot , hands off , I  check R3 voltage,   its a 1000:1 ratio , so well be 2.4v for 2.4Kev on the main HV feed.
If it reads 2.4v and not the panel meter (reads low) then the meter  in the panel is bad. (in some way) this mode of failure is a man killer,  STOP AND CURE THIS HARD FAILURE NOW.
Try to know that gassy arcing tubes blow up the meter (no mod.'s), so meter is always suspect , until meter MOD is done. 2 diodes anode of one to cathode of other across back of  meter terminals, added 1N4007. (its called a 0.7v bipolar clamp , in electronics)
More mods?
Do not use 250 or 600v rated fuses/holders on that HV bottom blue wire,  that must be rated for  2700vdc. or more. 3k is what it should be or more... ( I see others adding insulation to 600v fuse holders,  now that is ART)
This is why I chose the microwave oven fuse (and sold everywhere) , the assembly, (looks like tiny coffin  holder at 5KV spec.)and super easy to find and buy dirt cheap on Fleabay. <$5 and is 5kv rated,  for near nothing price wise.  (I'm using a 2amp fast acting  fuse)
Do not use cheap wire to replace that large blue HV wire, use HV rated wire only.  3kv +  (silicon rubber wire  or special Teflon) The below rules help decide what you can't use.
When first applying AC power, do so with no plates active. (anode power off, or caps of top of tubes)
Make sure the filaments are at 6.3v . (HV defeated checked) (CT 6.3vac means , 3.15v on each side of ground  ,center tapped is that) (do this test with wires connected transformer to meter, then hands free, power up , look only, then power down and chicken stick it.
Okay all done with mod.'s on my page , I'm in the end zone now, touch down time.
Real testing... 100% assembled.
Apply HV, now,  then make sure standby plate current is at 90mA.  if 0mA there are serious problems to address.  (mine had a blown HV fuse, chicken stick error , blush) other causes are grid bias is wrong. or and open filament ground (that pesky transformer CT  ground is bad)
One can test with 1 tube only,  then switch to other one by itself, and see if they are equal, with same exciter drive.
If only one tube is tried, 45mA? then reverse them running 1 tube?  (poor mans tube tester mode here, do not fail to use the Chicken stick each time, each test.)
 Doing all this with  power cord pulled and chicken sticked each time, right? and letting panel meter drop to 100vdc first. (takes 2 full minutes, mine is longer due to larger CAP.S)
If the tubes show gas?  (looks like red/blue/purple clouds inside, then try to cure that here.
If really paranoid, connect a 1kw dummy load and a sweeper gen on the input JACK (now a BNC) and sweep it 1hz to 100mhz, and see what it does. W8ji has all this documented on his web site.
Ok, seems like the amp is ready for show time. See my data here. (I will keep adding more data now)
I just hit 625 watts RMS  peak power (scope seen), in CW, today.(with 1964 tubes)
Last on the list is how to tune up the amp, well, the manual is free page 43-49,  clear as a bell, 
Once you have tuned it up, one time, the next time is easy, here is a short list:
Mind set, try to assume the antenna has problems, and win. I ran  a 1 million watt (peak)  R.A.D.A.R and we never assumed the antenna was good, ever. we started low in power and worked up, every time.
  • Set exciter to 20w (20% is same on my ICOM) (respect the gear, and win)
  • Set sb200 to off
  • Set freq, and band on external tuner. 40m? to a non busy (in use) channel. (no dummy loads, we are wanting to load up a real antenna here, on mine , I must select one of my remote antenna's via my remote switch controls)
  • Tune the 20w at the turner. to min VSWR , SB200 , using CW mode with the Morse keyline shorted or Morse key depressed. (do not use AM as its capped at 25w and hair pull time.)
  • We now know the ICOM transmits ok and loading is ok end to end, at this point. (if not? HALT, and correct this gross error now)
  • Energize the  Amp now , tune the amp. SB200. Per its manual. (takes but seconds to do)
  • Next up watts to 70% on icom to where happy, 625w (grin)  Do not use 100% exciter power,  Some amps only want 60w input max. My ICOM does not like over 75%, drive. (no surprise to me)
  • Fine tune the  tuner and amp until more happy. My SWR is so low I cant even read it, below 1.1 to 1. (on my PalStar)
  • Done.  Rag chew all day,
  • Next day?
  • if all was ok now and  same as the  day  before?
  • just turn it all on and fine tune amp and tuner, at full power. (trust in deity of choice , no antenna damage as you slept)
  • Takes seconds to do .
  • Only changing bands takes more work. (if yes, start at the top of this list)
  • So keep a log on dial readings so you can get them close, faster, at band change time.
  •  Try ALC mod, hear and run at 60% on the ICOM and let it intelligently drive SSB to huge peaks, as it was designed to do.
  • This is one fun SSB amp or with CW.


Plate Anode, VHF parasitic suppressor issues: 
 
Many think this device (plate cap) is the cause of problems, but is just GAS in the tube.
(your scope would have told you it was not VHF oscillations.)
If you did have that, I'd first fix the GRID grounds, and what is stated here.  not mess with magic plate cap coils from mars.
I first shortened leads on  my 572b tube grid pins 200pF(old guys ,uuf) caps by large factor.
These 200 pf grid cap wires were super long (2" total each tube), and the factory ground lugs, are in the wrong places.(too far away, 5minute work to move them (drill in hand)
But wait both ground screws were loose as a goose, after 52 long years, bingo loose, sure... but there is always room to make things better and avoid parasitic oscillations.
I re-drilled chassis holes, to the ground "STAR terminal Ring tabs" to be mounted as close to V1 and V2 pin 3 as possible then cut the leads on the 2 caps by 2/3s and re soldered them. 
This will make the grid a more perfect RF ground at all times and  at higher frequencies VHF.
No more grid inductance at the  RF grounds, or leads, for sure way less at VHF frequencies.
The better you can make this RF grid grounds here, the better you will be able to run those off spec. China fake 572b. (that  do not meet spec, in Mu and other gain criteria)
Next: (grids fixed)
The Plate RL cap devices are 47 ohms,  is spec.   I'm using mouser.com,  metal oxide p/n,  286-47-RC that are 5 watt rated.
My new parts are here:  (those coils , measured and  are calculated by me as 50-60 nH. (say nano Henry's) I am Not happy with the carbon oxide frail, skinny wire leads.  
  (read W8ji documents below  on this subject he is well grounded on this topic) hi-hi.
I will do these tests to prove it works right.

If you read W8ji  pages on topic , you learn how to test the AMP for stability and to tune  them for your amp., exactly if you so desire to do so. 
See his tests with a 68 pf (say 68 Puff) series cap in the Plate suppressor.! see his tests, proving the effectiveness there. But needs circuit card to hold a small cap.
My long gone SK uncle Seth (ham W6HNX) RIP would  say 68 micky mikes for  mmf.  (jargon long gone now, like CPS, cycles per second, now Hz)

Many think arcing is this SUPPRESSOR  failing , it is not.  (At VHF freq the coil there, is hi-Z  (X sub ell  is high,  Inductive reactance is high at VHF freqs)
 That  resistor is the VHF blocking device. (but uses a , scope/grid-dip-meter or FFT spectrum analyzer to see if it , is?)
Most hams do not have a full blown high end spectrum analyzer , but may  have a cheap lab DSO scope with FFT button (mines marked MATH)   This method is weak , and many not see a spur, due to poor FFT and filter and bandwidth performance, but you could use a grid dip meter.
This is a very crude FFT feature, but is not useless,  it may see the  Transmitter is producing illegal VHF spurs, the levels will read too low (accuracy ) but you may see them , if they are there, and that is what we need to know, just  their existence or small evidence of them to take further actions. 
Just scan above 30mhz for illegal spurs. (if there any at all?, I'd bet not, if you learn to ground that grid in the correct way)  This is what good looks like .
There is a very old saying told by old engineers,  "Amps love to oscillate and oscillators won't"   Less so today. 
Ok no spurs.
First off tubes not wrecked inside, will not arc, it's a hard vacuum inside and will stand off 8kv (as others have tested and proved) It sure will not arc with 2700vdc, It will not, what you think is arc is gas plasma. gases conduct as they ionize.
What you see is GAS: (killer gas)
Arcing is gas (in most cases, just look at the tubes in the dark see that illegal glowing?) (only the filament glow and that colors, are valid)
Don't forget , running the amp unloaded and the plate voltages will be way too high.  (operator error. or not tuning up under full power, then backing off to where you want to run, power wise)
See W8Ji  pages on running any amp over or under coupled and his arc pages.
 See the voltages that can be present, and the cause say, band switches to arc and burn up. (operator error or dead open or worse antenna)
If you find all is ok, with a dummy load and only the antenna will not load up, then switch to the other HF antenna ?  or use an  antenna analyzer to see if the antenna is really connected? and resonant at desired carrier  VFO.
OK no GAS  back to illegal Spurs:
What causes this, in an amp  it is gain in the band not wanted and then feed back. (the suppressor helps with the gain/band part)
 All tubes like this have a natural  180 degree (unwanted)  feed back path, we don't want, called C strays.(inside the tube and sometimes outside) Anode feedback, to the grid and cathode.
See the Miller effect:  (original and this modern version)
The Miller capacitance in a triode tube is equal to the plate-to-grid capacitance multiplied by a factor equal to the stage gain plus one. or at HAM school.
Quoting G8MNY: (italics mine)
"Common (grounded) Grid mode (drive on the Cathode) is used in some PA & RF amp designs with its reduced power gain, it has the advantage of no Miller effect as well
as applying current NFB (neg. feed back), making the stage stable, quite linear & with low impedance drive."
And J. Hapers great  page.
These natural stray Capacitances, are called parasitics.  (all elements near other elements have relational capacitances)  Even two garbage can lids in hand do. (hi-hi)
The path from plate to grid is a killer parasitic path, as that is the gain path and will like to cause illegal oscillations. ( this path is regenerative so must be controlled)
That grounded grid (correctly done) cures this best. ( if the plate charge can not wiggle the (E field effects) grid, then it can not cause this parasitic path) As you can see the external plate parts are way far away from the grid pin, not by accident.
The plate suppressor  is the insurance policy, 2nd tier) for just in case there are feed back paths causing out of band oscillations. This device, lowers gain above , our HF band. It's a stop gap device.
BEFORE PLAYING WITH THE PLATE CAP COILS, check the grids.
Our grid is at RF ground , but if its poorly laid out (long silly leads possible) , it may allow grid signals at VHF frequencies ,Via Miller effect that should be not happening here. (ah, fix that?)
(the grid will try to amplify this illegal but natural signal and if it does, it can in fact turn the amp into an oscillator. even up to  1000 peak watt transmission, but it will be what the tubes can do , age wise and total gain at that freq. The tube does not melt or explode. )
The tube is  spec. 30 mhz tube, but that only means it makes a lousy VHF tube, not that it has zero gain at 50 mhz)
Hint, use this page to do the CALCS: it free.

