AVO mk3 suspected low readings - UK Vintage Radio Repair and Restoration Discussion Forum

Richardgr · 3rd Jan 2021, 11:02 pm

Apologies if this is already a thread - I had no success with the search on my phone.

I recently tested a batch of EL34's and some NOS 5B/255M, and the anode current figures were all lower than I expected. It could be the tubes, but I was wondering if something could have got tired in my old tester, which is 60 years out of warranty ;-/

All the tubes tested OK, with gm just on the low side. There is a valve rectifier in the tested - how often are they replaced?

John_BS · 4th Jan 2021, 8:40 pm

Do you know if it's been modified (e.g. to protect the meter?).

As a first step you could run through the basic cal procedure as set out in the service manual.

John

Richardgr · 4th Jan 2021, 11:32 pm

Fairly sure it is stock. Thanks for the suggestion! I do have a uTracer, so I need to compare results on one tube, but neither tester is guaranteed to give a definitive result. I do have some Billingtons gold EL34's which are marked with gm and dissipation from standard characteristics, and it was in testing them that I saw the low dissipation results.

John_BS · 5th Jan 2021, 12:01 am

You've not said how large the differences are: are we talking 10% or more?

A simple check of one parameter (Ia) would be to open up the anode link and insert a current meter, which should read exacly half the nominal anode current. Does the gm come good if you adjust Vg to give the "correct" anode current?

John

Richardgr · 5th Jan 2021, 9:35 am

I think it is around 10%. I have some PL36's, so I can easily try your solution carefully testing the current from the top cap.

John_BS · 5th Jan 2021, 11:11 am

Those metal straps either side of the D-shaped Top Cap socket array can be opened to allow insertion of a mA meter, so you can check Ia for any valve.

John

Richardgr · 23rd Jan 2021, 10:57 pm

Finally got round to testing this. I left the 1M resistor in place when measuring the current since I assumed it was there for a reason.

Current was exactly half the reading of anode current on the tester meter.
I tried to find some information about calibrating the tester, but could not find anything helpful.

I tested a bunch of EL34s and the new ones I have tested lower current using the presets for characteristics testing, than some old pulls from a 70s amp, so it could be the tubes I bought are a bit suspect.

I tested a batch of ECC82s, so now I'll retest them on my uTracer to try and get to the bottom of it.

Another question: One of the ECC82s flashed internally when going to the 'test' position, and the tester made a noise as if a failsafe was triggered, but then the the tested OK. Can their be debri in a tube that causes a glitch, or is this a tube to be discarded?

Dekatron · 24th Jan 2021, 5:56 am

You should download the "Service manual for Mk III", it should be available here: https://elektrotanya.com/avo_vcm-mki.../download.html

Start by checking the calibration as per the instructions. For this you need a calibration valve which you can read in the service manual about how to make. Then continue to check all components, especially the resistors as they might be way out of tolerance after all the years since they were manufactured.

Don't change/replace or modify anything until you know for sure that something is out of calibration in your Mk III as it can be hard to get back to square one if it didn't need any changes.

Since it is a Mk III which normally uses EB91 valves in the circuit you can try a new pair there to see if it affects the measurements (unless it has been modified with silicon diodes). The pair of EB91's wear out just as any valve and sometimes needs to be replaced.

Richardgr · 24th Jan 2021, 11:44 am

Hi Martin, thanks for that. I have some EAA91 in my stash, should the rectifier tubes be tired and need to be replaced.

Thanks for the link. The procedure looks to be straightforward, compiled for a workshop guy with a meter and a screwdriver, so I'll aim to find some time down the line to compile the measurements.

It looks to me like an EL86 would be a good sub for a KT33C - low voltage tube with high current capability - so I'll try and set one of those up as a bogey.

I am going to be very reluctant to dive into this unit since it is totally original, complete with issue 9 of the blue bound test book in its slot underneath. I'm hoping if there is drift from exact values, I can compile some chart for extrapolating the correct values. I have the uTracer, and I am only ever looking to match tubes and highlight those to be binned, or tired, when I am testing.

Paul Galpin · 24th Jan 2021, 2:38 pm

Just to add further confusion --
When you get the book, there may be a bit of the history of the AVO valve testers. AVO did a lot of research, and found that testing the valves with AC worked just as accurately as using variable dc power supplies for the valve electrodes (Except the control grid, check the circuit, it has a diode there)
If the AVO meter says it is passing 100 mA, then another meter in series will show 50mA, because it is actually the positive peaks of ac current. This is in the design.
This saves on bigger transformers, hot rectifiers, large electrolytics, and COST.

