Hickok Tube tester Refurb  . ::::::  by me... K5jxh        
          




I really have no need for any tube testers but, I do have 3 Heathkit HAM  rigs, that use tubes. (I do want to own one of the oldest and newest testers WW2 to 1975?)
 (we can buy most tubes today for near nothing , and less as us boomers go silent key on the ham rig)

I think this one is about year 1945?. (I also have  a B&K747b )
It's not pretty. (Ugly comes to mind) Ugly means low price.
This one is a  model 532: 
 (Do not buy one unless the key parts are good, Meter , transformer and wafer switches, mostly are now  unobtainium )
Below is my day 1 photo's.  Chart out, and the bad hard to find Bias pot removed (now fixed).  (later all that red rust is gone)
I love the reverse embossed lettering,  (factory etched I think it was done there)

See these Selector switches, see the numbers?,  the numbers seen  from grid to Suppressor are pin numbers.
So 6V6 (JR5347-2 tells yoiu grid pin 5, plate 3,screen 4, cathode 7, Suppr. 2)  Knowing this fact one can add tubes not listed, if you have a calibrated tube sample.

The tester is simple in theory, but in practice the wafer wiring is insanely complex. All contacts there must be good and wired right.  (I have  9pin tube failure that makes me pull hair out)
My 4/7/8 octal pin sockets all work great but the 9pin Noval has issues, even testing a Triode.  (3pins, heater ok) This 9p socket was added long ago (see factory service sheets), and makes me suspisious of that fact (done wrong?)

More photo';s of mine here, &  here
The top plate is being treated with 10% Phosphoric acid, this halts rust , removes it and converts what can't be seen to a primer like compound , called iron-phosphate (a barrier to further rust).
Fixes and Mod.'s
1: Bad BIAS pot, cured.

Shock #1, (yes,pun )
 No case metal safety ground, very illegal now. (with metal cases) In fact,  is in  Stark contradiction to safety. 225vdc is not safe to touch ,  for sure at these currents., and the words " halcyon days of tube technology"

(think how well  that old  72 years wire insulation works, parked in hot attic, you trust your life to that wire?) Solution ground the top plate to the  NEC, spec. safety ground, using a 3 wire power plug.

SHOCK#2, is wire colors are not shown in the schematic, (dang me)
Shock #3 Is  there are no transformer pin's ID's shown in the schematic . (there is a transformer page I annotated it.)
My transformer is good and so is the meter.
I now have tested many tubes and it works, and my calibration pot added makes it plenty accurate, (not OCD rated (LOL)  but good enough)

Gm  or Transconductance,  I will call it uMho still, after all , our radios all have USA TUBES and has a matching RCA tube manual and others, all stated in Mho's. ( micro mho's not Siemens)
If you could go back to say 1960, and bought 1 type (same p/n) Tube from the top 10 makers then, how much would the Gm vary?  Answer (+/-)40%  that was the spec.  THEN. (per page  64  by R.Tomer)
These testers are a blunt instrument, it is not a labortory class intrument, not at all, nor is it the makers intent. (It was a simple tool to service an old TV or old radio)
This tool does not run a full transconductance scan at all, in fact , it picks one point on the spec, Gm chart and checks it there, crudely and just one spot.
It never tests the Tube, even at both ends and the middle (3 point testing is far better , no?) 
The books on these intruments never even tell you where they got there Gm spec from, at all.  The RCA book is not a full spec, its just what it might be, on average, the RCA book (or GE) does not even spec, out the full bath tube curve data
on performance.
In one of my better testers (B&K) it does not tell you what 100% means, nor on all 2500 tubes it can test. (using hi/lo plate and bias modes)
The data sheet were called then, a BOGEY sheet and they even sold BOGEY tubes (aka, standards) to match them. (some poor sod had to test 100's of tubes to find a bogey match) {In USN a bogey is a transgressing aircraft}
R.Tomer quote. Pg.64
The tube tester (any) makers offer no guarantee to even attempt to land on the magic BOGEY value. say 2600 umho at 250 Vp. 
At the least , it will fail there, due to using the wrong Vp voltage, of 195vdc. or the wrong load on the plate, preventing 15ma from ever happening,  (the load actual + Rp is total load)
In most cases the tester attempts to check Gm at say 200vdc, on the plate, then use the average seen on the datasheet for that voltage and then show 100% reading for that. or Bogey = 100%



