Questions on check and calibration of AVO Mk III and IV testers

[TomAW-60]

Hi,
I just registered as a new member and this will be my first posting on this forum. Although trying to read information on the AVO testers found here before posting this, I may of course touch on questions addressed in full previously, thus please have me excused if doing so.

I use a lot of old NOS tubes in my stereo system and in order to know the quality of the valves I put in to my components, I would like to measure them (especially check for shorts!). Over time I've acquired a TV-7, an AVO MK IV and an AVO MK III. The TV-7 seems to work within some precision, the MK IV is clearly off when setting the set-mark, which then brought me to jump on a MK III just recently that I came across. This one seems to give Ia and gm readings somewhere in line with measures written on some of my tubes/boxes acquired some 20 years ago. However, as tube testers even back then was of considerable age, would numbers be correct? I have no set of reference tubes to compare with, and then again, I suppose any reference tubes would carry the calibration accuracy of the tester making them. Then again, you can always just listen to the tube when used in a component for a practical test. But, perhaps being sentimental, I really would like these AVO testers to be properly calibrated, hope to bring them back to their perfect life again and serve their purpose as the lovely instruments they are. Myself being a newbie on electronics, I've reach out to a couple of DIY resources, however with little result when them having a try on it. Which made me feel a bit helpless. I then tried to find some information on calibration of AVO testers on net. Which led me to this forum - where I found real quality information on AVO testers! Which I find really great!
However, with my very limited knowledge of electronic and circuits, I struggle a bit to understand the technical parts of the documentation (and terminology), so method will be to read over and over again...

What perhaps was not so encouraging was reading this section (I've edited a bit):
.." most of the resistors and capacitors used in the AVO Valve Testers had an upper life span of 10-20 years at the most, some even shorter.

During restorations components have been found to differ by more than 100% from their original value (those are easy to spot). The more problematic ones are the ones that measure fine with a DMM or a bridge, but which don’t work well when high voltage or high current or AC voltage is applied to them.

Some resistors in the measuring circuit have been found to change their value substantially when they were heated but they measured just fine with a DMM; some have changed their value when higher voltages have been applied to them. So, if you want to get good measurements, then exchange all components."

Ok, that's a tough message. Reality checks in, and reflecting on it, it do seems to be telling the truth, although the "quick fix" of one or two parts is what one hopes for....perhaps thinking/hoping that status of parts will also be dependent on their extent of use/extent of the use of the tester over the years? (Hmm, 60-70 years of unknown history...) Adding to this, reading, I've come to understand that the AVO meter has to work flawlessly in order to get any correct measurement readings or even be able to calibrate the tester in the first place ...

But for a start: What would be needed of test equipment to follow the set up procedures/tests in the service manuals for an initial check? (Any specific DMM, potmeters, resistors etc? variac?)

And then I'm confused as to where to use a (modern) DMM (RMS as written on my DMM), and when to use a "Mean value" meter -references in service manual are given to an AVO 8 test meter - are these "mean value" test meters available today, or how to go about? (I presume this is somehow related to built and internal resistance of test meters?)

... and adding: I got approx. 240 (+/-) volts out of my wall AC socket, the fuse setting on Mk IV has a slot for this voltage, the Mk III even has a 240/250v connection position - would this high AC voltage represent a challenge to the testers/ the calibration per se?

Thanks for any feedback!

Best,
Tom, Norway

[Dekatron]

Could you please post links and references to the parts that you mention.

[stratfordade]

I’ll take the easy questions and leave the others to the experts!

If you have a primary tap for your mains voltage then select it. That won’t pose any calibration difficulties, on the contrary it ensures that the ‘set’ adjustment is accurate.

Yes the accuracy of the meter itself is crucial, but this can be checked without involving the tester at all, and in conjunction with a DC uA setting on DMM so no issue of AC measurements involving mean, RMS etc. of course be ultra careful making this check, so as not to damage the meter.

