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

[TomAW-60]

Waiting for parts to first secure the meter with the protection circuit (before I do anything stupid), I try to prepare for the check of the Mk IV voltages. I do struggle to get my head around mean and RMS voltages… as there seemed to be some kind of contradiction in the service manual….

Originally Posted by Dekatron

Now, don't forget that the calibration factor that AVO mentions as being 0.525 is the scale factor on the Grid-volts control and has nothing to do with any AC/DC/RMS or other voltages to do at all!

Thanks Martin, for pointing this out! This was clarifying, as I was (still is?) into a cognitive loop on how to match your explanatory documentation (linked to in your above post) with the service manual’s steps c) to e)... as they both have references to mean values and RMS values.

Because when looking into the service manual (page 6), while step c) references RMS values, then for e) (the grid V adjustment) the manual suddenly swaps to mean values! ... so here at e) this AVO conversion factor suddenly pops up … corresponding to …what?

The Mk IV service manual states the following:

(c) The unloaded anode volts measured with a standardised Model 8 or Model 7 AvoMeter should be such that the r.m.s. reading on the Meter equals half the calibration on the panel of the Instrument +6 -2%. -1V (+6 -3% -1V on instruments using two diodes in series).

(d) The screen voltages should be the same as the anode voltages.

(e) The grid voltage should be such that when measured with a standardised Electronic Test-meter, or other standard d.c. mean Valve Voltmeter, the voltage measured between grid and cathode (this measurement must be made with the link open and the meter switch set to 100mA) should be such that the mean d.c. voltage equals the calibrated voltage on the panel of the instrument × 0.52 i.e. 101V negative bias should read, (with the link open) 52.5V mean d.c.

So to me two challenges here:

1) Understanding the RMS and mean values correctly when in practice doing the calibration according to the steps in the service manual, and
2) I’m practically (and I presume most are) unable to measure mean values as in e)! (as my DMM - and the AVO 8 - display RMS values).

In document, section 2.1.3 (page 10) - you explain:

“...105.98V - This Mean DC voltage must be present at the junction of ‘calibration’ resistor and the SET Vg potentiometer, in order to reach the red line at the SET ~ region on the meter!...”, and further

“...How do we get these Mean DC voltages at this junction across the grid Volts circuit then? For the AVO Mk III, Mk IV and CT160 you will first have to open the link ‘Vg Link’, and set the voltage for the Grid Volts control to the required calibration voltage of 52.5V with the SET Vg potentiometer, and then close the link again, and now adjust the SET ~ potentiometer to set the voltage to the desired level to get the meter needle to indicate at the red line at the SET ~ region. The calibration procedure is described in detail in the AVO Service Manual.”

In the service manual e) The “AVO grid factor” =0,52. In your reverse engineering document I then get it is actually: 52,5/105,98 =0,50.

At this point my conclusion is:

I must convert all the calibration measurements steps to one standard, be RMS. And treat the grid voltage setting as an independent figure.

... then came up the question: What Ia voltages are actually displayed on the panel knob of the AVO testers?

You have written somewhere: “... set the anode volts to ‘200V’ and then measure the AC voltage at MR5 and adjust the SET-AC knob on the front so you obtain 220V (222V) RMS ±2% (with a DMM)..”

May I thus conclude that:

The values on the Mk IV panel switches are expressed in mean values. Correct?
The Mk IV meter then also shows mean values accordingly? Correct?

(.... this goes to my needed basic understanding...)


And now I just found the Radio Bygones article by Euan MacKenzie dated 2015 on Mk3 and Mk4 testers and calibration (attached, page 10 onwards). A very informative article (also answers the C1 cap voltage question) - and with the most updated circuit diagrams (Martin). I will have to study this article more closely. From the article, and related to the mean/RMS universe now discussed:

“... with the ANODE VOLTS set to 400V and on TEST, measure the voltage between the A1 and C sockets on the top panel; without anode rectification it should ideally read 444V RMS. Alternatively, if anode rectification has been incorporated, then it should read 200V on a DC (or mean reading) meter. In spite of anecdotal opinion to the contrary, digital multimeters (DMMs) will read the mean value of a half-wave (or full-wave) rectified sine wave on the DC setting..”