Solution 1:    FIX THE GRID FIRST , and leave the plate coils alone if there 47 ohm resistors are ok.  FIX THE GRID First, after all if they are truly at RF ground there will be no VHF oscillations, if seen at all, most don't even look for them. (spurs)
First shorten the 200pF caps on the grids and end VHF oscillations.  (those long leads are super bad here)  at 100mhz the Xc is 8 ohms to ground 3 times better than 30mhz but, the long 1 inches wires on mine are .023uH, and is 3x greater XL at 100mhz (14ohms) than at 30 mhz.
So that means the leads being 1" (times 2) long render the 200 pf cap. near near useless, just doing XL /XC math only you can calculate it. (USE C14/15 leads SUPER SHORT)

Plate CAP (anode) RLC tuning,  this device increases it's impedance at higher and illegal frequencies, and lowers plate voltage and gain as a result.
I was only curious how the factory plate cap coils tuned. (coil wise the 47 ohms carbon is obvious, its 47 all day long at any Freq) Even if the coil went to 1000 ohms Z that  47 ohms sets the gain of this tube, and is deem good enough by Heathkit to block parasitic oscillations.
Just looking at inductance alone shows what happens, here, C stray is very complex here, and I can not to that or total Z at freq. (Z= impedance)

 
The suppressor coil is  about 50nH  (I used online coil calculators, there are many and all agree)
If your scope or grid dip meter shows, parasitic VHF oscillations from 50 to 100mhz,

At last resort,  add one coil turn to the suppressor, but be warned it might drop gain  on 10meter band a tad.  (there are only a few prefect suppressors made, and takes a RLC device, not just LR)
I did this just for fun to see what the numbers  were, just for the main inductor ..
This plate inductor is 3 turns (3T)  of wire  0.5" long , .45" diameter, and is about 0.05 uH (50nH),   (measured carefully to  run the numbers)
(4 turns (4T) would be  .12 uH, which doubles it.  and I think is too much)
A 1 inch diameter  coil same spec. as  3T, and .5" long, is .24uH , so you can change this very easy. No need for magic inductors ever. Just this:
L= (d^2 * n^2)/(18d+40l)
Mhz          ohms           R/XL   ohms    (R=47)
Freq
3T(XL)
Z for R&XL
4T XL
200
75
29

100
38
21
151
50
18
13
38
30
11
9
22
7.5
2.8
2.6
 
3.8
1.4
1.3

As you can see, the plate cap resistance changes the tube gain, and reduces, fast above 50mhz,  and that 1 more turn of wire, huge changes....
The above calcs do not attempt to do stray capacitance values, or stray inductances using real RF network analyzers   (no rabbit in my hat, nor that expensive tool)


Be sure to  tread carefully here at the coil,  as its easy to lose gain on HF bands.
As you can see XL doubles with 1 extra turn of wire, at 30mhZ.  <<< this is a key fact here....
If you read W8Ji pages you see a capacitor added in series, to get max gain in all ham bands and no gain in VHF.  (and is used by some major amp makers today)
 Point of Reference 1:

This part below , (sold) , not easy to find, is one choice, will not survive next to that roasting hot tube, but is near the same spec.  I only did this to show, my  calcs at 50uH are close.
It's a good photo so see what 56nH looks like,  and does look like ours !, if ours was naked.
Sold at mouser.com and  melts at 125C  (shown only for illustration of what is sold now, and not likely to survive, and why they wound them on carbon comp. resistors.

If your scope or grid dip meter is not showing illegal VHF oscillations, then why are you here? (or the scope view or the Scope FFT view)

If grid current is high, check for gassy tubes first.

PTFE (Teflon) facts: ( material facts not UL or NEC etc)
The wire is silver plated, and a super good conductor. (and will not break when bent)
200C temperature rated, my 700f solder iron touched to it cause no damage, so 200C is a low spec. or life time spec.?
Will not propagate fire or make fumes harmful to humans (just like the fry pan)
The wire has  insulation break down rating  (truth) of  500 volts per mil,  that's  500volts per .001 inch (a mil is that) of Teflon wall thickness.
My PTFE added sleeving where needed,  is 0.5mm thick (OD-ID = thick) or  (.019") so is 19mil x 500v (or 9kv rated)
The MIL spec version this think is 17kv  )RMS  (24kv peak !)
  So to make hi voltage wire from LowV wire say Teflon hook up wire, just add standard  PTFE sleeving. (simple , no?) (with due care)
My 20ga. hook up, wire sliver 19 strand is 0.2mm wall (or 7.8 mils) or 3500volt breakdown,  UL would list it at 600vrms but they derate it big time.
If the 2.4KV blue wire is bad , try this, wire , called corona grade wire, it will last for ever. (or the 5kv silicon rubber wire)  In my case,  I sleeved all HV wires, in view.
On my rig I'm mostly sleeving the existing 800vac and 2kv wires. and replacing some that is just ugly , beat up, or solder iron damaged, or has petrified insulation.
I replace some gray wires (there are too many) with orange. and black, so I trace important circuits more easy.


THE Tx/Rx relay inside the SB200,  has silver plated contacts "looks black" (black is not bad), never TOOL file them or use  sand paper on them.
It's a Potter Brumfield. Relay. (now Te.com owned and sold)
 They make a 24/48v and 110 volt dc relay now.  see TE.com for data sheets. (using the wrong relay coil, will require a new set of resistors across it or in series . causing 24 or 48v to happen or 110) Lucky me my silver contacts are perfect.
To Clean them ?
 (one can run clean strips of printer bond paper through them soaked in Alcohol,  and never use oil or stinking nasty oil based penetrating water dispersal  WD40 in any radio system)   WD does make now a electronics cleaner but make sure no OIL is in the product.
Just exercise it and it will self clean, most times; the more you transmit the more it self cleans.

GAS 101:(the tube is a hard vacuum, if not?, it is more like a NEON  lamp, big gas is lost hope,  air has 20% Oxygen and O2, kills filaments fast , as per Thomas Edison learned long ago and Mr. Watson.)
Oxygen glows
Blue-violet, Neon red to orange. Pink is helium (rare) but does it matter ?, gas is bad, no exceptions,,, no it's really a mixture of gas that is ionizing like a neon  lamp, nothing but bad. All that matters is that gas tells you ,you do have an air leak.(or worse)
  • The tubes shipped new with a perfect hard vacuum, and pre-getter activated in the factory (some use electromagnet induction device stimulated, to fire off that getter seen inside, see all that silver on the glass inner sides, that was day1 factory deal there)
  • The old tubes can be DeGASSED, by running them with my plate fuse pulled. (or pull tube plate caps.)
  • Running gassy tubes may cause it to arc and wreck the tube further (or far worse), avoid that if you can. (gas invites arc, due to gas plasma effects)
  • Gas is all about bad news, this is a show stopper, stop now and correct the problems with gas. (I'd work hard to find NOS tubes, not from China) Hope you are lucky.
The old parked in attic tubes, or NOS (new old stock) tubes, can leak air into the vacuum of this tube. (and are called gassy)
They don't glow only orange (filament glow) as seen above, they make other wild colors, red/blue, purple are gases IONIZING in the poor vacuum in side.
There are ways to deal with this. I will make no claims on making tubes last longer, as many of these tubes last for 10,000 hours.  (I can't  measure that)

Gas:
Brand new tubes in a factory , are pumped down to a vacuum then sealed and then they run a getter process, that burns off all gases inside, to get a perfect hard  vacuum.  (but later after 50 years some atmosphere can leak into this perfect vacuum inside. (lead seal issues)
The top  list on tube issues, are  gas (arcing), effects.  (gas ionizes then invites arcing they work as a team here, all bad)  Do look at the tube, under 2700v power, and when loaded, to see any wild looking colors. (at a distance,  say 3 foot back)
In many cases, gas can not be removed by you, many try, but success is limited.  (sure try)!
It's usually too much gas inside. (hope I'm wrong) but if is only a  tiny gas,it  can be cured, DIY. (luck  can happen.)
If the gas is small and the GETTER is not dead, you can let the getter do its job with filament power only. (run it like this, say for a day, with no plate voltage) (I have and HV fuse, so I pull that , and burn off gas)
If you see a Tube arc  that is gases inside the  tube in most cases.
Just because a tube is old , does not mean they will have gas. (mine don't)
The  tubes do not arc ever (well mostly) in a pure vacuum at 2700vdc.
We have people that have tested this tube up to 8000v and no arcs. ( China tubes are not tested properly in there crude little factory, it's just a crude knockoff of the real thing.  Is this NEWS"? ) {do be ready to be scalped by them $$$}
Do not blame the VHF parasitic oscillations for odd plate or grid currents,  unless you prove NO GAS and  that with a  grid dip meter?, (Do use a scope DSO, FFT mode and see if VHF spurs are present? I bet they are NOT)
If you have  new scope the FFT spectrum mode will show you SPURs if they  are present (50 to 100mhz)  Look for any thing odd over 30MHZ.
Using grid dip meter to find spurs, is going to take genius level skills..  Using even a cheap spectrum analyzer even a cheap HF version you might discover spurs. 100x faster than someone fiddling for days with  grid (fet) dip meter.
You need to look for a spur up to 100mhz,  I bet you don't find them, if  your grids V1/2 are grounded correctly, with short short wires to 200pf caps.   Spurs means spurious signals. (wrong)
See my SDR box (device) below for examining RF signals from your TX:
Tube arcs show huge grid currents on the meter dial or spikes., and don't forget the meter diode protector  pair, mod. to save the meter as it is  no longer made or sold this  meter  200uA.  (now classed unabtainium with proper dial background)
If the tube arcs, the panel meter likes to EXPLODE (inside)"dial set to plate current mode"  There are 2 protectors for this, one in my NEW PS and one set of diodes on  the back of meter (back to back diodes) (both are MOD.s)
If you see odd grid currents (excessive or surging, shut it down and find the cause.