[Chrispy57]

Following on from stratfordade cherry-picking from the easy questions, I will offer my expertise and say that yes, Avo Model 8 meters are still widely available on the secondhand market, a Model 9 would be equally effective too.

Cheers
Chris

[TomAW-60]

Thanks for helping!

Status of Mk4 tester (says No 3661, Type 4):
When powered on, tester seems like functioning. However, the pointer will not reach the red set mark, but stops far below the red mark (I have 240V out from wall, fuse set at this position). When checking my Mk4, my DIY friends repositioned the fuse to a lower fuse voltage setting (210-230V), and using the rotary set-switch on the front panel, made the set mark (they told me) - before giving up when trying to understand further measurement results. Reading AVO MK4 documentation here (Martin F’s reverse engineering of the AVOs), I get this is no fruitful way to go - transformers will see a different voltage than indicated by meter (“fooling” meter calibration by adding on voltage to meter by switching fuse to 210-230v in order to get the pointer to the set mark). To my understanding: transformers and meter will now be out of sync, and this will not work.

As such this may then indicate that the sensitivity of the meter is too low - (with a correct setting of fuse position aligned with wall output): So is… initial voltage fed to the meter too low (transformer(s)?), or current fed to the meter too low… or meter resistance is off (meter itself and/or the 10K resistor)?

From the documentation I’ve come to the understanding that the meter needs to work 100% correctly, to measure according to spec (2,453?+10,000? - FSD of 39.75µA), as a precondition for the AVO to be operating correctly and possibly be calibrated correctly. I thus will have to check/measure the meter as a first step.

I have a UNI-T UT161E multimeter( with a lot of strange named positions). So, if to check the meter can I use this DMM (set at Ohm position), or will the test current from the DMM, when measuring the meter resistance, may cause harm to the meter due to the meter’s very high sensitivity? - and if ok, where to find the two correct measuring spots inside the Mk4 for the two DMM test leads (black and red)?

I read about an AVO meter test I can perform without touching the meter: To put a 1 MOhm resistor between the two anode links for A1 and A2 found on the top plate and then run the cold leakage test for positions A1 and A2 - and they should then both indicate 1 MOhm on the meter in both positions. But I presume this is of little value with the set-mark issue described above?

(Currently I have no resistors, potmeters etc, so would need to get some)

Best, Tom

[Dekatron]

Don't use a DMM, or in most cases an analogue voltmeter, set on its Ohms range to measure the internal resistance on one of these 30uA movements as they will bang the needle to the end stop.

Most multimeters use much higher current for resistance measurements than the 30uA the moving coil movement on the AVO Mk III, IV & CT160 VCMs.

I just checked my Fluke 87V and it uses 100uA when measuring a 3250 Ohm resistor which would be over 3 times as much as the 30uA. When measuring above approximately 5.25 kOhm the Fluke 87V switches to 10uA (there is a hysteresis span between 5.25 kOhm up to 6.62 kOhm where the Fluke 87V switches between 100uA and 10uA back and forth). If you are unlucky the multimeter will use an even higher 1mA range for the resistance measurement and bang the needle way harder.

I recommend that you either use a proper constant current generator and set it to 30uA and see where the needle ends up or that you build a design like the one I posted here some years back (can't find the link now though).

The circuit I posted looks something like this: just use a 1.5V dry cell in series with a DMM on the uA range then in series with a 47 kOhm resistor and a 5 kOhm potentiometer and then in series with the AVO moving coil meter. You will then get approximately 29.85uA through the AVO moving coil meter when the potentiometer is as zero and 27.15uA when the potentiometer is at 5k. Before hooking up the AVO moving coil meter you can use a 3250 Ohm resistor in its place to check the current flowing in the circuit so it works correctly. If you want to pass the 30uA mark on your AVO moving coil meter you will have to put a 330 kOhm resistor in parallel with the 47 kOhm resistor which will give you a maximum of some 33.8uA which won't bang the needle to hard. In my original circuit I used a switch and a potentiometer to switch it into the circuit to be able to check the internal resistance of the AVO moving coil meter, this was done by first setting the first potentiometer so the meter showed 30uA, then switch in the second potentiometer and set it so that the AVO moving coil meter showed 15uA. By measuring the resistance of the second potentiometer with the switch off you will then get the internal resistance of the AVO moving coil meter - this comes from the AVO moving coil meter and the second potentiometer working as a current splitter letting 15uA through each branch, half through the AVO moving coil meter and half through the second potentiometer which in turn means that they need to have the same resistance for this to work.

