Calibrating Weston 981 type 3 tube tester

[nzoomed]

I've got a weston 981 type 3 tester and have had to give it a going over, thought it was just going to be a simple recap but turned out that it needed those pesky little tubular federal selenium rectifiers replaced and the rectifier inside the meter was also dead.

These are called a zinc iron rectifier from what I've been told and I have no idea where to find such replacements.

Anyway, I've replaced everything with standard 1N4007 diodes.

Now the tester appears to be working normally, but I have no idea how far out the calibration is after installing modern diodes with way less voltage drop.

However, it seems promising after testing some OD3 regulator tubes and getting a near exact measurement of 150 voltage on the meter, im assuming this means it can't be far out?

Either way, my readings from testing various tubes are not alot different than ones tested on my mates tester, but generally slightly higher, but not drastically.

I've found a link to a calibration procedure and looks straightforward enough, but its for a different variant that uses a tube rectifier inside.

If anyone has experience in calibrating these, it would be great.
My main concern is if the silicone diodes will require more modification to the tester, but I think its worth attempting to calibrate it first and go from there.

[Bazz4CQJ]

I don't think the Weston is a very common model in the UK; I could only find two previous threads on this forum.

As you might expect, the Westons are far more common in the US and there's a thread on this US forum on their calibration https://www.antiqueradios.com/forums...c.php?t=158702. I think that it's always the way with test equipment that sticking to the procedures laid down in the manuals is pretty much the way to go.

B

[nzoomed]

Originally Posted by Bazz4CQJ

I don't think the Weston is a very common model in the UK; I could only find two previous threads on this forum.

As you might expect, the Westons are far more common in the US and there's a thread on this US forum on their calibration https://www.antiqueradios.com/forums...c.php?t=158702. I think that it's always the way with test equipment that sticking to the procedures laid down in the manuals is pretty much the way to go.

B

Yes I had been reading that thread. Iirc I think I asked a question some time back on that forum but was not able to get much advice on nodding, but I might ask in that thread what I should do.

I didn't realize they were uncommon outside of the US.

I've come across one of these before here in New Zealand and it was 230V from factory, so they must have exported a few.

[Bazz4CQJ]

I guess that in the UK, Weston would have been competing head-on with AVO for the market in valve testers. Hence, there are lots of people on this forum with various AVO models, but Westons are pretty thin on the ground.

AVO supplied lots of valve testers to the UK military and it's probably the case that many of those eventually ended up with radio hobbyists.

B

[nzoomed]

Originally Posted by Bazz4CQJ

I guess that in the UK, Weston would have been competing head-on with AVO for the market in valve testers. Hence, there are lots of people on this forum with various AVO models, but Westons are pretty thin on the ground.

AVO supplied lots of valve testers to the UK military and it's probably the case that many of those eventually ended up with radio hobbyists.

B

AVO certainly were popular here too, and there's lots of them out there.
I know Weston did have a UK division called Sangamo Weston and they made a lot of kilowatt hour meters found in many homes.

I must say I have not seen many of their valve testers here, but they must have been available.

I'm hoping that there is not too much majorly out on the unit, the readings seem only about 100 umhos higher than tubes I've had tested on another unit.

[Bazz4CQJ]

From what I can recall of umhos, a variation of 100 units is quite a small percentage. The AVO testers measure Gm in mA/V, which means that very many valves just have values between 2-6 on that scale, though the testers will go as high as 60mA/V.

It always seems to me that owning a valve tester is like a hobby in itself!

B

[nzoomed]

Originally Posted by Bazz4CQJ

From what I can recall of umhos, a variation of 100 units is quite a small percentage. The AVO testers measure Gm in mA/V, which means that very many valves just have values between 2-6 on that scale, though the testers will go as high as 60mA/V.

It always seems to me that owning a valve tester is like a hobby in itself!

B

Well thats what I thought too!

I guess the readings could vary between different tube types, but I guess I should have a better idea after recalibrating.

I was expecting it to be way out, but the whole point of having calibration pots inside the unit is to adjust voltages to match a reference voltage anyway, so you would think as long as its able to be adjusted within a certain range, it should be doable.

[Bazz4CQJ]

The only tester that I know well is my own AVO VCM 163. That has quite a precise setting up procedure in the service manual, requiring two internal pre-set pots to be adjusted, and if that is done with the correct readings, then everything is set. The controls on the front panel include a "CAL" check test to be performed, but there's no front panel adjustment if the CAL check does not give a good meter reading.

Going back some decades, AVO would sell you a "standardised" valve, which came with precisely determined Ia and Gm values, and you could check the tester against that. They are no longer available, but I have created my own standard valves, carefully going through the exercise of getting the results to plot Ia as a function of Vg. When the valves are installed in the tester, I'd say the results come out pretty close (within 5%?) which I think is adequate.