Which made me struggle again - unless the 200V referred to here is measured by a not RMS but DC mean reading meter - which is no longer around? Or am I lost here?

[vintage port]

Hi Tom,

I have just finished re-calibrating my Mk4. I am a relative noob and have been kindly guided by Martin (amongst others) here, it is difficult to exagerate how much help he has been, so a huge thank you to you, Martin.

The Euan MacKenzie article had escaped my attention before which is a pity because I think his magnet trick for the meter looks like a fabulous solution for those embarking on the refurb project, the picture is really helpful. Thanks for that link!

I implemented my op-amp solution which works well, photo attached, but unfortunately I cannot find the schematic. I have looked through my file and will keep digging.

Regarding the mean and RMS values you are right, it is immensely confusing. With my meter stuck into the Mk4 however a few things became clear quite quickly, and I write them here in case it might help other confused newcomers who arrive at these problems:

1. The reference point which is often missing in the literature is the zero volts baseline. I guess this is because it is too obvious for a pro! It is the cathode, the easiest thing is to stick a multimeter's probe into the cathode socket in the electrode panel thing in the middle of the top of the unit.

2. If you have the half-wave rectified anode mod (I believe most do) then when measuring the anode or screen voltage it makes no difference whether you opt for 220V RMS or 200V mean, they are both exactly the measurements I get from my meter and machine, so use either. I suspect that they remain the same as long as the sine wave is relatively clean, which mine was.

3. The use of 220 vs 222V may be slightly less important than it appears; it is chosen as a voltage in the middle of the available range. There will be some unavoidable inaccuracies to either side because of transformer winding differences and no doubt other variables. There are enough inaccuracies elsewhere in the machine (and the mains) when not using the 200V anode setting. Of course getting it right is means that at 200V anode the machine can be spot on. The main point (to me) appears to be to even out the other inaccuracies so that they are minimised across the whole range of readings. Likewise the 0.52 vs 0.525 factor. It appears to be a compromise figure. Wth my limited understanding, if you predominantly measure small signal valves a lower factor may enhance accuracy but will shift any non-linearity into measurements made of power tubes, whilst a higher factor may suit measurements of the more powerful and higher voltage tubes at the expense of smaller tube measurement accuracy.

4. Euan's point that DMMs read mean value in DC measurements and, (if configured that way) true RMS in AC measurements, has been borne out on my meter (Fluke 83/5). However, it flatly refuses to measure MR5 on any DC scale .

All, please feel free to call out anything wrong in here, I don't want to add to our confusion!

[TomAW-60]

Thanks again VP for your comments and input! Very valuable to me in my process with the mk4. It encourages me to not give up on this!

The Euan MacKenzie article is very helpful, more practical, especially the picture showing the plate with RV4/RV6 and measuring points - and the uncovering of the mysterious zero v baseline as the cathode!

First, I need to tidy up a couple of loose ends on my confused (non-skilled) understanding of multimeters. I now understand that RMS and true RMS only apply for AC. As EM points out in his article, all multimeters read mean DC… because it is a constant current. Simple as that! (talking to myself) So in the DC domain it is just a question of the meter’s accuracy, I would assume the more precise the better (it is still a puzzle to why AVO in their service manual says “..If unable to use the recommended meters (which are: AVO Electronic meter for mean DC, and AVO 7 or 8 for RMS AC) ensure that those used are of sub-standard accuracy…?)

To your no 2): Yes, it now becomes clearer to me that either use 200V Ia setting and measure 220/2222V RMS AC, or as EM in his article use 400V Ia setting and measure 200V DC on the top panel. They would then correspond and be the same! (with a clean sine wave) I have perhaps reached my Eliza Dolittle moment!

The thing here is then to see the grid voltage adjustment as a separate thing with use of this conversion factor - although the grid voltage calibration needs to be done at the same time of calibrating Ia (according to Martin’s documentation).

To your no 3): Thanks for putting context to factor 0.52-0.525 and the related linearity over the broader measurement area of the tester! I’ll think about that as I both use small signal tubes - various triodes in general - 5V/2-3A rectifiers and then these bigger 211/845 as output tubes (latter though I presume are outside of the range of the AVO’s).