One more test you can do is to put your DMM in series with the AVO moving coil meter in circuit and set it to the proper current range and check what current flows through the circuit when the Mk IV is in the SET AC setting.

These tests require that you really know what you are doing lest you might damage the AVO moving coil meter permanently and you'll need some components for proper testing.

I hope I didn't confuse you more.

[The Philpott]

Regarding resistance testing of meter movements and other sensitive devices, a manual ranging digital multimeter is a good thing to have. The 20 megohm and 200 megohm ranges on a couple of my DMM's are able to throttle the applied current back to a safer level.

Dave

[Dekatron]

You can switch the Fluke 87V to manual mode too, but it still doesn't have a current range that corresponds to 30uA, nor does most other multimeters according to what I've seen discussed on forums.

It's also slightly more dangerous as most DMM's automatically switch back to automatic mode when turned of and on again which makes it easy to make a mistake.

However what you can do with with this constant current source that's built in to a DMM is that you can build an external current divider by much the same way that I describe with the 1.5V dry cell above. But then you need two DMM's so you can measure the current at the same time as using one DMM as the current source on resistance range measurements. Let's say you have 100uA on one range, then you'll have to get 30uA through one branch and 70uA through the other branch - since you know what the internal resistance of the meter should be, which is 3250 Ohm, and the FSD current of 30uA this results in a voltage drop of some 97.5mV and with 70uA flowing through the other branch you get a resistance value of some 1392-1393 Ohm (97.5mV/70uA = 1392.86 Ohm). This effectively results in a resistance seen by the DMM of 1393 Ohm in parallel with 3250 Ohm which equals 975 Ohm which of course corresponds to the 97.5mV at 100uA through this resistance.

I clearly recommend the 1.5V dry cell way as you only need one DMM and you can easily measure the current flowing in the circuit. You also won't risk forgetting the manual mode of the DMM.

[Craig Sawyers]

Originally Posted by Dekatron

I recommend that you either use a proper constant current generator and set it to 30uA and see where the needle ends up or that you build a design like the one I posted here some years back (can't find the link now though).

I recalled that you'd posted a circuit quite a while back. Took a while to track it down.

Post 34 from https://www.vintage-radio.net/forum/showthread.php?t=166489

I clearly saved it! Attached

[Dekatron]

Thanks for finding it!

That circuit is for the VCM 163 50uA meters so you'll have to increase the series resistors some to limit the current to some 30uA.

Originally Posted by Craig Sawyers

I recalled that you'd posted a circuit quite a while back. Took a while to track it down.

Post 34 from https://www.vintage-radio.net/forum/showthread.php?t=166489

I clearly saved it! Attached

[TomAW-60]

Thanks for explaining and thanks for the detailed and very precise description for how to measure the meter in a safe way! This certainly helps to lower confusion resistance!

This arrangement I think I can make. I’ll take care and use that 3250? resistor as a check before I try it on the meter. I think I’ll even put in that switch and the second potmeter.

Repeating for my understanding: Using this battery/resistor/potmeter/DMM circuit I would then feed the meter with max current (approx 29.85uA - as FSD is 30uA). If meter is ok, I would then expect the meter pointer to show full scale (5k potmeter set to 0) - will the DMM itself represent any resistance in this circuit?

The red set mark will be 90% of FSD - 30uA x 0.9 = 27uA - which then corresponds to the potmeter set at full 5k resistance in circuit with current 27.15uA (I’ll measure the potmeter first then).