The GOOD Features:
The best feature of any tester or Hickok is the Dynamic mutual conductance test,  Gm is now called Transconductance,  and in an amplifier, is the  small change of grid voltage resulting a large change in plate current.) 
A Triode, (say) is tested as an amplifier, and tests the gain factor.  (cheap grid emission testers, do not)
Not all testers use the same plate voltages or plate loads, nor same grid test signals, or load, in fact some testers use way too much grid signal (overdriving it), and can even damage high gain tubes. 
Due to those facts, no 2 tube testers will show the same results for Gm   , read the data sheet on 12AU7 and see Gm 100v (plate)  is 3200 but at 250v is only 2200, so that old tube tester with only 100v for the plate can read more than mine.
Most if not all tube testers test only at one point in the family of curves, note the Gm curve is not linear, note too that some tubes have radical curves, or remote cut off or other odd qualities.
That means if testing at a radial point in the Gm curve any tester inaccuracy will have  a profound Gm error reported.   (the stimulace errors are mostly very bad on all testers, both Grid bias and signal and plate voltage.
If you just look at the testers schematic for 1 second what do you see missing (BIAS regulators of any kind or so crude as to be pathetic in todays context)

Here is  a 6SN7.

The RCA book (RC-29) shows, 3000 umhos for 90v Vp and  2600 umho at 250 Vp. 
In Europe the ECC32, is near OUR 6SN7, and they mostly use Siemens for Gm. (or like below) : MAX.Vp is 450vdc.  Max. Ip is 20mA.  This tube is Medium Mu twin triode , for more gain see? 6AN5 at Gm= 8000uMho's
All my respects to the man John Harper and his page he made here. and the great 3D plot.
Now look at our book by GE , back in October 1954.
If you look at this great graph,  you can clearly see, that Gm is not a constant or some simple number we can grab hold of  or memorize.
To read OLD data sheets, teh Arcane nomenclature comes into play.
Arcane element codes: Eb (E= volts for EMF electromotive force, the "b or B" means the main high voltage Battery, as used before we had electric lights in our homes 1906 era!) (Mr, De Forest patents all had those markings)
In the  dark ages, we had the "A" and "B" battery , In say in the year 1908 we went to the store and bought huge batteries, for sure big "A"(more often) yes, filaments can suck a battery dry super fast.
"A" was the  filament battery and "B" was plate (anode,) in modern times we'd call the plate Vp. volts plate.  easy, no messy funny words like Electromotive or Battery codes "B"  I think "C" code was fixed grid bias battery.
Oh yes, K= Cathode because so back then Ec1,  (the "c" meant control grid and not cathode) omg, what a mess then. So the solution was to spell Cathode wrong. ( But De Forset and Edison were the 2 Genius guys then)
(Ef below means Filaments are at 6.3vac spec. & Eb means plate voltage  Ec means voltage control grid, back in the dark ages.) (or Ecc1 grid, and Ecc2, screen grid.)
Look on the right side, see how Gm varies by all factors.  (where your tube tester lands, below is anyones guess, mine does 250vdc Eb (lower top Eb curve))
The load on the tube also dictacts Gm readings , and Ib plate current flows. (both in the tester and in the application)
Now put this chart on a wall, and throw a dart at it,  where it lands, is anyones guess. (6SN7) Not only do I not know what tester you have , nor how your application operates, making test data very hard to determine.,




Applications (a near infinite topic)
Keep in mind 1 tube can be used in vast numbers of applications, (amps , of many kind (audio to Radio frequences), and class of amp. A, B, C, oscillators, mixers, clampers, frequency multipliers  and more,  100 years of inventions....endless)
We see tubes used in CASCODE (not cascade) that fail for lacking grid current, (zero) oop's. (R.Tomer)
That means the tube tester can lie to you , if the application you are using , operates this tube at higher currents  or voltages,  the tube can fail in the gear but not no in the  tester,
No lie these testers can find dead/weak tubes, or like my 6146 RF transmitter tube, here,  reads 1200 and should be 7000 Gm. Fails in the transmitter hard. (has paralle pair )
(To see lots of failure causes,  read the book by, Robert Tomer 1960, "Getting the most out of vacuum tubes",  chapter 1, 2, and 3.  (fails) A free book do not pay scalpers for it.