Take John Harpers advice. THANKS JOHN! (any old tubes found parked for years, NOS ,spares, etc , may have gas)
"This is why tubes should ideally not have plate voltage applied until they are warm, since in this way the getter has a chance to deal with gas molecules which have appeared since the tube was last operated."
If your GETTER is still alive (means some are not), the filaments can heat it and it can eat up some gas. (worth trying the moment you see gas or odd grid currents, per above link method)
I guess you figured out the Getter is a gas eater, it gets the gas out.
end gas issues.

MORE MOD.'s to follow below, next up is ALC mode, that most will not do , it takes work. This is the most hard discover, due to Heath no documenting it, nor telling you it's got 1/2 the parts in the PA and 1/2 in the exciter, making it useless for other gear. (mostly)

ALC (Automatic level control)  It is an ICOM analog function,  0 to -4vdc,  0 = max gain, and -4vdc is minimum gain, inside my ICOM. REDUX #1
This topic is very complex,  only because HEATH KIT never published their rules. (a secret)
My SP401 here with SB200 teaches me the rules.

This section is too long to read, so  Jump to the conclusion.  This section  below shows how I  reverse engineered ALC (it's a non standard so needed  that to make it work right)
The ALC is NOT BUNK, it prevents overmodulation (exciter set too high) and YOU WILL do that , as  no operators exist that are error free.
ALC can be tuned, one great PA is the Kenwood TL-922, with the ALC tune pot on there (POT= potentimeter or variable resistor to some the  knob adjuster) The Kenwood has range of 0 to -10vdc, this is grid BIOS feedback.
ALC works dynamically even in SSB mode not just CW.
ALC prevents the PA from going to gross IMD products.
Yes, a modern PA, Fully SS solid state will so self ALC and not do IMD but that only works on modern PA's. (yes modern gear is far far better)

a.k.a: RF compression. (SSB talk power same)
My full analysis, next, (long and detailed)
There are 3 problems here,:
  1. The PA has no ALC adjustments at all, on the Heathkit (the AssUme you only own, HEATH GEAR ! ouch.
  2. The PA has no buffer on the ALC pins, (in case the PA ARCs.(or gas arc)  you do not blow up 20  year newer or newer SS (solid stated) modern transceivers. BOOM you now have $2000 damage. WITH TRIAD portector mine.
  3. The PA has 100k ohm ACL pin source impedance an fails in to any SS modern transceivers, do to ALC bus overload.(#2 is cure same)


 What is it, ALC why its so good (acts an RF compressor) and how to tune it is here?.

ALC is like what a tech learns is school AGC in audio amps and raids and is for RF/IF amp gain stages of any radio.
The some hold true here, it is voltage from 0 to -10v that is  counter BIAS from PA  to the Exciter, this is for the PA to tell the exciter your OUTPUT (my input ) has too much power and is overloading me, so the exciter sees the voltage and backs off on the power out.
This happens dynamically too.  Like when you yelling into the microphone, my TL-922 manual shows how to  calibrated this volage exactly but Heathkit does not, they asssume your gear is green with heathkit labels. Both the PA (most times) has its own ALC and also for the exiter input. (dual ALC)
It is true some modern SS solid state PA;s have  SMART ALC and never lets it self be over driven.  (even stated in the sales borchures)

DO not read all this , this is my analysis of how Heathkit did ALC.
The ALC in this Heathkit is a secret  (no spec), that I  have unraveled. (crossing fingers)
 This means they only tell you it controls bias on the main exciter RF isolator stage,  I does not document levels or signal shapes at all.
Due to the complexity (and fear of ARC and GAS) and all , most folks just leave this pin, NOT USED. (really, I do not blame you)

ALC has great usage to prevent flat topping the outputs (IMD horrid effects), on SSB. (running full gallon mode)
Neither books tell the full story,  they just skip that , because, all your gear is Heathkit and its 1964. (no SS exciters then) right?
Some will tell you Heatkits are junk no that is false if the maker sold it complete and they did only bad kit builder do that.

I read the HW100/and SB401 manual, to learn more,  but this ALC line (again) is not spec'd out, but with a good scope (DSO) we learn more... See the full SB401 doc's here.
  • The SB200 is designed to only run against, tube exciters,That have a 1 megohm source load or more. (hopeless to use on my ICOM 7200.)
  • The SB200 must buck against this R525 and my ICOM D502 ZENER, and  that is not going to happen correctly, I tested the loading factors and it about  .25mA.  (base current of Q461 and R525 too , against the 8v supply) 
  • The SB200 can not drive this below circuit. (prove it via math and  , measured it)
  • The SB200 generates a small negative bias to the old tube exciter or old tube Transceiver. It is only works on high Z, input electronics with tubes. This bias turns the old HW100 off at -5v (6AU6 tube bias rules) (data learned in GREEN)
  • The cure is my buffer seen at the end of this section, on ALC
Skip the below, analysis and research done by me, seems I'm the first.
This schematic excerpt is from the IC7200, ALC section.  As you can see this is nothing like any kind of Heatkit ALC input.(of any kind)
Photo1 : ICOM IC7200, actual ALC section.  The bottom right wire goes to the ALC pin jack. Note the Zener diode D502.
This is what I  do not want to blow up in my ICOM. for sure blowing up transistor Q461 is no fun.
note 2 errors here, no GAS TRIAD protectors and the 1k ohm huge overloading of all old TUBE PA  ALC, out.
 
 The ICOM limits ALC to -4vdc,  many a PA goes to -8vdc.
I tested and analyised most old Heathkit transmitters , SB410 and HW100, both.

 The book quotes do offer help in understanding ALC,  this is only to teach others all things ALC does and how.

The SB401 finals are limited 3 ways,  automatic tube  bias and internal ALC, and the ALC jack (to the SB200)  The jack ALC is predominant to the exciter internal ALC.
This comment in the SB401 manual, clearly shows, why ALC is needed. In fact it internally limits you, but with an outside PA, they work as a team,  .


I am gleening all I can learn about Heathkit and there manuals.

Text photo2: Hw100 excerpted. The same book also tells you how the RF compression works."SSB talk power" The book is free, to find and read .  As you can see the ALC is actually created inside the Transceiver (or exciter). D903 does that.
One has to dig deep in all heath books to find these words.

Lets go deeper into the HW100, called Hot water 100 by old hams.
This ALC signal wins all Gain issues in HW100,  at tube V2 , V2 is the isolation stage seen in photo 4.  (if gain is 0, so is all RF power zero.)
Next is the input ALC, the yellow wire, The input is clamped at D902 to ground (0v) and rectified at D903, to DC bias that  is now filtered at "F" below. (attack and sustain (decay) tuning Filter, C932 is sustain cap)
Last it feeds a 1meg ohm resistor to V2 and V3 and 3meg to ground to make D903 work better at low signals.
First note that the ACL input jack is 4meg ohms to ground. Input inpedance. and clamped at 0.7vd PLUS not negative only minus ALC is used., how much we know know that answer, is  minus -5vdc for the HW100.
If the PA say SB200 /sb202 or SB1000 is used, those PC can put out -5vdc ALC and the HW100 will down all TX output.
.
Photo3: The HEATH Hw100. (PA section)
 C932 is the decay (sustain) delay controller.C931 is the attach delay cap.  << the Secret is here for that...
 I am not saying these old  Filter delays are best with SS exciters, (it is a starting point , yes)
 The Tube V2/V3 turns off at -5v. easy.(at the RF Isolation tube , seen in Photo4 below)<< DATA learned (by testing)
See D903 and the 2 resistors below, these form the second have of the ALC split,  Heath split ALC.



and last, see R22 below , this is a very high resistor 1meg, and does not load down the ALC bias, like my SS (ICom7200) does.
My ICOM  overloads the  ALC line  the ICOM has near 1k  ohm  load that  many not play well with any old  PA.
The V2 tube below is cut off at -4vdc (1/2 mA Ip) at -3v grid its 3mA (Ip) at 0v grid (Ip) is 16mA, max.  The cathode voltage I think is  near +4vdc. self bios, means that -8v ALC will turn off the tube V2, ending excitation.
 The V2 Tube runs self bias via cathode resistor R23, 470ohms  Grid self bias, is then 100% ALC actions here.
So that is what the V2 does, with changing grid bias,, (out of the RCA tube manual at 150 Vp) {Ip - plate current, Vp = plate voltage.}
Photo 4: HW100 isolation amp stage,   below. This is the main, device V2, to shut down , if ALC gets too high (means too negative), even from my SB200.
The ALC becomes direct grid bias to V2 pin1. (and is over 1meg ohm input DC impedance here)


Due to the ICOM overload of this ALC line, the only hope is adding a buffer to the SB200 that will drive any SS amp. 
The ALC circuits are split 50/50, in Heaths design,  and is designed to bias a 6AU6 Tube grid HIGH Z,  not compatible with low Z solid state rigs.

Photo 5:(ALC, SB200) cut up (edited) to see ALC more clearly.