Of course, all of these instruments are getting pretty old now, and their maintenance seems to be becoming more of a challenge.

B

[CambridgeWorks]

David Simpson very kindly took the time and trouble to provide some standard valves foc to be passed around members for checking their testers. These would have been of excellent pedigree as David is very serious about his valve testing!
Be nice to see some of them re-offered rather than just kept. Just the cost of postage to New Zealand.
Rob

[nzoomed]

Originally Posted by Bazz4CQJ

From what I can recall of umhos, a variation of 100 units is quite a small percentage. The AVO testers measure Gm in mA/V, which means that very many valves just have values between 2-6 on that scale, though the testers will go as high as 60mA/V.

It always seems to me that owning a valve tester is like a hobby in itself!

B

Originally Posted by Bazz4CQJ

The only tester that I know well is my own AVO VCM 163. That has quite a precise setting up procedure in the service manual, requiring two internal pre-set pots to be adjusted, and if that is done with the correct readings, then everything is set. The controls on the front panel include a "CAL" check test to be performed, but there's no front panel adjustment if the CAL check does not give a good meter reading.

Going back some decades, AVO would sell you a "standardised" valve, which came with precisely determined Ia and Gm values, and you could check the tester against that. They are no longer available, but I have created my own standard valves, carefully going through the exercise of getting the results to plot Ia as a function of Vg. When the valves are installed in the tester, I'd say the results come out pretty close (within 5%?) which I think is adequate.

Of course, all of these instruments are getting pretty old now, and their maintenance seems to be becoming more of a challenge.

B

Yes thats what I believe I need for calibrating my unit.
Basically I need to find a tube that gives the same reading as one of the factory supplied calibration tubes.

This is what could make this old gear much harder to maintain in the years to come.

[nzoomed]

And to add, I've found the following info, am yet to try out, I will let everyone know how I go.
Am interested to hear from anyone else who may have used this procedure:

"Antique Radio Technical Forum
Weston 981-3 Tube tester Calibration

Posted by Chris V on 05/27/2008 18:35

Weston 981 Type 3 Calibration.
I found one of these testers in the trash a year ago, the last of the type 3s and realized that of US production, it's probably second only to the TV-7, and in several respects better. After all, if this is what the legendary Heath TT-1 in it's glueboard carton was a knock-off of, it must be spectacular.

Furthermore, mine has a late edition roll chart which has most of what I need the tester
for. So, I cleaned and painted its rusty case, gave it a new set of caps, and admired it.

One little problem however, I could find no extant calibration procedures, and the Heath is different enough (owing in part to the boost needed for its huge low-sensitivity meter) that its procedure is not very helpful. So I spent some time deconstructing its functions, and here is the procedure:

The first 3 steps are done using the setup for a 6C4 except as noted. Measurements
taken with a DVM or other high impedance VM. The meter must measure AC alone on AC volts, use a 2uF cap in series if it reacts to any DC component on the AC settings.
A true RMS meter should be used if you have dirty mains in the shop, otherwise not necessary.

All readings are taken at the octal socket for convenience. Make sure it has viable caps
and a healthy 3A4 before beginning, no roasted resistors etc.

Configure 981 for 6C4:

With test meter on AC volts, plug it into pins 3 and 4 of the octal socket.
Set filament volts to 10, and adjust front panel LINE CONTROL to obtain exactly 10.0 Volts.
Hold down LINE CHECK switch, and adjust R16 to centre the 981 meter on the red LINE CHECK mark.
Set GRID BIAS to 25L (centre of range). measure DC volts at octal pins 6 (-) and 7 (+). Set R9 to obtain exactly 2.5 volts on the meter.
Set GRID BIAS to 0. Set SIGNAL to 1. Measure AC volts at octal pins 6 and 7. Hold down GM TEST switch, and set R24 to obtain 2.0 volts on the meter.

Configure 981 for 0A2:

Measure DC volts at octal pins 1 (+) and 2 (-). The 200 volt scale of 981 meter should agree with test meter, if not, adjust R4 to correct. Switch the filament voltage through several settings and tweak R4 for best overall agreement between the 2 meters.

Alignment complete - It's that easy.
A few items of note, the bias range is minus 0~20V in H, and 0~5V in L. If there are no
component failures in the tester, very little deviation from the original settings should
be required, these are well designed units.

Pot locations are as follows:

R9 is a panel mount next to T2. From top (meter) to bottom (roll chart) on phenolic strip are R24, R4, R16.

It could be argued that the transformer should be loaded for the line check calibration,
but when a filament causes the voltage to droop, it seems to be accurately reflected in the line check function.