u = Voltage amplification.  (V input over V out) a.ka. Gain or Voltage gain.  (  plate volts/ grid volts = u)
Gm - Transconductance. (in Siemens) or stated as  mA/V.   (means plate mA /  grid signal Volts)
Rp  = Plate resistance ( that tube acting like dynamic varible resistor , not much different from a transistors gain)   and  Rp = u / Gm
 
and  gm = u /  rp

Mutual conductance today is called TRANSCONDUCTANCE.
The tube tester does a great job of checking  say Gm between 2 , of same tube number tubes. 
You can in fact buy tubes that are tested on a very accurate tube tester same as yours (on fleabay) and use that as your guide. (Hart Calibrated it's called)
A tube tester like this, does do a great job of finding dead tubes, shorts, and maybe gas. ( and for sure, tests for  gain)  {please do not obsess (OCD) over TUBE under test (TUT) Gain it's not critical in most cases}
As tomer says, all tubes have gas, he is right, there is no perfect vacuum even on mars. (the moon is pretty close) but your application might exploit your gas, and fail. (higher voltages can do that, even higher than my 100v tester can muster)
Do know that the tester in most cases never runs 600vdc on the plate. (like maybe your Application does? as seen in high power RF transmitters or on high power audio amps)
The only sure fire way to check tubes is in the Applications circuit and with a oscilloscope. (good schematics (like Sams Photo facts cover this , or you doing a tube modeling program to calculate gains , then use your scope. )
One guy in the shop sees gain bad on tube X and finds the cathode bypass cap is dead open or weak.. after changing tubes 3 times.
OCD: (obsessive compulsive discorder)  In the context here, thinking  to get a tube with max Gm and say the tube amplifier does not even care or ever use such voltages or plate currents ever. So is just  waste of effort.
One can use a tube tester to lower crossover distortion with matched PA tubes. (final Power amp output tubes).

Realities?
Lets say you found  a low Gm reading, on one tube ,does that mean gain is so low and  that your radio will fail?,  most times not, in fact many radios circuits are designed to not need much gain at all.  
Radios, designed by a smart RF radio engineer  to run with many gain factors, so the radio runs a longer life span with no service at all.  (+10,000 hours is not uncommon)  { some records show 200,000 hours at  Radio station}
Say that Radio is a 1960 AM radio, and your favorite station is 10 miles from you,  so ?  the tubes can be near dead and work great.  (Hell my FoxHole crystal radio works here and with a $1 transistor drives a speaker too.)
In General:

If the tested tube reads kinda ok?, it's probably ok.
Just because one tester dial reads "umho's" does not mean it's any better at all.  My B&K 747 , reads in percentage of uMho's and is just as accurate as my Hickok. (In fact match so close, it's amazing.)
Warning some tube testers can damage tubes with when doing shorts tests. (1volt tubes , with tight clearances deep inside?)
Do not let your tester use huge amounts of grid current ,or  the grid can be damaged easy, or damage the cathode.

end theory class....
One  up grade, is this  from the  maker,   the charts are available even  for the 532.  CA-4  and newer testers.
For Compactron's (3 tubes in one glass bulb!) I have  6AS11 radio. But refuse to buy this ,just for one tube. (grin)  One guy made 1 tube Compactron radio using a 12AE10. (thanks DAVID !)
So expensive now, I bought a 2nd tube tester that does these tubes.

Inspections and tests 101:  (I use electronic Jargon, after all it is Electronic here, see those TUBES?)

The show stoppers would be: 
  1. Dead meter. (or sticks)  {a sticking meter may be cured, if you look inside it, and find iron dust (or filings) inside, (the magnet loves to find ferris dust, and jam the coil) use tweezers to remove said dust}
  2. A Dead transformer (burned up or open / shorted) {not cheap to rewind one}
  3. Broken any of the 9 rotary wafer selector switches (Phenolic resin wafers damaged), near impossible to find matching switches today.  ( buy tester that has the largest number of sockets and less switches is a best idea)
  4. Mine has a Dead burned up BIAS pot (3k ohms 1w) I got a 4watt for $10 , (original is not sold on earth) look below for my reversed engineered data on this pot. 
  5. Last and very hard to detect, is some dweeb, wired up all those wafers switches wrong, oops... that be hard to fix. (buy 2 testers and make one good one?) (the wire colors are not documented at all)
My BIAS POT Story,
Power and POT.'s;
The first testers had a flaw, no protection from shorted tubes, to protect  my bias pot.
Then in 1951, they wrote and service bulletin to correct this and added #49 lamp fixes this flaw. (so could a 100mA fuse)  The place it goes , is on the center pot lug, in series with it. Mouser has the lamp and holder.
I found a generic new POT made by MALLORY (NOS)
The BIAS pot I'm now using is rated 4watts, but that is only when the dial is at 3000 ohms.  (here is a more a much better pot ,note how easy air can flow here...)
Note too that the old  pot casing, below (stinks) acts as a cup and holds heat here! and is crummy plastic , and stinks super bad as it burns.  nasty it is. Mines now GONE.
One more pot example , there are much better pots .
There are 2 kinds pots, carbon  , Cermet and WW, wire wound, as clearly seen below, carbon POTS are 1/4w  mostly and totally  useless here (burn up it will).  (but great or say a radio tone/volume control)