D18  is the key (no pun) part of ALC, this diode, monitors, Plate current, R14 is the meters plate current monitoring pin.  This pulsing signal as seen clearly in the next photo #6 is what you get.
The ALC is then a mix of RF and DC (plate current as voltage) on the ALC pin. Looks like this.
C15 puts the grid at RF ground, The V1 grid uses Fixed bias. When you key up, the Grid bias goes to 0vdc.
If a tube arcs (gas) the tube may send huge high voltage, spikes to things, you do not want. My Triad protector (TVS/GDT,Zener)should keep my SS gear safe.

Note the PA internal ALC bias is sent to ALC out. D18 converts the modulation signal to neg. bias.
As you can see the ALC out  is near 10k ohm source impedance and will not work with the ICOM (mine) low input impedance, unless I use  buffer.
The HW100 and SB401 are 1meg ohm inputs ALC.(aprox) and like the SB200 as we expect.
The unloaded signal at ALC out looks like this below in photo 6, in SSB mode, yelling in the mic, at full RF power.
 This is a good signal, but only happens if  ALC cable, is not connected to the IC7200 , if you do connect to the ICOM , the signal is swamped. (GONE) (1k ohm load is too much)
So, the only cure I can muster, is a buffer, or unity gain buffer.  (and a pot from the output of the buffer to the jack so I can run alien Transceivers and tune ALC where I want it using the full measure of -4v , if I want)

Sadly there are no analog low voltage supplies inside the PA (15vdc would be nice) the SB200, to allow this op-amp to work easy, but see my cure below. (now done and working.)
A bigger  PA. like the SB220, with 2 3x500z tubes.Will have higher ALC outputs, above -4vdc seen by me. (more on this later using my Kenwood)
My Kenwood TL922 full gallon amp is rated at -8vdc out max spec. (manual) (pot adjusted on rear)
Example mine:TL 922 (shows how Kenwood does this)  The ALC source inpedance on the KEN is 10x worse more.
14mhz 80w exitation, -10v (-8v listed in spec) -10 is for calibration settings, if thi test fails for too high negative fist the PA ALC feed circuit.
Set exiter or power max (120w is that)
Adjust POT on RR of PA for 500watt on PA.  (922 spec,max.: 2 KW P.E.P.-SSB and 1 KW CW/ RTIY)


I tested this ALC OUT  jack cable , not connected at PA , and then using dummy load resistors (tiny carbon 1cent resistors) to see the SB200 ALC  strength, answer is that it is very weak drive. (useless driving low Z, solid state transistors)
Disconnected really is an   illegal mode but is valid as  base test.,  the heathkit exciters all do not load down ALC at all, with 1meg ohm impedance on ALC jacks.
The below is MAX ! ALC not connected, once connected it's near dead, I ran experimental loads on this line and all failed. 
I only did this, lacking all schemations for all my radios. (have them and now see why)
Tests loads on ALC to ground..  (using test load resistors)
15k ohms , -1v max. (overloaded)
47k ohms,  -2v max, (still overloaded)     (my effective source  impedance must be in the order of 50k ohms and not any good for bipolar transistor drive. no good for solid stated radios.)
This ALC signal is gutless, and runs the HW100 ok with its 1meg ohm + series grids lines. no problems.
As you can see SB200 is  designed to run a 1meg ohm load max. or the signal is totally useless for solid state gear.
Again we learn , Tubes don't run SS. (except on the RF jacks as that is just RF , not controls) Back in 1965 (about) there were not many standards then, so all this is HEATHKit state of the  ART !
To answer one guys question, why -4v, easy (the minus?), the vacuum tubes shut off there.  (grid bias is negative) It's so common even ICOMS newest gear use these levels... (it only needs to be buffered and filtered)
-4v is kinda a quasi standard, I see ICOM uses this and so does Ameritron but the latter uses POT and has as slow as 1000source impedance.(nice)
Heath should have added one more tube to create (if crystal balls existed then?) and ALC out buffer, a cathode follower would work perfect here.


Photo 6:  ALC , no load, no cable, at jack.  (this is infinity load  , cable pulled)  Try as I might, this is all I can get. on SSB on CW I got,  -4.5v , as seen here clear as day... this is max.

That noise there is RF + modulation (7.2mhz)  That noise is gone by the time it reaches the HW100s ALC Filter and rectifier if was connected up..
Note my very short transmit time.a 1/4second. (this 500w is  Rel. power on  AT2kd meter watts, keyed) (I'm tuned to max at 40m) in to 1kw dummy load (gallon)
The real ALC dynamicaly seen !

My theory for a cure is a BUFFER (non inverting).   (or an  Emitter follower transistor using rare Germanium PNP for low Vbe drop, this can work in is just 3 parts for $1 the error with this part is 200mV, better is below.)
The other cure is  $3000 to $6000 new SS PA?
The buffer has  high Z input (means high impedance) and  a low output Z, (741 op-amp is output z at 75ohms) This is the perfect way to drive your SS transceiver and super cheap too.
The buffer will send in the analog voltage (and less noise) with any load up to 25mA.
The key features are , no loading effects on the SB200 and a prefect reproduction of the original signal and no loading errors.
If you have 12vdc power inside your modified amp now, as many do this is even more easy to do. with a DC/DC converter, +12 to -12.
A unity gain , negative supply buffer is hard to build lacking 2 power supplies, say (+8v and one -8v or more for the power pins)  
The 1968 to 2016,  uA741 OP-Amp works perfectly, (ti.com  still makes it)  The CA3140 is the Rolls Royce of cheap Op-amp  that is even better....)
Sadly , the  SB200 filament  supply , although has  robust (current) is only 3.15volts CT,  that will make only  4.5v peak DC but the diode rectifiers drop lowers the further 0.3 to .7v. all bad news. (we need minus -10vdc out for sure)
We need at least 5v + 1.5v  or 6.5v supplies, (for a regulator. we need what,  2v more volts (at least for regulator happiness) from 6.5v,  now we need  8.5v total , a 7809 and a 7909 fits this bill.  ( options are +- 15vdc wall warts but 12vac works  best)
or even 7812 /7912 will work as will 7805 and 2912 paired.
The two most common low voltage transformers are 12v and 24v.  (in the old days 6.3v was tops and 2 times that...)
The solution is 9v to 15v dual supplies, Plus and minus.
Goal to steal some juice for my new OPamp, and some juice , that will not go to 2700v, during a gas event. So I picked the BIAS base supply, 120vac RMS.
To do that I used the lowest current transformer I could find so that I did not overload this very weak gutless winding . The HAMMOND transformer here is perfect. Plus no need for a fuse for the HAMMOND. The main transformer is my fuse.


My cure:  MY HEATKIT ALC , buffer board.   
Built on a vector perf board. easy.
I used the ubiquitous uA741p for simple reasons, it will always be made.
I do not know of  a single rail only negative  op amp, (single rail + yes, minus only no , and this great set of examples is good.)
My Kenwood PA has ACL source impedances of 70k ohm and  7 times  worse than the Heathkit 10k ohms. 
My Kenwood Version will need a  7915 spot regulator. (just in case I need more range)  the Kenwood book only hints at -8v out  typicial (I will prove this and report it)
The Ameritron, ALS1300 (SS MOSFET PA) @$2500. needs no such card below it uses 1000ohm source impedance and just works. (with anything, I presume)

Qty: ID and description:   (all from Mouser)
1- R1 50k pot, linear taper.  All parts sold at Mouser.com
1- R2, is a 50k resistor, 1/8th watt makes a nice fuse. (lower the watts the better it is) this is my arc/gas/surge protector for U1.
1- T1, I bought a Hammond transformer, 120vac in , and 12vac out, and will connected it to the 120v secondary (bias winding) of the monster 500va, to this 1va tiny transformer and makes 15 to 17vdc with the next 2 diodes and then to the 2 regulators. (power loss will be nill  this way)
The transformer shows how to connect the pins for 120vac in and 12vac out.  (it's dual input and output configurable) This is a real nice flea power, transformer that will not overload my factory bias windings on MAIN transformer.(or allow GAS/ARC fails to land here)
The next 3 devices are what I call my TRIAD protection.
1-TVS1  , The part I use  is at Mouser, and 15v break down is my choice "COMCHIP"  1.5KE15A-G , @ 63cents
1- GDT, I used the lowest voltage and most powerful sold in axial leads,   75v arc,  Bourns serices 2049.
1- ZD1,  . its JEDEC part too..1N5929 (3W !)  this zener is super fast, each protective device has advantages over the next, speed/power.
2- D1/D2/D4 ,  1N4004, found in everyones scrap box.   one cathode band  to the right, one to the left for dual supples, plus and minus.
1- D3 A low V forward drop "Schottky" diode with its magic Low Vf  (low volt forward bias drop of better than 0.2vdc) I'm using "ON" semiconductors,  MBR160G  device, sold at Mouser.com
2- IC linear voltage regs  UA79/7809  50 cents each, all at mouser,  The regulators mean the ALC will not change with line voltage changes,  lets say the homes Air conditioniong plant (HVAC) kicks on , or ? what ever.