If now bringing in the switch and 2nd potmeter to the circuit, turning this potmeter so that the pointer shows half of FSD (the 2.5 M? mark on the M? scale). I'm now able to measure the meter's resistance exactly by measuring the 2nd potmeter’s now set resistance - and I would also get an indication of meter linearity at red mark and at half FSD, if understood correctly?

Then a confusion point - I read in document:
"The meter in itself has an FSD of 30µA and a total internal resistance of 3,250?. Together with the shunt at 10K? it will result in a meter that has an equivalent FSD of 39.75µA and an equivalent internal resistance of 2,453? 3,250? // 10,000? = 2,453? The FSD current for the shunted meter can be calculated from the voltage drop across the original meter divided by the shunted resistance 30µA x 3,250? = 97.5mV 97.5mV / 2,453? = 39.75 µA. The two silicon diodes, MR7 and MR8, which are connected across the meter connections in the AVO Mk IV, one in each direction, ensure that the maximum voltage across the meter terminals will only reach ~0.7V"

(0.7V - that would mean 7 times more than meter FSD? If diodes are still working, would this generally have saved the meter when/if tester was operated incorrectly when measuring tubes (wrong settings etc)? - I see there is a recipe for a better meter protection circuit using upgraded diodes and a 100uF polyprop cap)

Reading more; the meter resistance would be made up by meter itself, a “swamp” resistor (tuned by AVO) and the 10K? “shunt” resistor. So, when doing this test, I would then need to know if the 10K? resistor is in the circuit I measure or not. Thus, next question popping up is then where to find the correct measuring points for the 30uA test (vs the 39.75uA with the 10K? resistor included in circuit, if understood correctly?). For the 30uA test, would that be the two nuts on the plate inside located at the back side of the meter with green and blue wire connections (see picture)? Battery + end to be connected to green or blue wire nut?

The second approach: Check what current runs through the meter when pointer is set at red mark. So first, in my case - with current condition of tester -, I would then need to move the fuse to the lower voltage setting 210-230V (in order for pointer to reach the red set mark). And then I would need to loosen either green or blue wire from meter back plate, connect this loosened wire to one DMM lead and put the second DMM lead to the other nut point of back plate, setting DMM in position uA? (sorry for this “dummies” need for details). Then turn power on and adjust the set knob on front panel for pointer to align with red set mark. If now the DMM shows above the 27uA needed (or alt. 39.75uA x 0.9= 35.78uA with 10K? resistor included in circuit), my meter would have too low sensitivity, correct?

I now need to get the parts for making the test(s).

Making the circuit (1,5v battery, 47k? resistor, 5k potmeter in series with meter) would you solder most of this or would the use of wago clips be ok? (Would this test of meter FSD be something I should do/need to do periodically?)

Thanks, Tom

[Trigon.]

If a bench supply and DVM are available another possible method of measuring meter resistance might be:-



Adjust the pot for a null across the DVM, disconnect the adjusted pot+resistor and measure. (Sensitivity will be approximately 1mV for a 1% difference in resistance between meter and reference.)

Cheers

[TomAW-60]

Edit - for the current test - of course put the DMM in series, thus just put the DMM inbetween - loosen wire to DMM lead and second lead to nut point of the loosen wire

[Jennings]

Hi Tom. Maybe take some time familiarising and digesting the calibration process and how it all works before rushing too far ahea. Although my mkII meters have a different spec, it might help to have a read through the active thread where I’m also learning and trying to get to grips with repairs and calibration…and the kind, knowledgable (not to mention patient) folks here have helped and guided me.

[TomAW-60]

@Jennings: Thanks for comment/concern! Yes, the more I read, the more the universe of meter details come to surface. Starting from scratch with little competence, I’ll have to take it step by step in a very systematic way. I’ve been reading Martin’s documentation on the AVO testers (and thanks Martin for valuable comments here!) and a series of threads on the AVO VCMs. Your thread on Mk 2 and your work on your meters is very informative! Later today I plan to test my Mk4 meter - I’ll post the results later. Crossing fingers!