The stock pot was 1watt and pathetic, I'm later adding with the factory lamp mod. , a solid state fuse here,  (more later but...Bourns, makes these  ) Specifically this. at 72 cents!
I am told today, that the T-105 service bulletin , (missing this revision in my rig here) adds a #49 lamp (2v@60mA) to the R7 BIAS pot. So I  did it that way. (the above may work faster and better) The lamp is a fuse.
I'm attempting a 2 way protector the lamp and the Silcone fuse, is in series. (one slow other fast)

SECRET #1:
Lots of secrets,  here the R6/R7 act as a dummyload (and test bias) in the tester for the 5Y3 and my new diodes, effective kills the tubes on Vf (forward resistance, the 50vdc drop is gone)
My tube mode (solid state =SS) raises the 5y3 from 200 to 250vdc, upsetting the calibration greatly.
I also discovered a secret, a calibration copper tab on the 100 % end of the POT, CW (clockwise max) on mine top rear pot terminal.
This tab, I measured it's effect and can change the top end by 500 ohms. see right side of photo below for this tab, it shorts out the upper turns of wire inside  the pot.
This 500ohms is like 200 Gm delta (change) Not much there. (think of it as fine tuning)
Experiment 1:
I added a 500 ohm 5W pot  in series with the  pot below. hidden inside. on bottom 100% end, pot wire.  (3watts or more spec, any percent is ok)  Mouser.com has it, in stock $5.
The story goes:
I then discovered my errors here, I found out that my solid state diode, for 5Y3 (btw, both my tubes on the under side were dead)  puts out way too much voltage (and current) making R6 value wrong, dead wrong.
I did expect this at the get go, I've done many tube to solid state conversions and more in the US Navy.  The stock tube Vf is 50v, my diode is 0.7v Vf (voltage forward spec) yes dramatic.
In goes a new R6 series Pot,  (500ohm pulled and pitched to the side , all this testing  was down with jumper clips, in like 10 minute work...btw)

So using my Calibration tube 6K6 and    and  junk box POT 10k ohms I discovered 8.6K (plus R6 , 5.8K measured) is just right for BIAS (Knob=34) with my CAL tube 6K6 made with HART ,calibrated 6000a. 1500 Gm, on mine.
This is no true Bogey Tube. Nor the ideal  seen  in the RCA book.
 
R6 needs to be 14.4K ohms,  or padded in series to get there, but what wattage? seems smart to use less than 4 watts (or 1watt stock R7) to have it be sacrifical over the pot blowing up, but the #49 lamp factory fix should cover this.
Keep in the 6K R6 and adding a 10k pot off the end of it gives me a 6k to 16kohm range. perfect for many new 250vdc power supply inside, the new trim pot R6a is mounted on a home made "L" bracket.
 
I went to Ohmite and found out our resistor is 12watts.
This is so wrong, the best resistor for R6 is 6000 ohms 1 watt, not 12watt , so that R6 burns up first not that very expensive R7,  (and now impossible to buy, From Hickok or anyone on earth it seems)
Lets call new pot R6a,  so R6 and R6a are wired in series on the way to the R7 center tab.
The new R6a is 2 watts and  and will burn if before R7.  (or never if the factory mod, works)

I will add a 10k 2watt pot made by Bourns. Mouse shows 49 pots, in these sector.  $5 
The best solution is a pot, for sure, so we can set it or even compromise when say using 3 bench mark tubes, via Hart calibrated gear.
The R6 correction  correct my R7 voltage spans.
R7 end to end was 56vdc wrong. (SS power rectifier mod)
R7 end to end now 30vdc is perfect.  allowing me to set the dail to 34(6K6) and get the correct grid bias. The 0 to 100%? on the dial now picks off a range for 26 volts bias points.

Here is where I began, with this bad pot.
Here is what my stock BIAS POT Device...looks like,   buck naked:   (ie.  0.73" is caliper measurement of how tall the form is on this end)
My new pot is a  linear taper 3000 ohm pot.  (a key device in the gear,  and  is the calibration point... I learned) It turns out , in this modern world , 3k pots are like (chicken lips) Think a diety, for NOS parts.
Man KILLER #2, If you touch that shaft you get shafted. Yah, not cool.
That shaft on the pot runs at  full wiper tap voltage and they put  3 huge rubber washers here, and an insulated knob to keep this voltage off 2 places; YOU and the whole top deck  .
Warning, do not run this tester with the knob missing, or you will be shocked or KILLED ! (as much as 250vdc 100mA. & lethal) (was 200 now 250 due to my new SS tube. (SS= solid state)
The new pot does not have this little horror.