Weston 981-3 Tube tester Calibration
Chris V 05/27/2008 18:35"

[nzoomed]

Well, here is a quick update.
Will be helpful to anyone reading this who has had the same issue calibrating these.
I will go into further detail in another post, but essentially adjusting R24 to get the 2VAC was an issue for me, and has been for other people from posts ive read.
It seems that "true RMS" multimeters dont like the signal at all. I couldnt even get a voltage reading on my digital multimeter that was a "true RMS"
But the person on the thread in the other forum said he could only get 1.3V on his meter.
Turns out I had an Triplett 310-FET analog multimeter and it successfully worked and got an accurate reading.

The 2VAC is from the ocsillator in the tester that feeds the grid of the tube being tested.
Its certainly not a clean waveform at all, the manual says its supposed to be 5000Hz.
Ive attached a photo of what it looks like on the scope. I played around with time divisions and got a bit clearer waveform which I will upload later.
Anyway, anyone who wants to calibrate one of these, I say go and find an analog Multimeter!

Testing the tubes again, and believe it or not with the silicon rectifiers, and its still reading accurate.
In fact, it wasnt even really that far out, as the readings are virtually the same on the tubes I tested prior, and on tubes that i tested on a mates tester a while back.

What I can confirm to anyone with one of these, is that replacing any of the dead rectifiers as far as the tubular selenium ones go and the meter itself, is that its totally OK. This seems to be in harmony with what another member said that they replaced all the rectifiers with no issues.

[Bazz4CQJ]

Sounds interesting; do you have have a manual you could up load, or better still a block diagram. Seems like it feeds 5kHz in at the grid, and then it filters it out at the anode?

The Avo163 feeds in 14kHz at the grid (a clean sine wave) and then takes the amplified signal out of the anode and passes it through a high-pass filter, which takes any 50Hz and other junk out, before amplification and rectification then metering to read Gm.

B

[Dekatron]

Operating manual with diagram and also the alignment procedure by Chris V can be found here: https://bama.edebris.com/manuals/weston/981/

[Bazz4CQJ]

Mmm...quite a lot in common with the 163 and maybe shows that the HPF in the 163 is a bit of over-kill.

B

[nzoomed]

Yup, thats the link I was about to post, the manual explains in quite good detail how the tester operates. Shame that Weston never published any service manual or calibration data for these.
IDK why the signal is so dirty on the oscillator, but at least i was able to get an accurate measurement with my old analog meter.

The good news is that the popular heathkit TT1 tester is an exact copy of the Weston design and all the tube settings are the same for the 981 if anyone is looking for tube data. I suggest anyone who is looking at buying the Heathkit model, to go for the weston instead since they do turn up reguarly on ebay and go for a much lower price. I should also point out that you can test any loctal tube with an 8pin adaptor to plug into the octal socket on the unit.

Some notes on Chris' instructions for calibration, they appear to be for the earlier type3 tester, since there is some variations, for example "Gm Multiplier" is simply called "Signal" on the earlier version, it also has a tube rectifier.
But the main thing of confusion is his instructions for setting to the 6C4 are for an earlier scroll chart that has different settings, for example, the Gm Multiplier was set to 2 and the instructions tell you to turn it to 1, which i already had set my tester to with my scroll chart settings, but I dont think it causes any issue when calibrating. The bias settings were 30H on my chart, but im presuming it was 20H on his scroll chart after watching a youtuber follow his instructions with his tester.

Ive attached a couple more photos of my scope last night after tweaking the scope configuration a bit more. This is as clear as i can get it.

[Dekatron]

Page 6 of this manual describes the circuit and it is quite similar to how the VCM163 works: https://www.rsp-italy.it/Electronics...r%20Manual.pdf

I haven't redrawn the oscillator circuit but it resembled a blocking oscillator in its design.

[nzoomed]

Originally Posted by Dekatron

Page 6 of this manual describes the circuit and it is quite similar to how the VCM163 works: https://www.rsp-italy.it/Electronics...r%20Manual.pdf

I haven't redrawn the oscillator circuit but it resembled a blocking oscillator in its design.

Would that normally create a fairly clean sine wave?
From what I can tell, its not just me.
Either way, seems to be working now. Would still like to compare with anyone elses testers if they can scope it.

[Dekatron]

If it is a blocking oscillator it would normally not produce a sine wave, it produces pulses rather than a continously varying signal like a sinusoidal wave or triangular wave.

There is a good book on blocking oscillators here: https://worldradiohistory.com/BOOKSH...ors-Schure.pdf where they discuss the differences between a blocking oscillator and an Armstrong sinusoidal oscillator a little.

[nzoomed]

Originally Posted by Dekatron

If it is a blocking oscillator it would normally not produce a sine wave, it produces pulses rather than a continously varying signal like a sinusoidal wave or triangular wave.

There is a good book on blocking oscillators here: https://worldradiohistory.com/BOOKSH...ors-Schure.pdf where they discuss the differences between a blocking oscillator and an Armstrong sinusoidal oscillator a little.

OK, well it looks like its probably doing what its designed then, Does it look like that to you on my screenshots of the scope?