Yesterday I got hold of some parts and a box of different resistors, built the circuit - even soldered - and then spent two hours+ wondering why my first ever circuit did not work… AA battery measured 1.6V, resistors values ok, potmeter ok, switch ok - but no current running as DMM showed a constant 0.00uA. Googled Oms law … In Sweden there is this child character, a young girl named “Pippi Långstrump”, made up by the fantastic author Astrid Lindgren - her (Pippi’s) motto is: I’ve never done this before, so I’m probably very good at it! Me, confidence getting lower and lower by the minute, I then discovered I had to move the one measuring wire/pin to another hole on the DMM front….. so now back in Pippi mode?

[TomAW-60]

As there are only two bolts coming out of the back of the meter with no other connected leads than the blue and green wire (disconnected), I then settled for the 30uA FSD measure. I put in a 3281ohm resistor in circuit first and adjusted the pot so I got 30 uA exactly. Then checked the pointer movement direction with just a touch of connection - the pointer seemed very sensitive… (I got afraid I was overshooting somehow). But it was just a very responsive pointer….

I couldn't figure out the “current splitter” set-up/measurement properly, as I had to put in a lot of resistance into the secondary potmeter part to get down to half scale reading (switched in series with the initial resistor/pot giving exactly 30uA to meter) - measuring that second pot separately would read a lot of k ohms, nowhere close to the ohms of the meter - so I got this wrong (I think I know why now, it should have been a separate additional circuit over the meter and not in series). Anyhow, I did measure the pointer position on the 100 mA scale at variable currents regulating the resistor load, noting current and total circuit ohm load. Result were as follows (AA battery 1.599v) - (sorry for layout, didn't manage to load my table view)

Circuit ohm load - Current uA - Pointer at mA scale - Scale/current

47.65 k - 31.25 - 98 - 3.14

49.72 k - 30.00 - 94 - 3.13

51.37 k - 29.26 - 91 - 3.11 (red mark)

61.97 k - 24.38 - 76 - 3.12

69.50 k - 21.88 - 68 - 3.11

91.14 k - 16.88 - 52 - 3.08

96.27 k - 16.00 - 49 - 3.06 (half of max 98)

117.4 k - 13.29 - 40 - 3.01

135.4 k - 11.50 - 34 - 2.96

140.7 k - 9.20 - 27 - 2.93

Sourcing 30.00uA I got 94/100, so the meter is then undershooting with approx. 6%?

Linearity seems to be somewhat acceptable...? Although seem to loose senistivity gradually at the lower part of scale.

Meter resistance is perhaps still an unknown measure?
(or 3250ohm/0.94= in range of 3457ohm?)

[TomAW-60]

Measuring AVO meter resistance

I finally managed to make the current splitter set-up with the parts - I initially was a bit confused about the layout of resistors/pots in the circuit. A learning process. So, the circuit branch will be:

1.5/1.6V Battery(+) - 47k resistor - 5k pot - (100k pot) - DMM - Meter(+) - Meter(-) - Battery(-).

Checking the circuit I used a 3.3k resistor in place for the meter. I found that an additional 100k pot in the circuit will allow for running the meter at lower current levels than 30uA and enable check of the meter’s linearity.

The current splitter branch will be an additional branch from Meter(+) - switch (open/close) - 5k pot - Battery(-).

I preset this 5k pot to 3250 ohms (meter specs). When activating the current splitter branch the preset circuit current of 30uA (meter FSD) will go up due to two resistors (meter resistance and splitter branch pot) now being in parallel and the total resistance of the circuit is thus somewhat reduced. By tuning the 5k/100k pot in the first branch I then adjusted the current for the total circuit back to 30uA. Next, the meter pointer can now be adjusted to show half (50) of the 30uA scale value (100) by tuning the 5k pot on the current splitter branch (for me scale values are 47 and 94, respectively).