My new $10 bias pot (NOS) is MALLORY #M3MP   (data sheet lost in sands of time...) That does not run a hot shaft , btw. (by the way)  
Below see my charts I made for both pots. (old and new)
I took my dead pot and using ohm meter test lead tips I measured the wire at each 10 points on the dial, and is linear, until that skinny weak tip end.
This chart below is just to show ohms, the volts there are all wrong due to R6 is wrong.


Ignore right column.V

One IDEA : fails.
As you my not know, Gm normally changes wildly with grid bias. ( the more gain the tube has , the greater this effect)
What if  I reduced the 120Hz ripple there,.if one puts a large 300uF cap on the BIAS supply,  I did so and it had no effect, I got 1 more volt out using  cap.  and Gm did not change. 
I guess, as the grid ripple rises,  so does the plate and they cancel?
  

So what is my bias at dial 34 for my calibrated 6K6 (1500 Gm, spec tube actual.
Answer is:   20volts  across the R7 pot works,  35vdc was dead wrong.
from the bottom R7 pin to the center wiper I get 10vdc at Gm 1500. As you can See 34 on the dial is 50% ohms on the pot.

More inspections , cures, modifications, and refurbishments:

The wiring if bad, is an easy fix.  I'd replace mine with Mil grade stranded silver tinned, Teflon wire, as needed.
MOD's:
Upgrade 1  is , new 3 wire line cord with safety ground.  (I standard instrument cable and connector will be used on mine, just like I do on my Heathkits.)
 You won't like  225vdc on the case top, would you.? (yes, a shocking , era then... and very LETHAL)  The rule on this  (mine) enough voltage to shock you and enough current to kill you. (easy no?)
If working live circuits,  set up power unplugged, then connect meters and jumpers to nodes, then with one hand in pocket, plug in the line cord, to power, then keep both hands in pockets and read the meters or scopes.
AND LIVE. (have rubber sole shoes, avoid leather soles, that conduct )
Next up?:
The worn old  tube sockets are easy to buy new, the big pushbutton switches are life time serviceable, just clean them. (I burnish the contacts using strips of cut printer paper, never sandpaper of any grade )
I modified the 2 internal rectifier tubes for solid state plug-ins. (#83 and 5Y3, both sold on ebay, cheap)  (83 is a toxic mercury rectum frier)  A $20 cost or less.

The next mod. would be the classic 2 diodes back to back on the meter, limiting the input to the meter to 0.7vdc protecting it, from worst case shorts to power. (BTW: this meter is 100mV meter, so Schottky diodes be best here)
Best diode is...1N5817
This MOD can  in fact ,save the meter from blowing up. (the meter that can not be bought, BTW) If you linger on fleabay for 1year, you might find one, yes, RARE.


My Restore will be to just get it working.  (no beauty fixes, but will be cleaned OCD)
I have  schematic, a roll chart data ,  and the books on it. Best roll chart is 1961 issue.

Wafer madness, look , ponder, and be happy this era ended:
Leave them alone, is best
I cleaned mine in pure (near) Alcohol  (true name is Ethanol)

Last resort, for proven switch failures that Alcohol scrubs do not cure!:
If the contact silver is real bad I use, Tarn-x to clean the  contacts, but only for seconds and then a fast flush with Alcohol. 
Here is  Tarn-X , MSDS sheet, data.  (I  avoid any product that have secrets ingredients in them. I do know that most cleaners sold in USA all have a MSDS sheet and the word secret is missing ,except the snake oil makers.)
Excerpt: MSDS 2017:
note how the word SECRET is not here.! 
The never do lists, are endless, in today complex world: but...
NEVER DO:
Never use Tarn-x and just walk away, or risk loss of all silver plating, this product is a chem dip for silver, and works with no  scrubbing.
I use as little of this liquid as I can. (more is never better) Do not OCD, this step.
Never soak parts overnight in Tarn-x . (seconds only if at all)
Never use abrasive cleaners of any kind here.
Never use  lacquer thinner, or MEK nor acetone thinner on any phenolic of any kind, that includes, methanol.

" acetone will dissolve most plastics,"(for sure 1913 era plastics, it's not Teflon)"
Never power up the gear with
Alcohol  (true name is Ethanol),  wet parts. Bake it dry,
Never use water to clean gear.  (it's very hard to bake out and can cause damage if it gets deep enough, like inside the transformers or other places...)
Do not put oil or WD40 (its oil) on the these wafers, ever.  Do not buy super expensive SNAKE oil , to put on these wafers. 
The switches came new oil free.
Only the rollers on the bottom are greased. ,  I used BRAKE grease there. (will never melt ) 


The bottom line is do only what is needed to get the switches working and stop. (do not OCD it , as we say today)  The black seen above 100x more scary than truth. (it's mostly ok)
Mine now looks like this: (close up view)
I did not need to use Tarn-x (after a huge alcohol flush and wet brushing)  I do bake it dry, 120F oven or attic dried. (same temp btw)


How those wires can look and be good after 70 years, is  magic to my eyes.