2-  Caps, 220u at 35vdc,   electrolytic, do not install the minus supply cap backwards.  $ 88 cents
1- US741 op amp.  50 cents. (TI) 8 pins and  socket, sure. ( CA3130 or better the  CA3140  works too and is MOSFET inputs , with TERA ohm input resistance and far better built in,  input protections, see all that here)
No heat sinks needed.
Built on a simple perf board, sold  for penny's sold on fleabay and all way in my parts bin,
We now have just  4 wire leads, ground, 120vac, ALC in , ALC out.  It will drive, up to 20mA external loads,  (huge in terms of ALC)
This will drive any transceiver that accepts, 0 to -4vdc ALC bias.  most do. 
The T1 transformer blue wire input is 121vac, not keyed and 117vac keyed so there is no major sag on that line, to mess up my grand idea. 
Click the drawing to ZOOM it.
I have read the schematics of all old heath transceivers to make sure the detect is the same in all of them, they are for same era.
So I now added the mimic that matches what was present in 1964 - 1978 (tube years)
Keep in mind and external 12vac wall wort and a jack added to the amp, and you don't need, T1 .
I wish someone made  negative only power supply Op-Amp but seems not....
If our input signal was larger say -5v or more, I could use a classic emitter follower here,  a 1 transistor and 1 pot circuit.
Read the Ameritron pages on how to set, R1 pot.  They document this well, and is in all their PA manuals.
After lots of head scratching I took a BUD box, and added 2x BNC connectors and my pot, so I don't need to modify my metal chassis.
Schematic #1 (mine)
NO copyrights, this is free to any ham, or you sell them. I DON'T CARE ! The Transformer there means you don't have to find some free power sources in the PA. It now has its own. (and gas/arc overs can use power wire as  killer gateway to my board)
The pot seen here works just like the Kenwood TL-922 PA, rear pot. (not by accident, in fact I studied all PA  made, that have full documation or just a schematic) Click to zoom photo schematic below.
This circuit below is  current buffer, and voltage follower. Unity voltage gain and huge current gain.{for the SS receiver} (if your Receiver already has attack and delay built in ? delete C4/C5.  Most of the extra parts below are for GAS/ARC protection.

If your R1 pot above is over critcal try a 10k pot, pots are cheap. (some full gallon 1200w PEP PA's may need up to -10vdc ALC controls and just change U2, to 7915 to gain that ability.
 
One can make a case for moving all the protection on the right to the left and trust the 741 and the max 12v possible there.  A good case, but I like mine this why above. (all  parts to the left of C22 are sacrificial  say during a GAS event,)
Adjust pot  R1 above, so the  PA (SB200 out) is just below the non linear CURVE of power. In SSB mode it will then do RF compression. (nicely)
The transformer above is connected to the blue main PA 120v-rms secondary and does not overload it, nor weakening bias to the grids of the 572b tubes. (tested by me)
An alternate to the T1 above is a jack and line to a external wall wart transformer, 120AC in and 12vac out.  Some folks do it this way, so they can uses this Wall wart power for other mods.  (CW mod's a plenty)
I picked  a weak T1 so that I did not overload my 120v main secondary winding and then have tube bias drop (a bad deal there).
With a wall wart at 12vac  in hand (not DC) one  could build and outside BUD box (LMB) with 2 BNC jacks and the wall wart JACK.  (thinking out of the box here , HI-HI) 
Connect this AC wall wart to the 2 diodes left side, above and no need for T1. In essence T1 becomes the wallwart.
The grounds must all be super good using that outside box. or all bets off. (that is why my  board is where I put it, inside)
Ok, it's now done.
Now some supporting data, bench testing... The below is my tests done in order to get the job done, I head to reverse engineer the SB200 and SB401 (I have both now) It turns out ALC is 1/2 in one and 1/2 in the other, proving that Heathkit and no plans to support other makers exciters.
More progress , 1st I will show 2 photos of the SB200 both unloaded as stated. To see what is going on.
The tests are all RTTY mode, keyed, 7.060 mhz,  500watt (all watts are tuner AT2kd watts )
All test are with out exciter RF-IN  connected, the ALC line is open, we already know the SB200 hates modern SS loads on this ALC line. (SS means solid state, not hollow state (HiHi)
First is what  I call the naked signal.
This is with only a factory .02uf cap C22, there on that line jack, and this is what you get. (ALC jack unplugged)


Photo 7:  (what I call naked ALC,  no  (HW100 or SB401 ALC line connected)
The below (reality checks) shows doing tests to prove why  OLD PA will not work with new Transceivers (ALC) (the PA source impedance is totally wrong)
 This is pure raw SB200 ALC out, totally zero load on this ALC line.  (it is mostly meaning less, due to the end to end distributed nature of the circuit)
(all are photos, 2000v scope probe at 100:1 ratio, for safe testing)  The scope knows it's 100:1 and shows the correct data.
 You can clearly see the RF component of the signal. (it's a mix of DC plate current and signal)   I think this live signal , can turn down V1 /V2 bias , reduce plate current, for self ALC actions.

I tried dummy loads on the PA, ALC line , with above wave form and found any loads at all overloads the SB200 ALC lines, even 50k, ohms is too much load !  (this was a key discovery for me)
The normal load, (discounting caps charging in attack/decay filters) is about 1meg ohm..
Next is , I add the missing diode (mimic), to the line, that only rectifies  my signal ) This is just to demonstrate the above signal is not really important, what matters is with the Exciter does with this above signal.
This is the anode side, (negative rectified and a .02uf to ground added )  This is just to see and show that the 2nd halve of the electronics is not present in the SB200, but is all Heathkit exciters, (H100, SB1xx or SB401 )
Photo #8:  This is  crude load  next, just a diode and a cap. (a crude mimic of my SB401 exciter)  In the end, I mimic it all in my vector board , next up.

See that?  its now a -4vdc level, like we want, but will never ever drive any (Low Z input) SS exciter, because the SS overloads this line.
 This is at full SB200 power out. Max.  ( 624.8watts(avg) or 1240 peak watts)
My buffer board will be and was the cure,  in my case for the (Icom 7200)
A SB220 may do , more minus ALC, due to the more powerful 3-500z Tube pairs. (I will test this on my Kenwood, TL922. later)
 
One more thing I learned the signal above is slow to go to -4v,  a 5 full seconds. ( the reason is easy, the heathkit excisters are 1meg ohm ALC inputs or higher , high Z inputs)
Last, but not least its done. , and works correctly , but I tested my new ALC board, into a 330 ohm ALC line test dummy load resistor, to be sure output is still -3v at  PA exicted at 70watts in.  Passes the test.
The 731 op-amp OUTPUT sourced impedance is about  50ohms, more that strong enough to drive any SS exciter ALC input circuits. (SS means Solid STATE rigs, transceivers in TX mode)
Photo: 9 installing my new card. I like working with these hand wired PCB,  for years.



Next, are live transmission tests and ALC loop tests, after I mount this CARD ,  in the PS cage. (my PS cage has 1/2 size, modern caps now, so lots of space here ,now)
And the POT outside, in a box..
Lots of fun here getting the SB200 to do proper, ALC actions into my SS  exciters.  (ICOM)
I now have the regulation working, the whole loop, RF to DC ALC loop.
My  Icom rolls back power, as I turn my Pot CW. (clockwise)
I'm using the Ameritron's (technique)  instructions verbatim from there PDF docs. ( my Kenwood TL922 has pot for ALC but is useless for SS)
The word POT. means  this in electronics, Potentiometer or "a variable resistor with a knob" (ALC knob , ALC DIAL ,etc)
Last is my new ALC working:  (I've no idea at all on the proper attack (but did learn it later) and decay, but you can adjust C4/5 and R3/4 to attain what ever you wish , as seen in Schematic #1 above.
Photo 10:
This below section  is all parts working together, my Icom 7200 and SB200 working as one. with the new buffer board, and I can set the ALC level with my Pot. to what ever limits I desire.
The rumors read (endless searching on topic) are that attack and delay are 10:1 ratio.
Below as we can see it's about 40 mS attack and  100 to 150mS decay,  I have seen radio's do 1 or 2 seconds here on decay, not sure if that is good or bad. (so far, so good below)  I didn't have .02uF caps in stock, so used 0.01 in the stead, seems ok.
Photo below is the output of my new ALC buffer car, in a live system.  (all thing connected and transmitting full power to my dummy load)

Last is tuning ALC, I use the rules on Ameritron RF AMP.'s to do that.  (I won, I got solid state to like hollow state ! ) 
The AMeritron ALS1300 drives the ACL pin with  forward power live monitor and a single OP-amp  U2 MB and one diode and 1k resistor. so is  very low impedance driver. The stated 0 to -4 vdc is out put range of ALC and set it just to glow the ALC led at full power via ALC pot.
A fun project ,and 10x that knowing others all say it can't work.  Yes, it can...! and is BEST of all.
But in conclusion, add these words to the SB200 manual " this ALC system (not covered in the theory pages, sadly) is not designed for nor expected to work with other exciters , nor for sure with any SoldState gear."  YMMV
Here it is fully installed now.

 
Building , testing and debugging with out a scope takes a genius, not me. To build, design, align, repair or test your gear, or debug new mod.'s , you do need a scope.
What I'm doing here, is using  Voltage Follower,  as seen in the ARRL big book, and then I recreated (copied) the Heathkit, ALC detector and Filter, as seen in my SB401
I have not tested this with my SB104 yet,  but the 104 has a wrong 10,000 ohm input impedance...and will not like my SB200. Heath violated there own rules when the switched to SS Transciever.