Now deactivating the current splitter branch, the resistance measured over the pot in this branch will be equal to the resistance of the meter itself. If the 5K pot in the current splitter branch is preset to 0 ohm, there will be no current increase in the circuit when activating the current splitter branch (no overshoot of 30uA current in circuit).

Tuning the 100k pot in first branch allows for running different levels of current over the meter (below 30uA), repeating the test of an initial scale value set vs half scale value when the current splitter branch is activated. This allows for checking the meter linearity over the full meter scale. Furthermore, it is wise to have the 47k resistor directly following the Battery(+) as this will act as a safety measure for the max current running in the circuit.

The above is of course elementary to most readers on this forum. I’m sharing this for any potential newbie like me who wants to test their AVO VCM meter

[TomAW-60]

Measuring my VCM Mk4 meter I came to results: 30uA= 94/100 on scale, resistance= 3390 ohm which was constant over the meter on different current levels (I checked scale values 90,80,70,60,50,40,34). I presume this means the meter’s linearity is ok.

But the meter is then a bit out of the specification of FSD 30uA and 3250 ohm resistance (140 ohms is outside variation margin). A few questions pop up:

If this higher resistance is constant over the meter, may it be treated as a factor so I can interpolate actual meter measurements with factor 100/94, both for Ia and Gm measures? (thus leave the meter as is)

Can I adjust the meter in any way:
Is the swamp resistor (meter temperature compensation function) in series or in parallel with the meter coil?
Is this a re-flux thing of the coil? (which I believe is hard to fix from what I read in one discussion thread I came across)

Doing the “sideway” test of the tester/meter (meter at 0 in normal position): When the tester is tipped 90 degrees to stand on the left side, the meter showed 4. When tipped to stand on the right side, the meter indicated approx. -3 (below 0 point).

It would be nice to get some reflections on this before I go to the next step of the check process (to find the reason(s) why I cannot reach the red set mark with correct fuse voltage setting - which is much more off than 94/100). Thoughts appreciated - thanks

[vintage port]

Hi Tom,

These VCMs aren't the easiest to cut your teeth on, so chapeau for trying! I'm a beginner still, having calibrated my MK4 many years ago, I was very lucky that all of the resistors in mine were spot on still, it hadn't been messed about with or abused.

My meter was a few percent low, like yours. There's no benefit worrying about its resistancessince that can't be changed anyway! What you CAN do is to make a little, adjustable, ultra low gain opamp circuit presenting 3250 ohms to the VCM and use that to drive the meter. At 6% over spec amperage driving the meter you won't do the coil any damage but you do get the meter set accurately so you can move on. The next (or previous) thing I would do is make sure that the bipolar cap across the meter is of sufficient capacitance and fresh. These can go dry and a "reactive" meter will soon be a damaged meter once the inevitable dud valve entertains when measured or the wrong scale is selected by accident. My meter is heavily damped for exactly this reason: an overcurrent won't save the coil though you've usually a second or two to deal with that anyway, but avoiding the pointer swinging hard onto the stop may well save the mechanism. Hope this helps a bit.

[TomAW-60]

Hi VP,
Thanks for that key information and that interesting fix! How would that circuit with needed parts look like? Would there be a drawing on this and a description on how to tune the opamp?

Yes, and to your second point, first thing now, I need to secure the meter from potential high current overload. I found information on this by searching the forum. To what I see, Karsten Simon (“Karsten”) has provided a lot of insight on this, testing different approaches and discussing/documenting the results for an AVO 30uA meter. There is a first evolved “simple” solution (using 4 diodes and a 100uF cap) in this thread, post 58:

https://www.vintage-radio.net/forum/showthread.php?t=55782&page=3

and then later he posted a more sophisticated set-up using the tester’s T2 transformer to source the circuit (next page, on page 4).

In terms of complexity I think I should go for the first solution… . A question would be if the protection circuit can be implemented just as shown in the drawing (see attached), positioned at the plate behind the meter, simply connected to the blue and green wires going to the meter?

Then, would it be possible to combine the opamp fix and this Karsten v2 protection circuit in a nice way?