K5JXH. , (boat anchor #3)
Topic, Heathkit SB104 ss solid state Transceiver.
A Minor Refurbish , repair and full tuneup.                                   

My SB104 Journal:
First up is m my slide show of my rig.
Next is the never do list (short) Never power the rig up backwards on 14vdc power, you will blow the rig up (will !), then never hot swap circuit cards (PCBs) last never change bands while transmitting.
Ham radio is all about, experimentation, and modifications, and learning.
I will only modify (mod.'s) a few things, keeping it minimal as possible as it a really good design.

There are some mods. I will index them first:
  1. The Mic mods end to end,  I added phantom power (a 3 cent resistor) to make my ICOM mic's work and my Electret mics too, and if running a relic microphone (the low z mic mod)
  2. Clean power, That DC power line needs some protection and filtering. (My fix is brute force not the better FCC CLASS B grade)
  3. The RF RCA  antenna jack mod to BNC , at rear panel mount, the no drill method. (done), You don't want that silly RCA jack to fall off keyed up, for sure.
  4. IC socket upgrade, this is not really a MOD it's a repair action, and removal of all SILVER plated, IC chips, (gone) black death syndrome.
  5. Factory mods, are here. (engineering updates by HEATH)
  6. Dead 5vdc (or way out of spec) there are fixes for this in the service bulletins.(in both supplies I now run 1% 50ppm/c Resistor pairs, cost me  a buck)
  7. Dead  or off spec. 11vdc (same Serv.Bulletin fixes) and mine....
  8. What if, the receiver front end dual gate MOSFET blew up, and can't find one, so you mod it like this.
  9. See PA checks here.
  10. My last mod.  I added a 1st "I.F" wire to my spare connector, from card, E pin 6 and I added PTT out so my linear SB200 can key up, 1st "I.F" is for a future Panadaptor (dream, to be documented later)

I have the alignment pages covered, and the files to them, and one way to avoid $1000+ signal generators.
My files section has the first ever 200DPI, schematic set on SB104 (non A)!
Failure cures for parts no longer made or sold. 
The cure for for a dead 5vdc dead or 11vdc dead.  ( more than one cure, and the 11vdc supply is not really up to working off a 12vdc battery, unless you convert to an LDO chips)

Finding good Boat Anchors is not easy,  lots of old wrecks sold for big cash,  I got a GEM ,  it now works like new.
The Reciever  front end is hot !, it really gets the DX. with those MOSFETS. (there is no noise blanker in mine or even any audio tone controls, just Gain AF GAIN RF, and modulation level knobs)
 It's really a light weight boat anchor, due to no huge transformers inside.
This rig really is nice performer, lacks an AM mode, but  who cares?  The dial turns like stick in warm butter, just like I like, and makes fine tuning easy.
It really is 3 knobs,  2 Gains, and Freq.  unlike computer rigs and deep endless menu's , so this rig as JOY operate !
No squelch knob or tone controls for audio. (no DSP noise blanker magic) It's raw and real and FUN!

Capturing multiplexed display on camera is tricky see next photo.  (this is a FREQ. Counter, totally)
 To be truthful, there is one set of tubes here, the red display are not LED but are orange  neon gas displays, called Beckman Panaflex  (yeah a kinda tube"with gas")

No internal DDS VFO was  invented yet in 1974 , it's still an analog   VFO  but the digital display does not lie, what you  see is what you got, at  0.1khz,   (so 0.5khz above, can be .4  or .6, this is the natural error. aka  ONE COUNT )
Once rig is warm, it does not drift, I will hold SSB voices well. (once tuned) (what just  short minutes warm up)  I put in LED's for the S-meter back light.
 (if you discover tests with external Freq counter is ok then  replace the  "A" card crystal   if off 1.000 MHZ exatly,  new ones are here. 520-HCA100-13X  with 50ppm , from mouser !)  (see the schematics for the test points)

First, I cleaned the whole rig and ,  it was only dusty, no rust at all , no spilled coffee inside. (In fact a great beauty ! after blow off and a 409 clean job)
Check for bad soldering. On this rig is most easy to do. Mine was perfect.
Modular Designed, (like the old Motorola TVs "the works in a drawer"
Very low IMD, ratings  (truly a nice transmitter that needs no RF final amp tuning at all, to operate) This SS rig, ends RF tube Plate/grid tuning
The Cards all come out, in hand easy, look for damage, overheated parts , then look at all socket pins on bottom for bad or cold solder joints, make sure the  missing noise blanker card bypass is in place, per manual.
Based on Capacitor date codes and all chips inside  the counter will have date codes !, it's  1974 chips  (mid year chips) So can be 1975 easy, even January.
If the rig has problems look for missing parts, mine had 2 missing resistors, (cut off mid lead) . This killed the VFO dead. (I found this easy with my scope, omg, no VFO out )
I made a fast check off least of Bulletins, and to my horror found out why my VFO was dead (found this first with my scope, the finally the reason why, below)
By reading all Service bulletins, I discovered SB30 , the person clipped out 2 resistors, R1251 (10meter fix) and got carried away and cut out R1252, killing VFO dead. there are non-A and A cards and they are NOT the  same.
The newer card has no attenuator  like the non-A does, and the newer card Q1251 is  RF buffer amp, so the last owner/fixer cutting those resistors killed my buffer dead. VFO Dead.
The new card below has 4 wires and not 8 with the band switch logic.
The SB30 needs to be revised, and mine is and is in my files section.  DO NOT CUT THESE RESISTORS BELOW.
 In factwhen doing SBulls steps, read the schmatic first, before cutting things. ( or removing things)

Below is new 1516 card.  (The below VFO filter card, is the new card.  Old card is  1203 and new is 1516 card ) DO NOT KILL THIS CARD, do not remove parts from it.
This card does not have the Atten. pad  do not MOD this card.
In the same vain,  FYI: the "G" board, has 3 different cards, so don't do any Serv.Bull, mods, unless sure it applies to your exact card by part number.
Many cards were revised 1 to 4 times to attain, SB104A ("A") status, do not blindly do the service bulletins steps.