ALC short play story (condensed to the max), a fast easy list of cures.  ALC is like AGC for  a radio receiver, but  in this case,  the total gain of the Transmitter is limited by ALC.
  • ALC is non standard protocol, if you mix gear.  Each maker use their own secret (some not) methods for ALC.  For sure Heath did. (the Heath secret is, it works only on their 1meg inputs to tube Grids and not the SB104 SS Xcvr.)
  • That includes secrets for , attack time , decay time, voltage levels (and polarity),  and input and output impedances of the ALC line jacks. (I find 0 to -4v ALC bias works best, as ICOM states) It may be true that -4v is a quasi-standard now..
  • To make this work reliably , I added a buffer to the SB200 ALC line, this buffer board, protects the Solid state  ALC input to my ICOM Transceiver.
  • This same buffer card, takes the SB200 and scales it correctly and clamps all positive excursions to +0.2v max.  The buffer gain is UNITY or 1:1. before the POT.
  • I have at least 3 levels of protection to my ICOM, so that if my PA tubes arc over or go gassy, the surge does not hit my ICOM.
  • In the spirit of "Ameritron" Amps, the level POT. sets the ALC threshold. (and my great kenwood TL922 ! PA.)
  • I have it all working now, and can see it kick in, if you run too much mike gain in SSB. The ICOM even lets me see ALC activity live. (LCD front display looking)
  • All linear amps, (not home grown/ rolled) use at least an internal ALC, then if you add the ALC cable, you now form a loop (servo loop) this loops lets the exicter, now protect your linear,  It reduces exitation  (-4vdc does).   My hotter, kenwood goes even more negative -10v is max)
  • The loop too, allows some loop compression, due to the rules of attack and delay.
  • One can make small  box, (BUD) with the ALC buffer inside, and  12VAC (or 15v wall wart, transformer as an input, and 2 jacks, ALC in and ALC out.)  sure can.


end ALC.   (Heath ALC)




SDR:    Software Defined Radio's  (rule one, do not transmit in to the SDR,  by any means, direct or indirect, or near field inductions, or BOOM you let the smoke out, sorry!)
See my SDRplay used as PANadaptor.
The last test I will do is with my RTL-SDR (using SDR box as a spectrum analyzer like this HAM Lief Asbrink does, to be sure TX band width correct and  is not splattering (barn door effect) or IMD products are manifested , spurs, etc.
The chips designed for DTV digital TV, and used mostly in Europe , a TV Dongle  is common usage and mostly just junk, today. IMO.
I like my SDRplay  version 1A best, it has wider range of frequences it  can receive. I can use it as a poor mans RF Spectrum Analyser. (not $10000 but only $99)

OsmoSDR: (this will confuse 9  of 10 folks but is a good place start just to see how complex support software is...)

Larry D. Wong:

Here is on Windows version.
TIPS:
Make sure you PC is in top health not infected or a bad HDD, run test on it, first.
 Avoid Windows XP and Vista, support has ended there. XP is a virus magnet never ever use XP today, its a new world now, and support ended 4 + year ago on XP.
Run W7 on up.
Turn off the AntiVirus protecter before attempting to install any Panadaptor software.
Using MS defender?, click its ICON in the task bar,click the left icon there, threat protection, click virus and treat protection setting , then turn off real time protect, and install your software.
Make sure the USB drivers on you PC work ok , test them with a  mouse,  and then make sure after loading the SDRplay that its drivers loaded ok in device manager. (clicked in Control panel yes)
I run mine on W10-64b on an 8 year old laptop HP 8540w. no problems ever.

Examples of signals,  youtube has endless demo's so watch them.

Do so using large db attentuators in the NEAR FIELD RF (start at 40db and work down using PADs) in to the SDR input,so as not to overload the poor thing.... as seen here.
SDRPlay and attenutator and a small protable antenna .
Do no RAM 100watts in to the SDR or it will blow to hell. (no if's and's or but's boom you let the smoke out , grin !)
I use my dummy load first, on the PA, and I can look in the shack at my leaked local signals using any SDR.

This software can allow you to monitor your own transmissions, even with a tiny antenna, as  side monitor of your RF envelope.   (set the gain real low, so as not to overload the SDR.)





 


The SDRplay , review, of the many, just random pick here by me.   The Device is worth $99,  for sure the RSP1a , the older one has no shielding and is no good. (unless you fix that ,Like I did mine)
The RSP2 are mostly overkill and expensive.  I dont need it for HF so can't advise.

https://www.youtube.com/watch?v=RfBuLJgjAng&feature=youtu.be


Last is Cantenna.
Using pure refined RAPE seed oil, has the highest smoke point, and was never exposed to dead animals or anything, sorry no rancid deal here.. (Walmart brand $5 /gal)(aka. Canola, what perfect name for Cantenna oil), shocking, no?
It works like a charm, and will never short  out with our super low voltages here,  250v p-p is not , repeat is not high voltage. (at your power distribution station it's 60,000 volts or lots more even near 1million volts, and they now use , drum roll,,, veg. oil "the green planet stuff" )
The resistivity is over kill, leaving what? (conduction and life span)
The oil has great thermal-conductivity if you study that. (so works)
Last is life span, I know others running it for 20 years, and no problems. (in doors, only, outside, I've not a clue)
The spec, is clear. Read it?
SMOKE POINT >450░F
Flash Point and >615░F
Auto-Ignition Temperature   >650░F (would take something crazy happening  to do that)  Some transformer oils spec here is 518F.(nytro11)
Transformer oil wins the cold operation spec, at -57c, that I just don't car about, inside my home.
When you cook with Canola, do you walk away?, no , when you transmit do you walk way?, no, so if it smokes, stop transmitting, I abused mine in a test and it's OK. no issues, at all.
I also sealed up all the Cantenna connectors with RTV, so it can't weep.  and did not fill above the bail, bosses on the  side, yes, all oil expands hot. (as do most things) and the MFJ can has a special vent.
Later ,I added a BNC test jack so my scope can do real power out measurements.  (per K7ZOI Wes Haywards guide lines)

Mineral oil is dino oil, or crude oil, or rock oil, Rock oil  is a day1 name for crude oil and is a mineral mined from the earth (drill mined) But we all know rock oil is really mostly millions of years old,  dead  plankton and such. (hard proof of that!)
So, really both mineral and veg oil are the same, one is fresh and new and the  other is very very old dino veg, oil., But either can be refined, as much as we want.  (Refined , dried out, and keep the flash point high)
(in  fact, can be hyrdocracked to any thing they want) ( adding or removing hydrogen on demand) long chains to short and the reverse.)
They can take natural gas and turn it to oil.  (a lot of so called synthetic oil starts as Natural Gas).
I transmitted  a long long time  into my Cantenna, at 500watts (AT2kd watts) and it only got luke warm. (using all modes, SSB/RTTY and CW) checking for IMD, etc....
It's a winner, for  sure.
If it were to leak (it's not toxic at all , after all its cooking oil) then it gets captured, in my  plastic bucket (I want a red bucket for Christmas)
 
Dot dot dot dot, dot dot, repeat, = hihi

I used my Cantenna to make sure my bandwidth is perfect with ALC active, and at any time. (no spurs, no crazy VHF parasitics. , no IMD issues) And checked with my SDR.
One more joke, rape seeds are CANola oil, so use CANola  in your Cantenna, and do the CAN-CAN, dance?


End ALC, checks, 


Power out tests:   The big moment ! My AMP is now ready to ROCK!
This section below is under testing and analysis and tuning. (mostly done)
After getting in all mod's   this is my real power out (max) using a SCOPE to discover this, VSWR is near zero reflected. 1:1 ,  the SWR needle is hardly budged off stops.(AT2kd cheap meters)

First the final results, this is the scope reading out of SB200 in to my Antenna tuner , turned perfectly with low VSWR. (end to end perfect) WYSIWYG.
The below is at max  peak output, any more rise in ICOM  exciter power , gives no more output from SB200.  at 70% power on ICOM.
Tek 10: 1 probe, at RF out,  scope set to 10:1 so 100v is 100v per below.  7.28mhz 40m.  CW keyed up. NO ALC  This pure  full absolute power. To learn the other power equations read this first.
First off this is not an accurate 50ohms , I will use a 1.5kv dummy load later in this section, this is into the turner.(matched)
500vp-p is 250v-peak, vp
250V PEAK  (Power= E▓ ¸ R)  That is 1250w (peak) E is vp  , P=  250 squared ¸ R   (peak is 1/2 of peak to peak)
Vrms = peak times 0.707 or  176.75 vrms   (RMS power is only for heating dinner, in a microwave oven, Pav is talk power and so is peak (for DX usage)  In SSB mode,  your yelling loud makes max peak.( and when silent no carrier at all)
What matters here is Pav, Power average (as my cheap line meters show)
Pav = (Vrms)2 ¸ R   (R= 50) aka. P.E.P power, peak envelope power.
Pav = 624.8watts(avg) Not bad for 50 year  old boat anchors...
This ham covers watts pretty good.(by Adam Farson, VA7OJ/AB4OJ)
In my case, my scope has super nice low capacitance loading probes,  10pf and I can read up to 2000v peak.Safely and directly, no need for $150 monitor adaptors, but I did make one.
What you must not do is use a 1:1 probe here. Many scope are limited to 100v, on all inputs direct. (Mine are all 400v to 600v, but many cheaper scopes are super limited on direct inputs, 1:1)
To be safe use a 100: 1 probe.

My tuner presents the perfect 50 ohm load.  (from SB200 , AT2kd and then lowloss feed to my G5RV)
My VSWR is 1:1.5 or less, its so low it's hard to even  measure, (nice tuner I have! smooth action and dead reliable. AT2KD By: Palstar this tuner shows 500watts)
My 572b are old real USA tubes. "United Electric Co.) 1964 (same year as radio) stamped as  made "work week 35"   Hooo Whooooooo!   
A safer probe  is a 100:1 probe,  rated for 2000v.
The probe below is on  a direct Tee fitting PL259, at SB200s  TX output cable to tuner.  (the tuner makes perfect match so that you see is accurate)
It is stated in my book, by the Wes Hayward, MSEE,  that this is the  wrong connection method,. using my Tee fitting. 
He states that  I must use a closed fitted jack on my dummy load (I have it now) to get accurate RF readings. Seen in a great book in chapter 7 here page 9.
This is 100mhz band width scope and you can see it's clean RF (per below),  the FFT scope mode , shows its clean too, as does my SDR gig, (no clipping here full hog on , no spurs)
This  scope FFT is not very good, but is my first check.  See above ALC tests for better Spectrum tools , used and my SDR. (The below is signal over time. Not signal over freg. scans, this is the total combined, sinewave)
The first reading below is old, its with low line voltage 108vac rms. and is now corrected, see scope view 2:
Scope View #1  (raw RF power out, voltages into a Cantenna Dumm load)
The the scope probe is used 10:1 Tek probe, low cap.  Rated over 600v,  ( I also have a 2kv probe if need, "Hi-Hi" that be illegal power, sure)
This is before my huge antenna tuner. Tuned to nth degree,  AMP and Tuner, to load.  SB200 PA out, voltage Peak to Peak. (as all scopes do,)

RP power out. below revised is with   123vac, rms, 60Hz line power. (30amp real AC line power)  My HV is now near 2700vdc, (way more now)
The after tuner scope readings are 240v peak scope probe is now directly on the DUMMY load. 7.2mhz, using Wes Haywards connection rule, I now have a test point added on my dummy load, just  for this test.(and future tests)
SWR is near 1.05 to one, the needle just  moves the most tiny bit.
The new power feed is AWG 10 wire, so as  side benefit, my gear is now all grounded with that 10awg,  to my service entrance, 200amp, power center and the even more large ground wire to the huge zinc dipped ground rod and all SST clamps .