The 1M resistor across the link is for testing magic-eyes (tuning indicators, if you prefer!).

Dekatron · 24th Jan 2021, 3:15 pm

To me it seems like you have a hard decision to make - keep it original or refurbish it to get correct readings.

Some of the problems with the old resistors is the drift that they experience when heating up, not just drift due to old age. So making a table with tolerances might not work as the measurements will probably be temperature dependent which in turn depends on how long the valve tester has been in use heating the resistors.

I've seen very poor resistors in these testers that changed a lot just from heating them with your fingers.

If you go the way with a table to adjust measurements you will have to check it at several points after using the tester for some time so that it has had time to heat up due to the measurements, however if you decide to just make quick measurements it might not affect them as much.

Why not leave the AVO Mk III as it is in its original state and just use the uTracer?

Richardgr · 25th Jan 2021, 12:35 am

I have the Sussex tester to be completed as well (started before the AVO was bought), so maybe that and the uTracer solve all scenarios. The uTracer is not so effective at capturing truly defective tubes. However, the AVO is a joy to work with, and has every base I'd ever need, except oddly F8, as used by ECL11.

My next step is to create some tubes with known characteristics so I can get some quantitative evidence, following the procedure in the service manual.

3rd Jan 2021, 11:02 pm	#1
Richardgr Heptode Join Date: Oct 2019 Location: Täby, Sweden Posts: 704	AVO mk3 suspected low readings Apologies if this is already a thread - I had no success with the search on my phone. I recently tested a batch of EL34's and some NOS 5B/255M, and the anode current figures were all lower than I expected. It could be the tubes, but I was wondering if something could have got tired in my old tester, which is 60 years out of warranty ;-/ All the tubes tested OK, with gm just on the low side. There is a valve rectifier in the tested - how often are they replaced?

4th Jan 2021, 8:40 pm	#2
John_BS Octode Join Date: Dec 2008 Location: Wincanton, Somerset, UK. Posts: 1,780	Re: AVO mk3 suspected low readings Do you know if it's been modified (e.g. to protect the meter?). As a first step you could run through the basic cal procedure as set out in the service manual. John

4th Jan 2021, 11:32 pm	#3
Richardgr Heptode Join Date: Oct 2019 Location: Täby, Sweden Posts: 704	Re: AVO mk3 suspected low readings Fairly sure it is stock. Thanks for the suggestion! I do have a uTracer, so I need to compare results on one tube, but neither tester is guaranteed to give a definitive result. I do have some Billingtons gold EL34's which are marked with gm and dissipation from standard characteristics, and it was in testing them that I saw the low dissipation results.

5th Jan 2021, 12:01 am	#4
John_BS Octode Join Date: Dec 2008 Location: Wincanton, Somerset, UK. Posts: 1,780	Re: AVO mk3 suspected low readings You've not said how large the differences are: are we talking 10% or more? A simple check of one parameter (Ia) would be to open up the anode link and insert a current meter, which should read exacly half the nominal anode current. Does the gm come good if you adjust Vg to give the "correct" anode current? John

5th Jan 2021, 9:35 am	#5
Richardgr Heptode Join Date: Oct 2019 Location: Täby, Sweden Posts: 704	Re: AVO mk3 suspected low readings I think it is around 10%. I have some PL36's, so I can easily try your solution carefully testing the current from the top cap.

5th Jan 2021, 11:11 am	#6
John_BS Octode Join Date: Dec 2008 Location: Wincanton, Somerset, UK. Posts: 1,780	Re: AVO mk3 suspected low readings Those metal straps either side of the D-shaped Top Cap socket array can be opened to allow insertion of a mA meter, so you can check Ia for any valve. John

23rd Jan 2021, 10:57 pm	#7
Richardgr Heptode Join Date: Oct 2019 Location: Täby, Sweden Posts: 704	Re: AVO mk3 suspected low readings Finally got round to testing this. I left the 1M resistor in place when measuring the current since I assumed it was there for a reason. Current was exactly half the reading of anode current on the tester meter. I tried to find some information about calibrating the tester, but could not find anything helpful. I tested a bunch of EL34s and the new ones I have tested lower current using the presets for characteristics testing, than some old pulls from a 70s amp, so it could be the tubes I bought are a bit suspect. I tested a batch of ECC82s, so now I'll retest them on my uTracer to try and get to the bottom of it. Another question: One of the ECC82s flashed internally when going to the 'test' position, and the tester made a noise as if a failsafe was triggered, but then the the tested OK. Can their be debri in a tube that causes a glitch, or is this a tube to be discarded?