Only missing feature , is optional and rare noise blanker card, left front. (mostly useless unless DSP class,"smart NB" but does cut cars generator whine or spark noise. ("can") better is resistor spark wires and plugs)
My rig even is missing the NB chassis socket.
This rig runs on 13.8 to 15v at 20 amps max, (and not on 12v unless you do the factory mod.?)
Any good PS will work with proper current. (adjustable to say 14v) I run a SMPS, but do so with a Fair-rite line donut (or coffin box ferrite) device.
It will not run on 12v batteries, unless modified.  The Service bulletins have at least 3 mods, to make the 11v reg work better.
Making a 11vdc reg with only 1 volt DROP is nothing trivial, see my fixes for that here.
Most rigs, not this one, are  designed to be on a running car engine or not running, with Alternator on line. (13.3v is the low spec. on alternators, so.....) and ours does , but not on raw batteries. 12v.
The only bad thing in the total rig , is the horrible quality sockets (IC) on Card "A" clock.  Cured.

What you got:
There is no 800+ volts to kill you here, no tubes to burn your hand, or burn out, or to test, or fool with ever.
It's all  solid state semiconductors end to end. (but I do love tubes, they are so old now, they are sexy, watch the movie "Frequency" to get a feel for that...despite wrong SBxxx rig used)) 
The only HV is 150vdc on pin A of the VFD? display,  one lonely white wire.  Really is less than line voltage, 120v times 1.414 (rms to peak) is 170v(p) on the AC line.
If 150v on pin A, is dead, that is a DC/DC converter card, a chopper there.  That is only a classic, Astable Multivibrator  circuit. This voltage runs the NEON "Panaflex" SP-352 displays.
The FREQ display is JUST a , quality "TTL based Frequency counter", with an accurate 1.0000 mhz master clock.  It has Neon digital displays,  that cost $50 used.

 The counter has (100hz resolution(0.1khz plus or minus) is all the counter can do, but is linear to a fault) It can be calibrated and the steps are in the book.
Keep in mind, the real freq, can be off +-100khz  1 count up or down. (so avoid band edges that tight)  If you want better accuracty, buy a cheap hand held Freq counter, with 10hz accuracy or better.
The VFO is fully serviceable , it's not a Contract VFO like my SB401 has. (good !)
No more brown phenolic shabby PCB cards, but now real quality fiber glass boards with silk screen markings and back side solder masked, OMG . (a classy chassis !) In fact mine looks factory new, still.
The IC's have sockets, (a nice touch that is, but is cheap brand was used, corrected by me, to machine pin gold sockets, from TE.COM)
There is a LED solid state  display  upgrade card, that is $33 raw just for the raw card,  I think that is to much for just a raw card. (and the days labor with huge 25+ numbers of wires to change.)
The display card has a rare DD700 decoder (HV outputs) the easier chip to find, if bad is called NSC DS-8880 chip. (and 80v rated part) (the DS7780 is  ok too) The 8880 is near mil spec)

It is a really nice rig,  easy to tune and use.  Plus very affordable, I own mine just to experience 1975 again. (QC QC calling from 1975, gear)
I have many 14vdc supplies(analog and SMPS) , and this rig runs on my SS30  Here is the noise to expect.
The Vernier dual jackson drive main Frequency dial, is smooth as stick in warm butter.  I LOVE IT.  , and the 3 knobs to turn simple, and super sensitive, FET front end. wow.

Big Electrolytic CAPS:     (we never just assume they are good, we means technicians, we check them fast and easy with a ESR meter.)  I have  box full old caps, all passing uF test and leakage and fail ESR.
There is at least one cap (C12?) added in the SB, that can be refreshed to get clean 13.8v., check the ESR carefully on these,. Just the huge caps. and make sure the SB recommended ones are ADDED.
Keep in mind C12 was added in the first year of production so most have it, or  were added at the rigs first service, due to the Service bulletin tells you to do that.
Manuals and SBulletins  seen here.)  Check all Electrolytics for ESR readings. and test for excessive ripple.
I don't like bad ripple in supplies.
We check 3 things here, ESR, Ripple actual and volts, making sure the supplies are IN SPEC. Do not allow 5v below 4.75vdc better is 5.0 The lies about 4.65 are just that. (only military grade 5400 chips run lower voltages!)
This gear uses commerical chips not MIL !

