Screen voltage check on AVOMK3

[svantel]

Hi, this is my first tread here
I have bought an AVO MK3 that I try to calibrate, Have fixed the meter, some extra magnets to get 30mA for full gauge, and that 22,2 and 27 mA setpoints is fine, added diode protection and damper capacitor as proposed in this forum. Anod voltage a bit high right now, Filament voltage where it should be. Now I am trying to measure the Screen voltage. On page 11 in the calibration paper, it says that you should shorten the V1 Anod to cathode pin (pin 2 and 5), however on the top side connection Anod (a1) and cathod I would mark as pin 6 and 1. Is it in one normal valve holder I should do the shortage?

Have checked most of the component’s, adjusted the 200 ohm and 750-ohm resistor to within 0,5%. But kept the old copper? Rectifier is it normal that it’s needed to change that to silicon diode (2 for the create voltage drop I read somewhere?). the tube rectifier gives excellent value in my Hickok 600A, so they seem fine.
Also tried to measure anode current, by measuring voltage drop over 200-ohm resistor, and Ac voltage drop and ohms law divided by two corresponds almost exact to meter reding, is this a valid way to measure current calibration?

I believe screen voltage is wrong on high settings, since tubes using a few negative voltage on screen corresponds to Hickok readings, but for example an 300B with -97 screen voltage give very low current (below 10ma ), have to go down to around -75 to -80v on the scale to get 80 mA.
Hopfyll for som tips here
Br Svante / sweden /Stockholm

[svantel]

Hi.

After reading another post here, I realize it's the V1 D77 rectifier valve that is to be shorted, does someone know if it's the left or right valve (it's not easy to follow the wiring to see which one is connected to screen circuit without dismantling again).

Br Svante

[Dekatron]

I did check the service manual on page 11 and it is another of the "famous" typos in the AVO documentation, where it says V1(a) it should read V2 or V2(a) as V2 is the valve used for half-wave rectifying the Screen-Voltage, see the schematic diagram. Since both halves of V2 are in parallel you just need to short one of them.

Then in your last comment on the 300B I think you mean Grid-Voltage and not Screen-Voltage, correct?

Anyway there are many discussions on testing 300B's on AVO testers and I think that the test data in the AVO VCMs are incorrect when it comes to the Grid-Voltage value. I suggest that you search the forum and the Internet on discussions about testing the 300B.

[svantel]

Hi Martin

Thanks for noticing the typo, and now looking at the schematic, it's obvious that it is V2, you don’t happen to know if that’s the tube close to left hatches?
Now I can continue to try to make the meter work at least close correct.

And of course I meant grid voltage.

Since I noticed you are an expert on these equipments, I hope you have time for the second question. Does it work to calculate mA by measuring AC voltage over 200 ohm shunt resistor (calculate and divide by two), or is there some more correct way to direct measure the current with an modern DMM (true RMS type >=1000hz)


Br Svante

[svantel]

HI again

One more question, now I know how to measure screen voltage, but for the grid voltage, its state that you should us a DC valve volt meter to measure that, can I ad an “shorting” resistor to an DMM to get the correct measurement (I don’t have a valve volt meter), or can it be compensated by some calculation?

Br Svante

[Dekatron]

Please start by reading the document about the comparisons of AVO valve tester circuits here, if you haven't done so already (in Appendix 1 you'll find information what instrument to use and on what setting): AVO VCMs a comparison of their designs - huge document split into chapters..

In general you can use a modern True RMS DMM, like the Fluke 87V", instead of the Valve Voltmeter in all measurements but beware of older "True RMS DMM's" like the Beckman RMS3030, as they only show True RMS voltages for pure sinusoidal wave-forms and not for other wave-forms like half wave rectified sinusoidal voltages (like those used for some voltages in the AVO VCM's) and neither for square- or triangular wave-forms - some do have compensation calculations for some wave-forms but they are in my opinion not usable and some have big errors in their measurements.