24th Jan 2021, 5:56 am	#8
Dekatron Octode Join Date: May 2009 Location: Linkoping, Sweden Posts: 1,465	Re: AVO mk3 suspected low readings You should download the "Service manual for Mk III", it should be available here: https://elektrotanya.com/avo_vcm-mki.../download.html Start by checking the calibration as per the instructions. For this you need a calibration valve which you can read in the service manual about how to make. Then continue to check all components, especially the resistors as they might be way out of tolerance after all the years since they were manufactured. Don't change/replace or modify anything until you know for sure that something is out of calibration in your Mk III as it can be hard to get back to square one if it didn't need any changes. Since it is a Mk III which normally uses EB91 valves in the circuit you can try a new pair there to see if it affects the measurements (unless it has been modified with silicon diodes). The pair of EB91's wear out just as any valve and sometimes needs to be replaced. __________________ Martin, Sweden

24th Jan 2021, 11:44 am	#9
Richardgr Heptode Join Date: Oct 2019 Location: Täby, Sweden Posts: 704	Re: AVO mk3 suspected low readings Hi Martin, thanks for that. I have some EAA91 in my stash, should the rectifier tubes be tired and need to be replaced. Thanks for the link. The procedure looks to be straightforward, compiled for a workshop guy with a meter and a screwdriver, so I'll aim to find some time down the line to compile the measurements. It looks to me like an EL86 would be a good sub for a KT33C - low voltage tube with high current capability - so I'll try and set one of those up as a bogey. I am going to be very reluctant to dive into this unit since it is totally original, complete with issue 9 of the blue bound test book in its slot underneath. I'm hoping if there is drift from exact values, I can compile some chart for extrapolating the correct values. I have the uTracer, and I am only ever looking to match tubes and highlight those to be binned, or tired, when I am testing.

24th Jan 2021, 2:38 pm	#10
Paul Galpin Triode Join Date: Feb 2009 Location: Port Elizabeth, Eastern Cape, South Africa Posts: 21	Re: AVO mk3 suspected low readings Just to add further confusion -- When you get the book, there may be a bit of the history of the AVO valve testers. AVO did a lot of research, and found that testing the valves with AC worked just as accurately as using variable dc power supplies for the valve electrodes (Except the control grid, check the circuit, it has a diode there) If the AVO meter says it is passing 100 mA, then another meter in series will show 50mA, because it is actually the positive peaks of ac current. This is in the design. This saves on bigger transformers, hot rectifiers, large electrolytics, and COST. The 1M resistor across the link is for testing magic-eyes (tuning indicators, if you prefer!).

24th Jan 2021, 3:15 pm	#11
Dekatron Octode Join Date: May 2009 Location: Linkoping, Sweden Posts: 1,465	Re: AVO mk3 suspected low readings To me it seems like you have a hard decision to make - keep it original or refurbish it to get correct readings. Some of the problems with the old resistors is the drift that they experience when heating up, not just drift due to old age. So making a table with tolerances might not work as the measurements will probably be temperature dependent which in turn depends on how long the valve tester has been in use heating the resistors. I've seen very poor resistors in these testers that changed a lot just from heating them with your fingers. If you go the way with a table to adjust measurements you will have to check it at several points after using the tester for some time so that it has had time to heat up due to the measurements, however if you decide to just make quick measurements it might not affect them as much. Why not leave the AVO Mk III as it is in its original state and just use the uTracer? __________________ Martin, Sweden

25th Jan 2021, 12:35 am	#12
Richardgr Heptode Join Date: Oct 2019 Location: Täby, Sweden Posts: 704	Re: AVO mk3 suspected low readings I have the Sussex tester to be completed as well (started before the AVO was bought), so maybe that and the uTracer solve all scenarios. The uTracer is not so effective at capturing truly defective tubes. However, the AVO is a joy to work with, and has every base I'd ever need, except oddly F8, as used by ECL11. My next step is to create some tubes with known characteristics so I can get some quantitative evidence, following the procedure in the service manual.