Tests:    But inspections are first. Look for damage, water damage arcing damage,  things broken , melted or burned up and missing.
The rare to find extender board is not needed, unless you  must  do the "G" card tuning and alignment steps, other wise trouble shooting is easy , just solder tack a wire to any signal you want and put back the card, look ma, no extender...!
The "G:" card is  hard case, to align with 17 trimmers... for sure...
Inspections: (and tests):
First do the ESR tests; the electrolytic (large)capacitors first (CAP.'s for short) above first. ( a special tool all shops need)
All grounds checked for tight and good.  Every ground tab that is on a screw on the bottom,needs to be tight.
Inspect inside the gear for any thing loose do not power it up if any coax cables on the right cards are loose, fell off or are missing.
This RIG uses a large number of sheet metal screws that love to go loose after shipping the unit. (note the better machine screws of old gear, (screw,nut,lock washers)
Large Resistors, all check out. ok. (mostly just the big ones fail, from heat, and are not many here)
Check the Relay , Inspect the relay for damage. It don't do much until in HIGH power mode,  in low it selects ALC mode. In HIGH it sets final amp bias.
"Today we have access to excellent gold contact plug-in relays (such as Potter Brumfield R10-E4-X6- S320 12vDC) with more than adequate current/voltage ratings. K5BCQ says."
The socket up grade, is nice here. (In most cases, it's good as it is.)
Omg (oh my gosh) ,  no  more 100volt  relay coils, what a blessing that.
Inspect the small number or rotary wafer switches for damage,  many can be cleaned easy with just alcohol.(pure and oil free)
The band switch, clean it now. (use no silver dip{mom's} (its an acid) or oil or any chems or powerful solvents here, ever. Alcohol is plastics safe.
The VFO ( has  "MOLEX" pin connector on the side. Called PL201, see Q1205 to 12. for details. J11 is the LMO out plug/cable and you can unplug it and test it directly with a scope. (check for drop outs,etc)
Make sure the 2>5 jumper in in the rear panel Jack "ACC" and the VFO jumper IN TO OUT is in place ,nearby.
The main book, on the RIG, tells you how to power it up the first time, or after a long shipment or parked in the attic for 40 years. (pull all PCB cards, it tells you and then watch the DISPLAY.... read it, do it...all.)
My VFO was dead, at the next VFO buffer card, due to sabotage.

I power it up and on for the first time now: (on/off switch to on)
All lamps glow, and display 7 segments in   displays, and meter back lamp. (the card out test has you check for FFFFFF on the display, F = all ones in hexidecimal math.  1111 = F. welcome to the digital world)
After completing the manual stated, cards out tests.  (this prevents damage from cascading)
We next test the rig fully assembled.
The panel meter shows about 14v when 13.8 button is pressed , power on button pressed on?
First are power supplies, (correct voltage and ripple)
Here are my actual RIPPLE readings. (all with the SCOPE)
At the bottom of chassis, the "B" card is the regulated output, pins, (read the schematic seen in my FILES section below)
Check all 3 power lines, on the TX-AUDIO_regulator board.  5vdc , 11vdc, and  13.8vdc for out of spec. voltage or ripple excessive. (oddly there are no filters caps at this point (PCB card) they be hiding else where...
The spec. on 5vdc is absolute minimum,  4.75vdc, (spec is, +- 0.25v) not 4.65v as many claim and are wrong. The 54xx TTL for military does do 4.5v on the low side, and do work there.
If out of   spec. correct this first, bad caps or bad regulators.
  Here is my DATA:
Check the VFO  next , see that it works, no drop outs  and and the display freqs. all look ok on all bands.
The VFO has a dual JACKSON drives, a reduction device, make sure it works and the main shaft turns from min. to max. if not correct that  next, we soak the Jackson drives (ball units) in paint thinner (oil based) and carefully re grease them.
Do not attempt to re clamp for change the ball clamps in the Jackson drives, just clean them and grease them.
Do the 0.65v transmitter power transistor bias checks. At L952/953 shown in the manual and below at key drawings.

A scope, is mandatory, as only a scope can tell you if you are reading just a bunch in noise, (hopeless) or a real signal and clean. Not owning a scope is like driving blind. (really)
If alignment is off, you will need  signal generator, the 100khz , square wave generator method is near useless. I use  and my scope and DDS.
Lacking a scope you  dont know if you  are tuning real signals, or noise.
But the first book shows how to align the rig, lacking any tools at all.  (the non-A book tells how) It's crude but works. ok,  if just ok is good for you, do that.
Once aligned the Receiver works perfect, using a real antenna.
The transmitter will be damaged, if you TX at full power, with no dummy load,  band switch transmitting or transmitted in to no antenna , or shorted feed lines or a dead external tuner.
Test the rig in to a dummy load, the attempt to use low power into the antenna feed line, see if you get crazy high SWR, if yes, stop , fix that now. Do not blast 100watts 1dB, in to a dead or shorted antenna, etc.
Do not key up the transmitter, with the rear Antenna feed line missing.
The TUNE BUTTON, do the tune (a button selection ) tests.
 The tune button activates low power mode, CW, so you can safely TUNE your antenna tuner, if not using a dummy load.
Get low power working ok first, before attempting high power mode.
The right bottom  marked POWER /HI , is the High power RF button. 100w
I check for clean RF with a scope and using FFT mode in my scope.  Look for illegal spurs too high .
Photos:   This is the band switch , to clean, easy to  remove. I use only Alcohol.

So by now,  the Receiver works on all bands, then you test the TX on all bands, and most work or some are weak no matter, we then tune this puppy up.
Make sure all internal Oscillators in the gear are working, no dead or missing crystals.
Do all checks in the text file that I modified. see my files section below.

Skips: I don't need VOX ever. (PTT rules my game or  brass key if off my rocker.)

Alignments:  (see my new DDS page, how to use this technology and save a bundle of cash)

The 3 books show the alignment 3 ways , 4 ways if you count the statement about using a real signal generator. (in all 3 books) the 3rd book is no book really its a sheet, a service bulletin sheet.
The whole topic is fully covered in the DDS page above, below is an overview.
In truth here  are all ways, there are to align: ( the only hard part is card "G")  the TX side takes almost no tools at all ,its only the RX side we have so many options.
One way in the books are to use the 100kHz , clock square wave as a signal generator found in the display counter board,  I'll skip how to align using odd harmonics of a square wave, good luck if you try !
  1. Book 1, I call this, it's just what the SB104(non-A) operator manual states (a prime goal then....btw), no tools at all, read it in my files section,
  2. Book 2, the SB104a  , read the book (alpha book) in my files section. above. It will ask for VTVM, no, don't rush out to buy one, use any modern DMM on A.C volts. and my data seen here,
  3. SB11, service bulletin #11 calls out a huge list of tools and does the alignment with some of those tools.  (avoid this SB11, is my advice.) the SB11 is here with my comments.
  4. Use a  DMM and  good commercial SIG gen you have and do the alignment using that, or use a good scope and a good sig-gen. (good means calibrated and clean signals)
  5. Use a DDS as a SIG-GEN.
  6. Use a sweeper + spectrum analyzer per below. (a huge cost for equipment but is best of best way)
  7. Last is just to make this complete, and this is aligning the radio , with no tools (sure a slug tuning wand) but just listening to radio stations you trim both ends of all BPF filters. (I will pass on this...)