  
Over 70 input the amp pumps, but  not now, with my new ALC working !!!,   my ICOM 7200, CW mode. 7.2MHZ, and with new line voltage 123vac, brand new 30amp service just inches from AMP.
All readings in to 50ohm dummy load, 1 kw rated.  All watts are from scope calcs.   Static plate voltage was 2700v , and sags to 2700vdc @TX and the peak voltage in SSB , is 220v, limited by ALC.(whistle testing in my new electret mic)
Icom %Power out
into RF amp.Excitation
Grid -I
max.
Plate-I
 V-pk
 W-pk
Watts RMS
Pwr avg Watt
Or P.E.P
70
340mA
350mA
240v
1152w
815w
576w
60
250
340
240
1152
815
576w
50
200
310
200
800
566
400w
40
140
270
160
512
362
256w
30
100
220
150
450
318
225w
20
70
175
120v
288w
204w
144w
 So 60 watts  input from the exciter is plenty,  My new ALC design, makes my ICOM love my RF PA. they are now married.  (My  Pulstar AT2KD,Antenna tuner power meter matches same data, as the far right column)
The tube spec. for grid is peak current of 150ma, (in a 2 tone test) so would be 300 in a pair of tubes.
() in each case of grid current above, is max Ig found,  if I adjust the load knob, so watts stays highest  and cause Ig to drop,  I can run way lower Ig currents and keep 500w full power, this the best way.,  See tuning here.
Do not overdrive your AMP or risk getting into the non-linear region, and IMD as door prize .  It is said by TOM, to always use max power, to keep IMD lowest. 

  Conclusions: (a short list)


To avoid all the above, do this?
Boat anchors: ( I say anchor, in a  loving way !)  Heathkit's,  love them all.
Best is to just buy this AMP:

LINEAR TRANSISTOR AMPS:

I really should  buy a ALS600.  but I sucked into the vortex of , Heathkit's(tm) that I have fond memories of,  built many back in the day.. 1960s. for other hams. (I sold them in a HAM store then)

They are vastly more easy to get, and find, used, and USE state of the art, Solid State  HV FET drivers.
Can be had $600 used...  half orig. MSRP.  They raised the new price by 400 bucks, dang....
The parts for ALS-600 are off the self. At the makers website, 100% full support. (DIY or they fix it) But no lie, this SB200 does work very well.

Jargon,  lots and lots of Jargon, for HAM and Electronics tech. and EE.'s  RF EE's
MOD. = modifications
Clamping devices, this is not some radiator hose clamp or Tye-wrap, it is and electronic clamp, the most simple form is just a diode. (In my olden days  a tube clamp in the form of a DIODE) (one running joke then, is the COMMODE DIODE, used to filter out all the crap)
A  SIL. diode can clamp a line to 0.7vdc (silicon) or 2 in series to 1.4v or a zener, diode to many different voltages, see 1N59xx parts. so will GDT and TVS.
The TRIAD (my work)
If super paranoid we use 3 clamping devices taking advantage of each devices prime advantage, speed or current or surge capacity. (this is very common for modern equipment for halting indirect lightening surges or killer ESD , electro static discharges.)
In our case we must clamp, spikes caused by bad TUBES (say gassy) to SS exciter getting  damaged. I use a TRIAD CLAMP set.
SS = Solid state , also known as TRANSISTORIZED, (or in the half baked world, (history) a hybrid (mix)
HV = High voltage (and DANGER) or the word Lethal.... (no second chances)
VTVM = Vacuum tube volt meter, long gone in a modern world, is not DMM ,Digital multimeter and if savvy (safe) its rated over 2kv , for sure if the AMP has bad ground, do not blow up your meter.
kV = kilo volts , so 1kV is 1000 volts  2.4kV that when touched by you, are instantly teleportation to St. Peter. (no oops, no 72 virgins, hihi, no second changes, sorry.)
kW, same "k" rule but kilo watts. 
One old joke, is the wall sign "Danger High Resistance"
mA. mili amps,  1mA = .001 amps.  (or 1/1000th of an ampere.)
PA= not public address, it's old ham  jargon for, Power Amp, or in the context  of  R.F Power Amp. (linear in this case)   Also called "Finals"
Full gallon,  1000watts (as in a gallon of dummy load transformer oil,  be that mineral oil, or real transformer oil ,sans PCBs)
PCB, = printed circuit board, not the old "hand wired chassis" of Zenith TV fame and funny commercials then.
PCB is also the nasty toxic substance used in old transformer cooling oil called.  POLYCHLORINATED BIPHENYL. < bad news.
Barefoot, no large PA active, no Full gallon,  ( I think it means you can walk about in the basement, bare foot and not get RF toe burns doing so.  (as seen by me, at 10 years old in 1958, In my uncles huge powerful ham shack (2kw+), he was a Collins RF engineer )
CRT= Cathode RAY tubes,  this is technology (not the inventions pre 1900) used for TVs, but used by real citizens started in 1929 on the East coast and for sure not west.
For me January 9, 1947 - KTLA (one of the first in the western US)  Most homes in our town did not have  TV until about 1950's , the CRT died the day LCD screens came in to view. LCD = Liquid Crystal Displays.(now OLED)
RF= Radio Frequencies.http://www.oldradio.com/current/bc_tv.htm
FET, a field effect transistor that works almost  like a very tiny TUBE (tube means a Fleming valve ,electronic)  but the Bipolar transistor is a current mode device (Bell Labs fame), with gain.
MY TRAID SET: (or DUET)
GDT, gas discharge tube. This and the next device, prevents cascade  damage from ARCING or GAS , tubes. (I used 3 parts in a  TRIAD this is #1)  This device is the FIST or Hammer , it's slow  ,a bit late but STRONG. (all the parts here are very inexpensive and all sold on mouser.com)

TVS  , a Transorb« device.
  5KP100A   It acts like a pair of Zeners  in series. (Triad 2) This device is super fast and medium strong, This is a rapid advancing technology. (Not tested by me,  chicken I am) It is only insurance.
The below is not used on my PA RF input jack, only on the ALC and Key line Jacks.
Zener diode, 15vdc (-10v max working Volts on the ALC jack 1N5929  anode to jack center pin) and key line  only 5vdc max 1N5918  key line is this Zener.
Gas,  not flatulences, but is what the vacuum inside the TUBE hates,  and is sometimes,  from the air you are breathing , as somehow got into that TUBE, or by other means.  GAS can start  ARCs.  Gas ionizes, just like a neon light does, (LAS VEGAS style) and always bad news.
Tubes overheated (from faults) can self gas up inside, (element destruction , like making the grid or plate glow red. or orange)
ARC, means a 2 points+ in the tube have flashed over,  or created a plasma?,  a new tubes USA made, can run near 8000 volts and not do that. so... it's got 50 year old age issues or the  tube is from China, (clones,etc) even parts inside the tube bent/cracked or loose can cause this.
Arcing is not plasma, plasma is gas  ionization and conduction. A pure vacuum tube will not arc. (not at these levels ever)
"No material has a permittivity lower than that of a vacuum". If you add a gas, it will fail, the gas will ionize and conduct illegal currents. Gas is 100% totally unacceptable in this tube.  Like peeing in the swimming pool,  fail.
 Keep in mind I had transmitters that had plate parts that run in pure  water, that did not conduct at all at 25kv (read magic filters). So vacuum is a pussycat for this tiny 2700 volts.
I love lab tests , and this one takes the cake,  they had 4000v stand off's run at 1 um. (1 micro meter) about
.00004 inches,  gap., keep in mind our tube is at 2700vdc on the plate, and NO RF applied and some static  90ma current flows. The vacuum arc gaps are HUGE, in the tube.(insane huge)
My point is that even small gaps as tiny as 1 um can withstand 4000 volts in a hard vacuum. The perfect HV medium that it is.
Once the electron cloud flows, and hits gas, things get far worse. In a pure vacuum the 2700vdc is going nowhere. The tubes specs have nothing to do at all on standoff voltage, only when under load Vp max,. to avoid excess plate power dissipation.
Look up dielectric spec of a pure vacuum.
For FUN, Here is what ARC-ing looks like (a 1930 coil Ford Model "A") this is what we don't want.(anywhere and at anytime or place in our systems)

IMD, = Intermodulation Distortion
Super slang is this: SWEEP.  (most persons today have not clue what a CRT is, nor what SWEEP means, at all, why should they?)
Sweep tubes: This is jargon from CRT TV days, the horizontal output yoke driver "SWEEP" tubes were huge, and for sure on Any TV 29" diag. screens and larger. 6LQ6 were in most homes in the USA long ago, and were used for transmitter tubes. (these tubes are now RARE)

Bias, = means the bias on the tube, that sets its conduction points.   (on, off or running CLASS "A" mode, this is what allows easy Linear amplification)  Even your modern FET uses bias, but much smaller, and for the same reasons.
Gain, all tubes and transistors have gain, that is there purpose even if used as a switch.  The tube has Gm and the transistor has HFE or (beta gain) They amplify signals as in this case. The bias on a tube can allow the tube to run in CLASS "A" mode.  (see data sheet on any tube)
TUBES , is not a pipe or cylinder, it's an  Electron Receiving tube, as seen in all RCA books of the era (see photo below) and  see Fleming above.
We call them transmitter tubes, here in Gren'go land.  The tube is just an  early technology, It is an early  electrostatic device, controls using "E" field in the grid , that even FETS use  today for enhancement or depletion mode operation, E field devices.
Electrons can be controlled with E fields or H fields.  (see how TV CRTs use yoke coils to sweep the   beam of electrons, my old scope uses E-fields only to sweep my beam)
Triode Tubes come in many sizes, from tiny to huge.  (I've even owned portable radio's with micro sized tubes inside, some had 3 batteries then, A, B and C sized, now you know why those  names started and are still sold today. AA/AAA/C/D but not B)
People  had portable radios, before they had 120v AC power to their homes, they all used Batteries. (x3) (I owned a hybrid once)
Our MAIN pair of  tubes are giant versions of the below, but with very high currents up to 500-800 mA paired value !. (I have even worked with Military gear that had sub-miniture tubes. and HUGE  water cooled tubes and Fluorocarbon cooled high powered transmitters)
Some of the largest tubes were used for cold war,  (base ELF transmitter for Submarines comms)  Some huge tubes have no solid-state equivalent, or affordable, like what is  in your microwave oven or in a huge TV stations or 50,000 watt super radio station. (KGO)?
See GAS found in tubes here.