Looking from the back of the AVO VCM Mk 3 V2 is on the left, the one closest to the three green power resistors. You can see that it is so in the included photos as the two diodes are connected together via the white wires, whereas the grid volts diode has one connection going via the grey wire to the SET mA/V link. The easiest way to short V2 is to short the two yellow wires on the top of the SRBP as those two posts connect directly to the two rectifier halves - just look at the photos and you'll see it clearly.

If you are looking to measure the anode current just open the anode link and connect your True RMS DMM there set to mA DC and double that figure to get the correct value - just as described in the service manual when checking the tester, described in paragraph (h) on page 6 of the service manual.

Likewise, in paragraph (d) on page 13 you can use your True RMS DMM set on the DC Volts range to measure the grid voltage.

[svantel]

Hi Martin

Thanks very much again!
Yes I have read your comparison, but doesn’t catch 100% of everything you write (I’m an mechanical engineer, so I can maybe blame my low understanding on that).
I uses an APPA 98III DMM, that I think is “real “RMS, and an Keithley 2015 THD that I can use for this measurement.

When I follow the instruction on page 13, first setting correct anode voltage, and then measure grid voltage, I get the correct (almost) voltage if using AC-volt setting om the DMM, but the text state DC valve voltmeter, and switching to DC measurement I get 105 volt instead of the stated 52,5 volte, is it an typo ther also? The double value follows al different selector setting for grid voltage.

I now se that this measurement should be done with link open, I read that as “no tube” in the circuit, but it maybe is referring to the small metal link close to V2 on the board?

When I brake the anode line and measure current, its like 50% over the meter reading (if I remember right), so there must be some problem then that I have to address

Thanks again

Br Svante

[Dekatron]

The Keithley 2015 THD seems to be the better meter of the two when it comes to measuring half wave rectified sinusoidal waveforms, better accuracy from what I can see in the manuals.

Did you open the grid circuit link when using the DC range on your DMMs? You should get the stated voltage with the link open, and yes it is the link close to V2 that should be opened.

I don't quite understand what you mean by the last sentence, what measurement are you doing then?

[svantel]

Hi again

I have not measure grid voltage with link open, didn’t realized that it was the link before I saw you photo on the power supply board, so I will do that in the weekend

Last sentence, that test was with a triode connected by cables to the AVO meter, and a ampere meter in the connection to the tube anode (so not in the link on the board)

Br Svante

[svantel]

Hi again

with excellent help from Martin, I have now checked Filament, anode and screen voltage, and they are well within the allowed variation stated. I have also calibrated screen voltage, need to ad a 47k resistor over RV5 potentiometer, to get correct reading at both end of the turning of the potentiometer.

However, I can’t measure the anode current as stated, when putting an amp meter in the open anode cirque link as described on page 13, sentences b (under sentences d the link is called grid link, but I presume it’s the same link?) I get an fixed value, around 5 mADC (around8 mA AC), I can adjust it with RV4 to another static value. But measuring on a tube (300B right now), that readings stays the same (around 5 mA) whether I set grid voltage so I get 20 or 50 mA on the AVO meter. What wrong I’m I doing?

(I have the V1 rectifier valve in, I presume that’s correct)

The R19 resistor couple is fine, and the AVO meter resistance is 3,26 kohm (measured by the way described in this forum, by a 10 turn parallel pot and turning to 50% of FS)

Grateful for some more help

Br Svante

[Dekatron]

The Anode Links are the two found on the valve holder panel on top behind the roller selector, to the left and right marked, they can be seen in this photo: http://radiosyd.se/pictures/avomkiii-13.jpg and have black thumb screws to be easily opened. There is one for Anode-1 (A1 Link) to the left and one for Anode-2 (A2 Link) to the right.

The procedure described in paragraph "(d)" on page 13 of the service manual is for the grid volts control and refers to the grid link which is found inside the tester whereas paragraph "(b)" refers to the two Anode links found on the valve panel.

[svantel]

Hi once again

I wonder how I could have miss that, it’s even written on the side A1 link, anyway thanks!
The current reading is of with some 4,5%, but since AVO state 5% is ok, I didn’t try any corrections.