The great K5BCQ way,  KEES A TALEN

Thanks Kees ! I love your simple and clear, easy to read web site, even very efficient Black/white photos, of the scope screens , nice.

Then there is JANIS nice site,  and a great review on the SB104,  priceless advice. THANKS!

I will report all I learn on this rig. When I Get a  "Round Tuit."!

The way do to the MDS checks is covered in my DDS page.  (takes real  and calibrated tools to do MDS at 1uV sensitivity)

MOD.S (modifications) (under development)  I put this device on the DC power cable so is really not a rig mod , but just the power line mod.
For sure few, it's a nice rig in its own right. 
A modern mic connector, mine is missing. (both ends gone)
I put a Fair-Rite donut #43  for a $1 on the 13.8vdc main feed line and the SB (service bulletins) added huge cap.  
The donut keeps RF in my box and RF from sneaking from SMPS to my box (sb104 box)

The Ferrite core fix.  (I plan on using 2, on at the SUPPLY 14v, and one inside the gear,  due to horrible filtering on my SB104.... to be addressed soon and documented.
Set the TURNS on the CORE  below to what ever band you want to be cleanest. 3 turns should do it.  (that ohms line, is the impedance the illegal signals see,)  ..

I run this device, #43 core.

did the RF antenna jack mod to BNC , panel mount, the no drill method. (done)
I may do the added TX/RX relay socket mod , that is  add the nice socket mod. using the nibble tool method. 

The Next MOD is the Micro-phone and Jack mod, most gear this old the jack is broken and dead, or you can't find that relic , obsolete mike  2pin connector. sigh.
My rig came with a  1/4" stereo head phone, jack for the MIC connector. It's in the junk box now and a new ICOM jack is  installed.  (8 pins)
(ICOM/Yaesu/Kenwood use, the 8pin Falcon jack), (the OD is .625") aka. the C8P , jack.
I wired up 4 pins,  (phantom power and mic signal to pin one,  then 6/7 ground and last PTT on 5) Done and works.
Next I will cover microphones and the phantom upgrade, and choices.
The STOCK Heath 104 has a HI Z, input  MIC radio. (Z means impedance) So needs no changes, for impedance.
To get Phantom power wired in my old SB104, I connected a 4.7k resistor to 11v supply and pin 1. BAM it works and sounds perfect. 
The old gear (tubes) the MICs are tuned (scaled) for Low Z, 600 ohm Mic's back, so if using LOW Z. passive dynamic "coil mics" then change this next resistor.
Last is the mic. Low impedance mod, one resistor change, R203 on the "B" board from 22k  to 51k ohms (avoid this , get a better mic!) This is for running very very old low Z mic's..

Examples of new gear help:
The below shows my IC7200 and the 1k ohm input impedance and is the phantom power seen here.
 See Icom's way to do  Phantom power using 8vdc supply, and how it shares the mic signal line, as you can see its cave man simple. 
C1033 is the DC blocking cap. Icom has 1 OP,AMP and then DSP

the Black wire  "MIC" here is  mic, at pin 1. The wire is power and audio , all in one wire. (aka, Phantom power)
Here is my new mic, wiring in my SB104,  only the red underscored wires are connected, below.
The below is a ICOM page snap shot HM36.
The IC7200 , hand mic, and shows me how to wire up any new mic.  I wired my SB104 to mimic this ICOM standard. Just  4 wires. Seen with red underscores.

I got this working and SOUNDS GREAT.  (I talk in to my RF dummy load and the other RX hears my voice and I sound like and old fart, but clear)
See my microphone choices and modifications page here.( See here for ELECTRET mic, upgrades.)
This mic now runs all 2 of my ICOM rigs, and the SB104,  with great sound.
End custom mod.'s begin factory mod.'s.

The Service Bulletins: Factory UPDATES. SB's for sure.
 Each SB Mod, applies to just to one of  up to 4 cards, boards. PDB.  Be sure the card number matches.
The SB only applies to the SB104 non A or to the "A" model do not mix them up.
For example "CIRCUIT BOARD G [PN 85-1708-2] " do not do this step to the the other 3 "G" boards used.
So I warn you not to do the steps blindly , make sure the revision is correct for your card.  Do not do the SB30 on the VFO filter if the card has 4 pins, DO NOT. as was done to mine...
Some Factory Service Bulletins,  I read them all and see if they apply to me, for bands I use, or problems I see using the rig.  They have lots of 10 & 15 meter band mods.
I found out that my April 1975 rig, has most of these UPDATES already.  Mine is  a NON "A" rig.
For sure test your gear out 100% and if there are issues, read this first to see if your issue (problem) was solved long long ago.
I'm about 1/2 through and only one resistor is wrong, the Audio speaker driver, is not .82 ohms, not .33  R-532.  (parts on order)
These really are just  engineering change notices and also productions changes.(many apply to sistor HW104 too)
SB-104-10  (only if VFO drifts)
SB-104-11  is the updated alignment steps from HELL here it is with comments I made and in PDF form..
SB-104-16 (if true)
SB-104-19 (SSB)
SB-104-24 (all)
Do not do SB30 on the newer SB104a or the VFO will be dead after.  Do not touch the 4 pin card !
SB-104-39 (huge)
I will stop here, as this sure is a  cold winter project.
These checks above or crazy tedious,  I had to look for the part on the schematic, then the location diagram and then find it on the card.
I read the schematic, and did so to see if the schematic was corrected and to see if the mod makes sense. (oops some don't in fact it killed my VFO filter dead, previous owners)
Buy and large all mods are done to mine, on or after April 1975. (my rig has  tag telling build date)

All FILES and documents, related to the above .