Day1 Tests ?,
Turn it on, if will not glow the power lamp (if lamp not burned out) then the fuse lamp is burned out. (replace, them as needed)
The tests here work good , for 6v6,  then  check the bottom of case tube voltages and there DC (pulsating) outputs.

I test all resistors and switches, and clean all switches. (all my fixed resistors in mine , test good)
Keep in mind they are all +-10 or 20 % spec. accuracy back then , for sure on all wire wound resistors and I'd hope the silver 10% band (CC-carbon Comp)or missing tolerance band is (20% they are) is clear to you... 
Repair all bad wiring, (teflon wire is best )
Repair all bad sockets, (buy new ones with gold pins and ceramic are like $2 now, bingo)
Replace all bad fixed resistors, seen at mouser.com . (look under all wafers, see hidden  47 ohm carbon comps there, sure. test them , in circuit.
Pray the English DIAL POT , never fails. (ganged pots are super super rare now)

Reading a cool web site called the Tone-lizard, I passed all his tests but  failed hard the bias pot test.
My pot worked for like 1 day and failed, and the tester passed my 6V6 tests.
My rig turns on and my line set test works.
My meter works.
The bias pot is dead. (using a wrong pot as a test only  , I discovered my 6V6 and 6L6 tests work, ok, and all my receiving tubes in my Heathkit radio's work (spares I have) until it went 100% dead/
Got a new generic 3k pot with 4 times the watts rating.
I added  BIAS meter so my 3k-4watt, linear pot, (see classic tapers a here, the USA tapers are C1 C2, etc, and is a Hickok SECRET what they used)
I added the Hickok factory mod (1951) that puts a $49 lamp on the center BIAS lug to prevent Pot burn out with shorted tubes.
This tester transformer adds 5vac RMS signal to the DC bias (7volts peak that is and is a sign wave on top of pulsating DC,  there is no name for that mess,,,, but will be in SYNC as will be the plate voltage, pulsing)
I added a CAL POT (500ohms)  ,to mimic the factory copper trim tab seen above. (it is near useless)

More secrets:
The main Gm meter is full scale at  0.114vdc(my testdata) (100mV that is) (I installed 2 back to back 1N5817 diodes across that meter for protection. as seen from 1970s to now , on meters like this)
This mod at least assures 250vdc never ever lands across those 2 meter pins.  (2500 times too much is BOOM let smoke out , failure mode)
I tested my meter  un plugged.  so FS is 100mV. (so the magnet is 70 years old and still like day 1)
The BIAS  Pot wires are .002 inch diameter, or 44 AWG. (hair sizes.)
My new  linear pot works great, as does the calibration secret  (my new pot  is 4 watts, 400% stronger)
I have lots of secrets on the transformer, seen below.  (turns ratio, amps, and volts) Amps is not actual, it's max allowed, by by any output , no others used. Its a max per winding, exclusive of others.
My wall power is 123vac, not 115vac like back in 1943. But the line level trim on mine works like day 1.
 
Calibration:

The best document I've seen as  guidelines is this, by Mr. Shoo.(539c) use it as a baseline guide to doing yours.

I use a calibrated TUBE from a seller using my tester and sends a tube with Gm readings (sticker) he measured,  and then I adjust my new Secret pot.