The HV chicken stick (in real English, the High voltage safety discharge probe)  I call it the Brave stick.
  1. Rules for the this safety stick.
  2. Make you own is OK, use PCV pipe 18" or longer, for the handle and and old screw driver for the end. proble. Even duct taping the handle to the screw driver wins Duct tap awards.
  3. Pick a resistor to match the voltage you are discharging, 5kv is goal here, (so 1Meg  meg ohms is ok ) 10x less than the bleeders in the PA now.  (a Alligator clip works ok but Crocodile is best)
  4. Let PA sit  for 1 hour first. (then open cases or covers)
  5. Ground the chicken sticks ground clip to the PA case metal grounds, only. 
  6. Discharge all HV nodes 3 times for 1min, using this probe tip, make sure the ground is connected 3 times looking. (learn where HV exists inside the PA first OR DO NOT DO THIS ACT)
  7. No kids playing near by or Forest Gumps, messing about near by.
  8. Like guns (assume loaded)_, assuume the PA bleeders are dead. and the PA holds charge for hours, don't not trust your live to any bleeder.
  9. one other choice we learn is after step 7,  is to use a leaned long screws driver against the case ground and then land the tip of driver (same time) to the HV node just done in step 7.
  10. Spec. 1meg Ohm is 20seconds for 1TC, T=RxC, for 63% loss
The 1meg ohm resistor is there to not discharge the CAPS too fast and warp there internal CAP plates.  At 2 Time constants  86.5% of the charge is gone. 40 seconds time. 400vdc remains.  
4TC is 98.2% discharged in 80seconds (2800 v to now 50vdc) safe,  5T what is called fully discharged. at 99.3% 100seconds(almost zero not quite) If super impatient use a 100,000 ohm resistor (1watt or more) for 8second safe discharge.


This is how capacitors discharge TC= R  x C.


Pure data:   (First fix AC POWER IS BAD)
(In this case, my poor home 120vac power wiring) "CURED"

My SB200 A.C power line draws 1.13 Ampere (rms) at idle, not keyed. (receive mode)
Then when I run full hog Tx, (RTTY + 65w exciter drive) SB200 draws  10amps. (from my pathetic 16AWG power feed (AWG= American wire gage) NOW GONE ! (10awg now)
The RX voltage (weak too) is  118v that drops to 109 ac on TX.  From bad to worse, but the  has shack too many things on this feed.
That drop includes the current sucked by my exciter (100w?) ICOM 7200 and other things.
During transmit the filaments visibly dim.  at 1090va Line load. (this is NOT GOOD)  (that be a 6.3vac to 5.7v drop on them.)  (my drop is equal on both Hot  and Neutral lines, so is copper losses, I squared x R= 90watts wasted, copper wire in the walls)
The cure?
14awg  power 20 amp feed is way better., but better is 30amp feed, since I must pull in a new line (power) best is to go to 30 amp service. (Now way do I need 240vac here)
Heath only states, 16 amps is max. line current, (clearly using my 15amp service is and was wrong ,not to mention my DEN/SHACK runs off one breaker,  ouch)
Heath does not state minimum line voltage spec, but would be known  (calculated) by using the known rules of the filaments and max allowed sag. (in my case this fails.)
The  low filament voltage spec. is 6.0vac (+-0.3v) so that means line must not drop below "114vac rms", or the filaments will be out of spec.  (6.3-0.3= 6v) The transformer turns ratio is 19:1.
The cure for me was 75 feet of 10/2 copper ROMEX  feed cable (2 +ground) and  a 5L-30r , receptacle jack, and new 30 amp Square D breaker , now, I have 123vac full time no matter what. (at the RIG) No more filament dimming.... hoorah.
That is right a real 120v 30 amp power feed is easy to, from main breakers to the rear of the PA.
I have 14 more volts, now. 
MY  SB200 has its own private power feed.  Just for all my PA's
PLUS  my RF power meter went to 500 watts , on my tuner. wow ! (got way more plate voltage, too, as expected)
Now the fear of damaging my filaments ends. 
I do love this SB200, it is a really cool rig!! (I just added a SB104 and SB401, to the rack) Long live the memory of HEATH.
The 2 A.C. non NECA/UL rated line caps were changed too. (at the power entry point) Now you can keep the PA plugged in when not home and not catch fire.
The C1 and C2 caps need to be replaced with CLASS X1/Y2 rated UL safety cap.'s,  so of they can not short out,  sitting on the line power 24/7 .   The correct caps are sold at  Mouser, cheap.
Cap. maker  Vishay #
VY2103M63Y5US63L7

Tuning the AMP.
I first tune he antenna or dummy, to the exciter, with this RF amp off.  (SB200 off is bypass mode)
Using my AT2KD tuner and the 1kw dummy load active.   CW mode. (keyed up to tune) ( tune the tuner now)
I make sure SWR is good. say with 20watts exciter power. If not?, stop and find cause. (do not just slam 1152watts peek in to bad antenna or feed) Never use your rig at any time with the SWR meter off.
Do not tune the BAND knob , while keyed up, transmitting. (or you will damage the band switch) There are no computers here to protect you from yourself ,no plug and pray(sic) and  no SWR NANNY.
The SB200 2 knobs are  BAND at 40m and TUNE at 40m.  (example)
SB200 is turned ON now.
I then tune the AMP using the tuners, forward RF power meter, as my guide not Plate current, like the Heath  book says.  This external meter allows me to see forward and reflected power at the same time and is accurate. (I tested mine and it is)

I use the SB200 tune knob for max power.
and then use the Load knob to get low grid current Ig , in spec, and lower is better. keeping power high.  ("I " means current and "g" means grid.)
Never run the load knob at full mesh (to new guys, that means never fully load capacitance, on this gear CCW max is full mess, #1 on the dial , my dial has  protractor scale and 0 , is full mesh , max cap.) CCW means counter-clockwise.
Next ,I set my exciter to 70w, this is as high as I can go and not have ALC kick in.
Keep the grid current as low as you possible can keyed up,  the book for the gear says keep it  in the white grid square of the AMP panel meter, on the gear. (true words ) If grid currents are wrong, suspect gas. (purple/blue glow in the 2 tubes)?
On mine with 500 watts showing on the tuner watt meter, my grid current was 350mA, I turned the Load knob and  Ig drops to 200ma, before watts even starts to drop, this is the sweet spot; here (no need to run max Ig at all, as you will see)
I then flip the dummy load switch to antenna, and trim all the above, if needed. ( while not walking on other hams, usually on the highest end of band of choice...or as you discover it...)
Antenna's do not tune up , like a dummy loads, so this takes experience or better keeping logs of all tuning knobs, learned before.  (I log this every time is go on the AIR,  have 4 knobs I log, 2 on AMP , and 2 on Ant.Tuner)
 I upgraded my load knob scale to 0 to 180 degrees as seen below...

Ig = current grid,  as seen in all data sheets.  (Ip would be current plate(anode))
Full mesh?, take your 2 hands, spread your fingers, wide and mesh your left hand to right hand fingers, this is full mesh. (max "C") These CAPS are AIR CAP's. Air Dielectric.
At 40meters, the load on mine runs at 25 degrees.
 On a stock dial, is about a 2.  I'd say avoid 0 at all times.(aka 100% meshed, max. capacitance.)  I keep the load turned to the right, just a tad, as I see real power drop , just a tiny bit and grid mA alot . (ex. 40 meters)
Treat the grids with TLC.
Temporary ,now looks like this..Thanks to Ham!,
Kevin Mahoney KD0ILM / XV2KM

VSWR , SWR:
   if you use and antenna turner, all worries are over.! The PA is mostly safe.  If running low loss coax, most of your transmitter  power lands on the antenna, where most folks want it, and not into your home.
Read this page by Wc7i,  It really is the best of the best.  If nothing else, read the last paragraphs, please.
http://wc7i.com/reflection%20section.htm

RMS ?what is Root Mean Squared, (it's the square root of 2) press 2, on that calculator the square root button, bingo for 1.414 , now hit the 1/x button, called the reciprocal.  0.707.  (end math, just 2 button presses )
RMS purpose is only one , and is for heating up heaters.   an AC powered 100w rms heater is same heat output as 100w DC heater. (easy way to think about it  or are Edison DC watts the same as Tesla AC watts, sure they are if Tesla uses RMS meters)

Problems with ICom 7200, Send relays solved on my Panadapter page.

 




THE END OF MY JOURNAL:
73's to yah.

Parting shot:  For all my missing files or  404 ,etc, see here. HiHi



rev 4   ++++ 8- 12-2016  K5JXH

ALC works now and even works with my old SB410, SB104 and my ICOM7200 (with my ALC switch box)
Now the PA and whole rack, is upgraded  with full AC power, 123vac (@30amp true service) line voltage , all the times. I now get SB200 2700vdc power unkeyed, and now the tube filaments do not dim, when transmitting (yeah)
73's