The schematics are here,  and I have added red stars to all Test points in  my schematic, that are stated in the official alignment calibration steps.
Most of the alignment steps only connect to external jacks, and what-not, see TP-A and TB-B in my new schematic,  I will add the secret crystal calibrator test point soon... on the "A" card.
The Web Schematic 104A.

My SB104a schematic  revised and added test points for the alignment.

Assembly Manual 104A. (real) With parts lists, and all numbers for parts.

Same manual but 8times larger but fully OCR support can be searched for any text (control +F). 66MB large !

This text file  below is a gold mine, a treasure.  (I show tricks to do, lacking a $500 -1000  SIG GEN.)
The Service bulletins and all alignment steps and trouble shooting hints. in text form. 
This text file covers common problems and solutions to each. 
I made a PDF out of just Service Bulletin 11, alignment with comments.

The stock 14v power supply is here. HP-1144

My new 104 non "A" schematic set, (I scanned it and is huge 35mb PDF at 200dpi 1:1 scanned )<< the best quality on the net. (schematics are best, at HR, highes resolution !)

In the above new set is my full system schematic, for the Non A. (this year,  they do a quasi- schematic and block diagram)

And last, the full non-"A" SB104, alignment pages in 100 DPI resolution.(10mb size)

Key drawings:  I used 16ga house wiring solid copper for the jumps then add the 10gage(AWG) cable, and my 20Amp Maxi fuse, and toroid. (photos soon)

.push tune or key up to get .65v

This really is a buy and hold rig, if ever there was, this is it. (and the SB200) both run on 120vac and do not pop the breaker, hark !

POWER SUPPLIES (14v) ! (no lie a linear supply is best of best) but I will cover why, and what options are available to you.
My SMPS noise (~10mv) does NOT GO PAST HERE, ever. (by design)
I upgraded my SB104 to work more like new rigs.
This rig does not run on 12v , it needs at least 13v to run at all, do to the lack of LDO regulators., nor does the rig have line filters to allow rig to run clean in a CAR, (any vehicle that all make HUGE conducted noise)
This rig is only designed to run with the HEATH external power supply (costly too)

Most modern radios run in VEHICLES and that means 13.3vdc to 15v dc,. running and most gear (not this one) run down to 11vdc just fine (ignition key off as the battery drains., but not SB104. See 11v mods.
( Modern gear that is designed for CARS do that; (do not plug your radio into 24vdc Aircraft)
If you have  6vdc Desoto car? , buy a 6 to 12vdc converter, but be ready for HUGE NOISE !!!
I run mine on 14vdc. With no LDO mods.

 The Astron SS30 can be had for $100 + ship,  a China "Mean Well" $50 + ship, (I have these and others...)
You  are buying a power supply 13.8v for many reasons and issues, when you buy a power supply.
  • AMPS,   (extra is better) Your transceivers spec. needs and tad more. (or for your needs next year, with newer gear )
  • Quality ( a  moving target today)
  • The "Crowbar" option  (if the series pass transistors shorts ,do you want 20vdc or more hitting your $5000 ICOM?  I wager not...? Best practice is to tune the Crowbar, for bet ops, TX and RX, so it don't false trigger.!
  • Current roll back, more than a silly fuse.? It can be a blessing.
  • Over temperature Shut down (supply, internal)
  • LOW NOISE. (for sure, running relic gear,can be hard,  but hang a scope on the pins, got noise?, over 30mV? Did you add the Toroid donut Ferrite device yet? (I run a Toroid in the middle of my power line)
  • Must run ok on low line voltage, common in shacks lacking 240vac MAINS.
  • A steel case, my SS30 is (and works quite well)
  • A recognized brand?, you're kidding?, they are all made in China now. Who  knows what you will get?.(Mom to Forest Gump)  (click here, recognise any names, here? me neither)
Some unregulated supplies can work too,  @14vdc,  It is just a transformer, bridge rectifier and some huge caps, and nothing else much, it sure can work.
It's 14vdc  radio off or a  tad less , transmitting ok, if you purchase extra rated amps,  (go one size larger than the radio TX needs)
The trick with UREG supplies, is finding one, that does not exceed your HAM RADIO input spec. voltage, under all operating conditions, TX and RX and Full modulation and , AC line high and low normal conditions, Good LUCK THERE.
SOME SSB rigs have 60 and 80 amp surges,with (linear Amp rigs) Here is where I've been working, no room for a pro grade line filter.
The C12 is Service Bulletin mod (under Chassis), and I added 2 more caps,  the blue metal film, and 1 ceramic.
 The 3 way cap fix, cleans up noise in 3 different noise classes. The TVS solves reverse power connection and also overvoltage failures.
Keeping leads short (and inductance low) is very hard giving the physical lay out. The blue cap is metal film.  The cermaic is .1uf.
Keep in mind you are filting incoming noise and out going, that is also why the power cable has the Fair-Rite donut.
Click the photo for more information.