That Transformer:
The only page I have is below , that I horsed up., and the pins letters can be moved, if the maker wanted, but was told to mark them. (mine is not marked)
The turns ratio dictates output voltage. (basic transformer theory 101)
The input will not be 120ac, due to the knob marked, LINE ADJUST.  (different tubes suck more current, and this is the corrective factor (knob).  (all this is here, is a poor man's VARIAC)
The unmarked pins can be discovered by the adventuresome tech, by using a voltmeter. (do not get shocked on pins with high voltages  230v/70ma can be deadly)
The unmarked pins can be checked with a good ohm meter on the low range, to see which wire is which.   (also find pins not used, or pins open coil easy)
Cases in point,  a reading of 1/2  voltage or 1/2 ohms in a 3 wire group, would be a center tapped coil, like  "N" is below.
The group btw, is AWG American wire gauge.#  what is not know is the length of each coil group. (only turns and gauge is stated) The inches show are pig tails.
or, if checking group 19 , heater secondary wires,  then the ohms goes from low on bottom to higher on top,  easy to do in a 2 ohm range, as seen on most meters made. (modern) 
The 80.5 to its pin 0, heater will go from near zero ohms to 0.25ohms on pin 80.5 (to pin 0)  all pins read less than 1 ohm here in this heater group. I think this group can hump out  1 full amp.
If your transformer is very different then step one is what? FINDing the groups first is the only answer, then testing each group. (do note some tabs are not wired)
Lets say one pin has  weak volts, there are now  1 to 3 tests to do , to prove the transformer ok or to prove the chassis wiring has a short. (always blame chassis first, as below is not sold today)
Volts test wired, fails? then ohms test, then volts unwired, and even a load tests not wired..
  1. Unsolder the pin, with power off, and use and ohm meter on this group of pins, lets pretend now, that pin K   (check L to K, and K to J ( 0.2 & 0.4 ohms),  24AWG it is. 26ohms per 1k feet.  (K is 7feet of wire 17 turns)
  2. Power up, Unsoldered "K" still does pin K read 3.7v, then transformer ok , 0v now is bad transformer, but even with 3.7vac we can have a current supply problem.
  3. Now the acid test, a dummy load test. Just like testing a Ham transmitter we use a dummy load.  The hard parts is each group has different amps ratings. (sadly not documented) What I do is use 1/2 the amps rating for load, so lets do the math on that, and say pin K again is in question, 8  ohms at 2 watts or more., and uses that as a load here, using my huge scrap box of used resistors. I have one other trick, I can, load down pin "L>J" and see how much load it can take with test resistors. and not drop below 10% voltage, if it can handle  1amp load use that for a load on all those grouped pins.  The 228v pin is my guess 70 mA max. I'd shoot for 50mA or 4700 ohms  15watt resistor or more.
  4.  BTW (by the way)if you overload a coil here, the "fuse lamp" glows protecting the transformer from any damage, yes, pretty genius that.. and is common on some shorted tubes tested)

. CLICK PHOTO TO ZOOM IT !  I solved the volts and ohms here, using AWG rules and turns ratio rules.

The maker did not spec, out (means published a full specification on) the coil OHMS. (aka, DC resistance, nor reactance, nor Impedance) So here is my weak attempt. at DCR, Amp.s and volts.

An old song tune: "Ohms Ohms on the range, where the deer and the antelope play"..... ah...
Using my wild guess Coil circumference constant of 4.7 inches (ruler in hand near COIL CORE) C= Pi x D   (or 0.392 feet per turn)
Now I'll guess ohms crudely below,  based on 1.5" diameter avg coil bobbin size, times  pi, = 4.7" (circum = Pi x D) yes, crude as I can muster. (sorry not unwinding my rare working transformer, ever)
So Primary side COIL length is  , 550 X 4.7" (div by 12) = 215' of wire. So,, (0.2k feet times (65ohms per1000 ft)  13 ohms,  (65/.2 = 13) That is direct to primary, not line cord. (my DMM reads 17 ohms (DCR)
Next : (example)
Group 33, I do pin  "V', gauge #33,  is 207ohm/1k/feet , is what, 411ft long so is  0.4k feet  x 207 = 82ohms DCR. all this is ball park numbers... (if had true diameter, I'd be dead on)
I will do one more, using 4.7" coil loop circumference, per turn,  I do coil Q, #31 gauge. is 130ohms/1000ft  so is 31 turns of #31, (oddly)  4.7"x 31 is  146" div, by 12 is 12 feet, long, and is  1.6 ohms DCR, DC, resistance.
These are all crude averages.
With my examples one can test all coils with any cheap ohmeter,  both ohms and volts. (In most cases if the windings read right, not soldered to harness, the transformer is most likely OK.
Some call it the Loom, one can loom around a loom for hours. (In fact, if you wait long enough problems can loom up)
For currents I used the wire guage max current spec, or if the watts exceeded 250w then set the limit there.
The primary coil is limited to 2 amps max.  (or overheats, easy) 2x120 is 240VA. (apparent power that is called)

B&K :
I also own a B&K 747B tube tester, 


Phenolic facts:
Printed circuit boards (PCB), as seen long ago, due to  them being dirt cheap then... (we use epoxy resin PCB's now)

PCBs (then) are in fact made from Phenolic resin, seen as brown, and seen in all my older Heathkit gear, FR4 rated and way cheaper than better green fiber glass epoxy PCB, but both clean up with alcohol for over 50 years.(no problems)
The reason I mention PCB's  at all, is we all know now to clean them and what do use and for sure what not to use... (oil never, ethanol yes)

Early Phenolic radio case. Natural colors are amber, for phenolic resins. (Alkali's is #1 enemy of this plastics as are many aggressive solvents and are toxic)
on the right here is pure phenolic resin.