The only still present issue I have to calibrate is the mA/V potentiometer calibration (paragraph C on page 13 and 14).
Setting the volt grid pot to 105 mv in 2,5 mA position, and then after switching to mA/v position and adjust to 0v with the mA/V control in position 5 worked fine (had to move the scale from position 4)
Then again setting grid voltage to 0,52V (in 2,5 mA position), and mA/A at position 1 also reduce to 0 v again.

However when I set the grid voltage (middle sentence under c on page 14) to 52,5v and control to 10. It stays the same high voltage as before when turning the switch to mA/V, is that the way its supposed to be? (It’s unclear written for me)
I can’t for sure back it down to 0v again, regardless where I set the mA/V knob on both inner/outer scale , the voltage stays close to 52,5 V.

I thought it might be some problem with the cupper rectifier, so I added a silicon diode in the line from the rectifier. But it did not change anything.


Br Svante

[Dekatron]

It has been a very long time since I last calibrated a Mk III but I quite don't understand your explanation, can you add some photos of what you are doing and where it doesn't work?

Do you need to change anything or is it enough to check the voltages from the Grid Volts potentiometer that they are correct?

[svantel]

Hi Martin

On page 13/14 (se World fil for that part)

I don't follow the middle sentence:
The above procedure should be repeated with grid voltage adjusted to 52,5V and the mA/V control set to 10.

For 0,105 V and 0,52V, I can get the reading to fall back to 0v when turning the switch from 2,5 mA seting to test of mA/V (conductance) on 5 and 1, as supposed.

But what about the sentence when grid voltage is set to 52,5V?
What's supposed to happened after that setting?

Br Svante

[Dekatron]

Now I see another typo in the service manual!

It should read 52.5mV where it says 52.5V!

If you check my circuit comparison document in the chapter about the mA/V control you can see a chart on page 25 where values are listed for the mA/V setting as well as the voltage you should be able to measure for each position. In short the positions tested are (with expected function of the mA/V control in parenthesis to get the fudge factor 0.52/0.525 correct):

1mA/V <-> 520/525mV (1*520/525mV = 520/525mV)
5mA/V <-> 104/105mV (5*104/105mV = 520/525mV)
10mA/V <-> 52/52.5mV (10*52/52.5mV = 520/525mV)

[svantel]

Hi Martin

Then I have the meter fully calibrated, within spec. Anode current shows some 4,5% lower than actual, but no mayor problem, especially when I know it.

I had to remove the silicon diode that I hade in line with the cupper rectifier, else the back-of didn't work properly.

I still get very low anode current on 300b valves, when using -97v on the grid (8mA), but if i go down to -75, that seems to be a more normal value at 400 v anode voltage, i get the expected around 40mA, so maybe its an typo in the handbook there also? Other valves, like 45 shows normal current values when setting up according to the manual.

I'm not sure that I will keep the meter, its quit slow to work with, the back-of drifts, so it seems that I have to do the back-of maybe 3 times before i can trust the conductance value (maybe the AVO needs to stay on for longer time in order to get stable due to temperature drifting), anyway my Hickok 600A is much faster to work with, but of course it cant test at 400V.

But thank you martin, I would not have succeed in the calibration whiteout your help!
Br Svante

[Dekatron]

Thanks!

Could it be that your 300B draws more current on the grid than the tester can supply - maybe the grid voltage drops due to that?

Also, you should wait long enough when checking a valve so that the Anode current has stabilized enough for an actual measurement to be taken, otherwise you will experience that the backing-off has to be adjusted several times. Some valves experience a rising anode current for the first few minutes and then drop off some percent so taking a measurement too early will result in overly optimistic values!

[Dekatron]

One question: What is the serial number of your AVO?

[svantel]

Hi
Yes both anode voltage an especially screen voltage sinks if i push up the anode
current. However using the defined -97 grid voltage (meter reads 105 v or so) giving 8 ma current, responds to an conductance value of 1800.
If I instead push the tube to max (I used an cheaper tub doing this test VS last time), almost 70 mA, the conductance reading shows 3800, which is exactly the same as on my Hickok tester. so i believe its withal to get the current in to normal amplifier using area, in order to get correct conductance?

The tester has SN 1580

Br Svante

[svantel]

by mistake same picture 2 times, her the one with only 8 mA current
/Svante