POWER CORD design and filters:  (use a low ESR cap, the best you can find)
At the rear power jack I run a Littlefuse(tm) maxi  #MAH1, inline fuse holder , with 20amp maxi fuse. and my #43 toroid. To my SS30 supply. 
My power wire pair is 10gage and 3 foot long. short is best. (keep PS, 1 foot away from radio or more)
You can see the RCA jack (ANT) is now BNC. (will never fall out transmitting) The bottom RCA jack there is seperate receive antenna, that I will never use.
The TVS (just now putting it in) and line cable fuse, can save any rig from damage if the main supply goes, berserk.
The best filter for HAM usages of SMPS below,  and might fit,  this one is small.
Power-One (BEL)  makes and affordable filter, 2"Lx1.5"W & .46"deep.  smaller than MOST.
This would be a great filter, on any DC gear lacking CLASS B , filters like mine.
I am attempting the usage of brute force cap and toroid filtering and testing its true effectiveness.
My finished cable is here. (click it to zoom) (all ends tinned) (all soldered with a  100w iron)
 Shell not snapped on yet, so you can see my clean solder work. The Toroid is seen clearly. 3 turns. Using real Pb lead solder 60/40, for max strength.

The SB104 was NOT designed to be connected DIRECTLY to any Vehicle, 12vdc power system, It was NOT ! nor connected to any SMPS.  (do not forget the fuse on the red wire)

What if the 5vdc regulator MFC6030 chip blows up IC202 on the "B" card, you know, the one you can't buy now.
The chip should not go bad, it has  high rated input of 38vdc and 14 is there. (IC202 loafs at  12mA current (can do 200mA), due to Beta of 100 at 1 amp Ic, on the transistor driver MHT9210. (2n3055 good sub, T03  or TO204 package)
The sure fire way to blow up both 5v/11v regulators is connecting the up backwards to the 14vdc supply. (I put a TVS diode on my power feed ,  to avoid this, damage)
By the way stock normal Motorola  IC202 output can be tweeked to 5v exactly by playing with R252/253 on the "B" card.
Here are the 2 easy solutions. (mine and then Fairchilds (ON-Semi)
My easy fix is a 7805 :(uses as many existing parts are possible)

Better is this:
Fairchilds. (will take changing the series pass transistor, a harder job)

I made a data sheet on MFC6030 (found on the dark side of the moon, under alien wreckage.) 

11VDC horrors and issues: (only when not using the HEATH provide power supply)

The 11vdc regulator is  harder nut to crack if  it blows up.
Same "Chip not sold" again, and  running on a solid 14vdc like me,  then the LM317 fix works for me, but not  on battery power, say , in a car, or field setup?, it ill not work right.( at 12v)
If the LM317 is ok for you, runinng on say a 14vdc hard supply then set the 2 (r1/r2) resistors for voltage out, using the data sheet equations.

It would need an LDO upgrade to do that, although easy in concept, is hard do to the lack of  wire leaded parts, (aka. through hole)  fitting SMD parts size of a flea, to the "B" card will be problematic.
One could make a SMD card then have wires leave it, and fit to the IC202 through holes, sure.

The dead 11vdc REG might be cured using an LM317.  But how?
 The trick is getting the LM317 to work at 13.8v input !  (note that's only 2.8v more that 11v)  at room temp.'s the LM317 drops out at 2v (Vin - Vout) at a 1 amp draw.
The drop put on the MFC6030 is not totally known, that data was lost in the sands of time, but the data sheet shows 3v is what it can be.
The data sheet states worst is 3v LDO drop. (I'm sure that is at a full 200ma out current)  In truth is better because our output currents are tiny, (1/10th of max, ) 
The problem is no LDO, I can find will run with 12vdc input and 11.0 out. (with a series pass transistor for the full amp rating of  pass transitor that is 4 amps.)
The best I can find is the Ti.com LM2941 (TO220-5), designed for cars, and 12v vehicals, with full protection from same. 
 Getting this to run off a 25% charged battery at 12.06 volts, is not easy.
The nomral current out of the Q2 (2n3055) is about 1amp, so the base current is only  ( 1/50 or 20 mA)  is only 1/10 max. on the IC.  The Q2 can run up to about 4amps before the IC runs in to overload.
So the total dynamic range of current on the 11vdc buss is  1 to 4amps. (the 11vdc supply only runs RF low power amps, IF and oscillators. never does it sourch TX power)
One could try setting the LM2941 to 12vdc out (R1/R2)_, and the VBE drop on the 2n3055 will give you 11.3v out. 
So the input to the radio would need to be over 12.11vdc. or the LDO Chip drops off line.
The red line is what we have now,  adding the LM2941, can work here, at 1amp in the series pass Transistor and 50 (Hfe) times less base current, drop out is 0.11v.
Tune R2 for 12vdc exactly and the errors in the VBE line will drop you to 11v, and can vary , but is now .33vdc, error,  and at 4 amps (impossible in truth) is 1.32v drop error on the .33ohm resistor)
12v is 11.3v on emitter and is 11vdc on the out wire there. marked 11v.
at 4 amps (impossible)
As you can see if you run a 14vdc hard supply on thie input, it works perfectly.
lets calculate R2.     R2 = R1 ((Vout  / Vref ) -1)  (page 13 in Lm9241 data sheet )  The data sheet flat says to use 1k for R1, leaving just R2. Lets do that now...
R2 = 1000 x ((12/1.275) -1)
R2 = 1000 x ((9.41) -1)  {8.41 now} doing in order of precedence correctly and doing it backwards too..(as s cross check)
R2 = 8412 ohms .1% 1/8  low ppm/C rated both. means no cheap resistor, like carbon types for 3cents.  Best is to find standard sizes like .1% 50ppm for 60cents here.(8450ohms is a standard size)
end theory... on how I'd do this , if I had to.
Keep in mind this does not need be accurate regs, they only don't want  11v wandering from 12v to 16v seen in a CAR.
Keeping the output at 10.5 to 11.5v is all that is needed. The TX currents do not run through this node so , is easy to regulate.