Topic , on snake oil sold (no names here !) (tubes, bottles or cans of this stuff,  secret stuff) "MAGIC" do you really believe in MAGIC? (why?)


I have a theory? (Using EPA, rules and Occam's razor logic)
List all things it (snake oil) can't be and your answer is before you, (CAN be !)  
1st hint:(and the MSDS line, smells like oil) { that is because it IS}
Why search the list of things it CAN'T BE (Occam ) The Razor part means cut out what it can't be. This can be done .
We know what? it is not toxic,  as the  MSDS list states. (it's not water) In fact the MSDS lists it as a Secret, if you study EPA laws, you learn that its safe, even to drink it. (that tells me it's food)

Note also the MSDS  states a flash point, yes, it burns.  (you heat it near 500f and bam, it fires off and smokes before the flash, I bet.)
They claim it does not conduct electric currents and states it's ok for wafer switches, so is good up to 500vdc as the wafers makers claim?   ((can be transformer oil))
I conclude the secret ingredient in the bottle is mom's cooling oil, Vegetable oil and is why all contents are  stated as non toxic (MSDS), (unlike most petroleums and Dino Oils are...)
My second guess poor(cost more so..and flash point is 360C (680f), is 200f more that above.) if it's transformer oil, but that would be in the MSDS  and is missing. (unless it's new non toxic transformer oil, not cheap, nor is it food grade)
eg1: Natural Ester Envirotemp, FR3 fluid is listed as food grade. (EPA listed as safe)  (based on Soy beans and is modern transformer oil)
The EPA would require this truth be known, if it was DINO Transformer oil or  the above.
If my theory holds, on all facts above, then the secret is just moms cooking oil.  (so ask mom for some, or not?  I say not, do not use oil or grease on wafer switches)
The profit mark up on snake oil is ?: (supermarket oil mom buy's)
Sold as  $20  for  0.25 ounces of snake juice,  times 4 for onces and x32 for quart means  $2500 (rounded) a quart, and  1200 times more that moms $2 cooking oil. A 1200% mark up.  (easy money no?)
Mr. T Barnum Bailey was here. (and I did not bite) ROTFLMAO
So never buy products with the word "secret" in the MSDS sheets, (as a  technician, or engineer, or even a HAM?)
IMO, all this is in my opinion, do your own research and form your own opinions.

Snake oil for persons or your (equipement) give you just a  thin wallet and bad results.
I just heard a savvy guy say on Netflix, (a Doctor of med.), "if it's sold on TV, don't buy it.  "   Now with the internet, times 10,000 that statement.

Best links
http://www.jacmusic.com/Tube-testers/Hickock/750/index2.html

Mr.  Robert Tomer 1960, "Getting the most out of vacuum tubes" (page 59 to page 79  )   165 pages total, short and one of the best books in print on this topic and is free to read, on the internet.
I have great  DVD with 250 books (pdf) , on tubes, but the above my favorite.

I get my calibration (refr) tubes from this Guy. "These tubes tested good on my Hickok 6000a tube tester. The tester was just calibrated by Paul Hart."
This tube "bogey" (old name) is not really a "calibration standard" as we say today, at all, it's only close enough for me.

What is the distribution bell curve like on new tubes.

A great page with correlation  data is here.( I was shocked at the good consistancy )

Best tube data sheet archive? Franks (and mirrors) SUPER GOOD SITE !!!

If you dig deep you can see what B&K uses for Gm model 500.  (so when you see 100%  on there newer tester for 6L6 , they mean 6000 Gm) 100% means BOGEY tube.

Unified B&K Tube chart.


What is a BOGEY tube?,  it's a tube that matches the Datasheet, which is an average of all tubes made by the maker, (who fails to publish the +- deviation, we think it is (+/-) 33% of bogey.   .66 times or 1.33 time the listed data sheet number is the true range of real Gm on all tubes made by them..
Big enough, as we say to drive a truck through  that window.

So, if Gm (datasheet shows) is 2000 at 200Vp,  and grid drive is proper, then the Ip might net you, 1333 Gm on  many tubes in a case of new tubes, but not all. some will be 1.33 time more. at 2660 gm. Now you know why making matched pairs of tubes is so hard, might take you days to do that job.

These Real Bogey tubes are not sold now,  the makers of tubes are not in business now to make tubes. (so can't make you one, can't speak for China cloners)
There are others that try.  (ebay)



version 3.  1-01-2017  (added, bias pot cure, and Gm  facts)