All this is theory, I have not built it, yet. Do you own bench testing first.
This chip has over voltage input shut down at 26v (over) and reverse power connected to say a battery connections wrong., and 60v surge protection (jumper cable wild actions , and Air.Cond. clutch inductive kick backs))
The output is current limited too. and has a thermal overheat shut done.
It's  a dream chip, no lie. And was Designed for cars. (12v vehicles)
A better design is to make voltage sense at "A" feed back to ADJUST pin, for full 100% regulation, if in a OCD mood.
I am now told the LM350K chip works here (the K means steel TO3case) and is now harder to find.  Ebay has loads of them,   Ti.com still makes and sells the TO220 version. (if drilling your chassis sounds fun like)
It can hump out  up to 4.5amps but we only need 1amp. + tad more) for sure its an amazing obsolete part in this  steel case. (it will burn  near10watts out the case, in the 5v position, use good heat sink grease. )
The 350k you can just plug in to the V1/2 socket with the pass transistor pulled and then a new voltage divider added, that looks just like the above. See TI.com data sheet page 7, for resistor calculations, simple it is.
Found one on feabay for $2.25 delivered free. (for my spares)
Now begs the question will it work in the 11v socket?, yes/no !, down to 13vdc it runs, at 1amp  Drop diffr. is, 1.75vdc or 2v if super cold. The radio will work best at 14vdc on the main line. 
It wont like sitting in the field, connect on to a battery of 12v. Not at all. In that case, the above LM2941 beats this hands down..
In almost all cases, the chips blow up if and only if the owner connected the rig backwards to the power feed. 

I Got one more fix.  A blown front end.  Those RCA 40673 Dual gate MOSFET's

For a dead receiver pre amp  card the Dual gate MOSFET burned out. (shorted?)
The FET has 10vdc gate protection , internal diode clamps.  Almost  like TVS's function, but can be blown up. (TX'ing in to it would smoke em', as would lightening hits)
The RCA transistors(datasheet) Fets are very hard to get, but not the surface mounts.(today)

The  BF chips SMD chip versions can work  too , adapted. like the photo below.  
 The SMD (SMT) MOSFETS, come in 2 class sizes,  the large size like below and BF992so small (3mmx1.2mm) only a flea can carry one, as seen in stock at Mouser.com.
The BF98x, is large SMD."surface mount devices"
BF981 but are NOS only. "New old stock"  and has 1/2 or less the gain "gfs" as the original.
The better  3N212 has 17-40m gfs.  (can be  hand picked for best gfs, gain too)
The better sockets are at Mouser.com, in teflon, for a buck, not $20 from fleabay scalpers.  Best is no socket at all, to keep leads short and parasitic inductances low.
Near hits, try them? HF only, not saying any work above 30mhz.  See  list of Mosfets here for parameters
To say the least about testing these parts, use the MDS checks for receiver sensitivity, using a real signal generator at 1uV input. 1.8 to 30mhz.
  • RCA 40673 , the real thing.  As found in the SB104's  (400mHz,VDS=2v,330mW,-2v gates thersholds and 12,000 umhos GFS.) Note the 40m is better gain above.
  • NTE222 or ECG222.
  • 3N204 or 3N211,3n212,3n213 family up to 317. See them all here.(I say this 212 is best part to buy and find )
  • Motorola SFE1612,  rare now, see here.
  • Motorola MFE3007
  • Top cross refr. sites show up these, 3SK45, all Japan top brand makers like HITACHI, 3SK-xx, where xx is: 37,40,45,51,61,74,77.... as you can see these were made as family of  MosFETs too.
  • 40823 more rare.
  • No lie , there are people hording these rare parts. (to make an easy buck)  best is find some  cheap, in advance of need.
The  UGLY fix.  Dead dog, style. (using wrong packaged part same in a socket standard.

My actual RIG, first day, owned, problems and cures.
  1. Supplies check good.  13.7v, line (noise < 20mV,  11v = 10.45 (noise <5mv) , 5v,= 4.78 (noise  30mv @60khz) With my SS30 as input.
  2. The Receiver works on 40m meter is live with signals on 7.2mHz. Strong and clear, LSB.  (8pm light off)
  3. The Display test (manual book stated) , for FFFF passes. (proves only the display card in panel is ok.
  4. The 6.000 mHz test in the book alignment page pre-checks fails.
  5. I grab scope probe and find the Signal,  to the counter is dead ! At coax pin to it. I trace this signal (schematic and real wires) to the VFO buffer card. It's dead too.
  6. I check the input to the VFO buffer, and its perfect,over 1v perfect sine wave.,  I look at this goffy card, with 1 transistor and low and behold, some bad guy, cut off 2 resistors dead center of card, (I wonder why?)
  7. I put them back (1k and 470) cost me 6 cents,  and bam the output of VFO buffer is perfect.
  8. But the display does not count. (stuck at 444..)
  9. I then scope  the whole chain of decade counters here on the "A"card  and output pin12 on (IC101 is  ok)
  10. I find most things ok until IC118 pin 12 is dead. (Q output) (no extender card, so  with the power off, tack solder  a wire,  card back in,  power on , scope it repeat , method. (slow yes)
  11. In a flash back from 1974, I remember then in our factory, our buyer, bought these chips in silver plated lead option,  and like then, as now, the book says, do not use silver leads in sockets they are direct solder only. (fail it WILL)
  12. After discovering BLACK DEATH, syndrome, I buff the IC leads (pink pearl) and the counter works in fact the whole rig does. even high power, TX is 100w + (whooo whoo)
  13. 7  new chips on ordered (and received) and 7 new gold machine pin sockets, 14pin/0.300 wide skinny DIP.(aka, 7.62mm wide) $1 each at Mouser.com. 571-2-1571552-3  
  14. The chips (all here) are hard to find, got 74LS196 more easy for 53cents.  I side stepped all $50 scalpers... hihi
  15. The front panel display is dead accurate, it shows, WWV 15.000mHz,  zero beat,  and with-in 1 count, (+-0.1khz ) the right zero is 0 at zero beat, dang what  trip that is? (the alignment , 1 step, puts this dead on)
  16. Last is a full alignment per the original book using a real SIG GEN, and not that infernal 100kHz square wave noise maker .
  17. I scanned my non "A" radio schematics, and is in the files section here. in 200dpi resolution too.(sems

The whole center row of chips (CARD "A"), 6 total, the sides of the leads (pins) are all JET BLACK (silver oxide).  The new sockets are so  tight Like10 times, maybe enough to work with silver, but I didnt change it.
After the cure, Bam ! 15mHz WWV  is 15.000.0
The "A" counter card facts:  You know the black-death chips fail when you flex the "A" card and it fails. I solved that.
(watch out for date codes, mine are all 74xx and means 1974 work weak xx , do not let the 74 fool you,it  is the year.  443-xxx is the Heath numbers. 
Keep  in mind 54xx is mil spec, parts, and are SUPER GOOD.
The secret part numbers are documented in the operators manual, on page 279 and page 280.
You can in fact mix some 74xxxx logic families. (a complex topic in so many sub families).
The thresholds all work  ok in some  cases VIH and VIL, but not the fan out. fan out is the loading factors on output to inputs,  IOL to IIL (and passing VIL thersholds, can be problematic).
The Fan-out rule, means that many logic families TTL can not drive say x10  7400s, for example the 74HC can drive only 10x 74LS and only TWO, 74Sxx chips. but the input to 74HC is a floating gate and no current flows, DC.
For the TTL logic beginner, FAN-out is loading factors (easy to understand using load word) and is the ability of the driving chip to handle many other chips, in PARALLEL. (and will be less!, at super fast A.C speeds"frequencies")
To mix logic families, we must  know these rules.  I did change these BLACK chips, to "LS" with no problem.  but IC112
 I am running 74LS196s now in the center row and 74LS73s IC118.., with new sockets, the factory "A" card has horrible sockets, nice idea, bad execution.
 I upgraded mine to what are called gold machine pin sockets, on all SILVER leaded chips are gone now. These new sockets are best of the best."tight fit they are, wonderful tight"
The tricky chip is IC-112, 7400, it is,  as it runs the 1mhz crystal oscillator and using the wrong chip here can kill that oscillator, 74LS00,74HCT00 (for sure fails) the 7400 is used as linear amp, and must  work in this mode.
I'm also fooling around with 74HCT75s lower noise. (the 74LS75 work already)I just put in the 74HCT75, they work.  HCTs are newer and can be more easy to buy. Just don't touch IC112.
Success, my "A" counter  card is NOW solid like ROCK, I can flex it wildly while turned ON, and it's SOLID no drop outs  ! Click photo to see original card and the silver black death chips issues.

The LS logic I used , reduces loads on the +5vdc supply and makes less noise (radiated and conductive), (good for radio work, that)
Note how Heath avoided 2 layer boards, using simple jumpers.  and the usage of ferrite beads, to cut RFI.
Gold sockets from MOUSER.com and new chips found searching the WWW, end to end.  Found some in Dallas. BGmicro.
The machine pin sockets are super super tight fit, they will NEVER fail you!
The rigs display is in Mhz  and looks like this with the coax RCA pulled.  The top counter is from China, from SANJIAN Studios, off Fleabay. it works great see my revised doc page on it here.
As you can see with the RCA pulled below (and the internal counter has no input) the 6606.6 is seen here (tuned to 7.000 mhz exactly) the Freq. counter shows 10.394 mhz. (my new scope shows the same)
1975 smacks into 2016:

End  "A" card and all.

The PA card in the transceiver has 4 power bipolar transistors.  2N6456 NPN bipolar , Q951 to Q954. Covered 100% in the free manual, and all details there.  2N means standard USA JEDEC part made by many makers.
The 4 transistors are stud mounted to the Heat sink,  each handling only 25watts of the 100watt total.  (Transistor NPN Silicon 40v Ic=7.5a Po=50w Rf Power Output )
They can be had even today. ( price no object easy)
These 4 transistors are only used in high power mode. (button front maked "HI")
Before condeming any PA,  check DC power and for noise on the power bus,  bad caps in the power supply are always a first check thing, then check excitation for sure in not "HI" mode to be sure that is good first.
Then PA checks. (transistors)
I use a scope to check for clean excitation first. in non "HI" mode. (if wrong ,weak or noise ridden we fix that first)
Heathkit                    JEDEC # STD.                   Acrian & Communication Transistor Corp.  , and last unknown by me.






Original spec.:

Material of Transistor: Si

Polarity: NPN 

Maximum Collector Power Dissipation (Pc): 75 W

Maximum Collector-Base Voltage |Vcb|: 45 V

Maximum Collector-Emitter Voltage |Vce|: 25 V

Maximum Emitter-Base Voltage |Veb|: 3 V

Maximum Collector Current |Ic max|: 12 A

Max. Operating Junction Temperature (Tj): 200 C

Transition Frequency (ft): 60 MHz

Forward Current Transfer Ratio (hFE), MIN: 10

Heathkit cross references:

My old scope pages, and boat anchor manuals. (2 links)

You can RTM,  read for fun?, some free boat anchor scope manuals here!

Here is my scope fix page.( if you can't afford a scope buy a broken one and fix it, , buy TEK)

I do love this rig, it works so good, I have to pinch myself, the FET Front end is HOT, I can hear tons of DX. (I toggle  3 radios now via antenna coax switch and this rig above, shocks me, its SO GOOD !)
Got my first contact this rig, on 20meters !  whooohoo

version 6  9-29-2016    , any questions or need help, contact me on my EHAM.com , page, under my K5JXH,  name, there. I will respond.  (50+ years as an electronics tech.)