And thanks for the picture of the opamp circuit - looks nice! I will have to think about my 94/100 sensitivity and 3390 ohm meter situation later. For the “magnet trick” however, see Jennings thread https://www.vintage-radio.net/forum/showthread.php?t=222411&page=3 posts 40 and 43, onwards. Some experience drawn there to be included. It seems like it’s either an opamp solution as your own, or take the big step to open up the meter to work with magnets to increase the flux and the resistance by adjusting the swamp resistor value. All just theory in my head for now….

Down to earth, I will first build my security circuit for the meter and then start troubleshooting my Mk4 .. first thing being the low pointer reading when it should be at (or close to) the red set mark.

[Chrispy57]

Hi Tom - as a boy avidly reading all of the articles and adverts in old Practical Wireless magazines in the local library circa 1970, I came across the term "sub-standard voltmeters", which were invariably more expensive than ordinary voltmeters, and wondered why anyone would pay extra for a crappy voltmeter. Years later I came to realise that this term actually referred to a voltmeter that had been individually calibrated against a voltmeter that had been calibrated against a physical standard. This may answer the puzzle you have mentioned?

Cheers
Chris

[Dekatron]

"Sub-standard" means "below-standard" in terms of precision/tolerance of a standard (standardised voltmeter) in the case of AVO calibration meters used - so it is a better meter than the ones AVO recommends when it comes to precision/tolerance.

[TomAW-60]

Haha! - these “AC and DC enigmas” of mine really got my head spinning… thinking the secret key was somehow embedded into the "sub-standard" meter introduced in the very first sentence of the service manual…

[Chrispy57]

Hi Martin - I would normally defer to your expertise in these matters, but wonder if you might find the following details of interest, copied from a university laboratory manual explaining the hierarchy of Standards?

"Theory: Calibration of Voltmeter, Ammeter and Wattmeter is the process of checking the accuracy of the reading of the ammeter under test (AT) by comparing it with the standard (AS) one.

In other words, calibration checks the correctness of the instrument by comparing it with the reference standard. The various types measuring standards available are given as below.

International Standards: These standards are defined by an international agreement. These standards are accurate to the extent the production and measurement technology of the time allow.
These are periodically evaluated by absolute measurement in terms of fundamental units. These international standards are kept and maintained at International Bureau of Weights and Measures at Paris. These are not available to ordinary users even for the purpose of calibration.

Primary/Basic Standards: National standards laboratories are responsible for keeping in maintaining the primary standards. The primary standards are independently measured by absolute measurements of the national laboratories. The result of measurements at various national laboratories are compared amongst themselves. This brings out an average figure for primary standards all over the world. These are available for calibration purpose outside the national laboratories. However the basic purpose of maintaining the primary standards remains the verification/calibration of secondary standards.

Secondary Standards: These are calibrated with the primary standards. They are maintained by involved industries. They are sent to national laboratories for calibration and certify periodically.

Working Standards: These are used to further calibrate the instruments which are made for the users. These type of standards can be named as sub-standards."

Cheers
Chris

[Dekatron]

I'll go and ask the calibration company I use and see if it could be a language/translation problem when I discussed the calibration procedures for the AVOs many years ago. I calibrate my DMM's and AVO 8's at their place when I feel the need to nowadays, we used them at my former employer regularly and I snuck in my instruments then, but nowadays I have to pay for calibration so I do it more seldom.

[TomAW-60]

Hi again…
Still waiting for C1 caps that never seem to arrive… anyway, I made the first step building the “Karsten” protection circuit for my Mk3 and Mk4 (picture from Mk4 attached). I measured both meters before and after installation with 30uA and 20uA, and got the same pointer measures. So far, so good.

Working on my Mk4: According to schematics (and my understanding) the “protection part” for the meter consist of the C3 cap (8uF), MR7, MR8 and the R35 (10k?). My version lacks the MR7 and MR8, and the R35 measures only 1,56 k? … (that is if these resistors actually can be measured in circuit, without disconnecting one end ?). The original C3 cap has at some point in time been replaced with a 10uF cap. I tried to follow the black and red wire connecting it - the black was connected to the Electrode selector, the red to the Meter switch.

1) Is this black and red wiring really correct? It seems like original wiring.

And as the “Karsten” protection will replace it…

2) May I now remove the "original" C3 out of the circuit? (will it add 10uF to the protection circuit - (not included in test) - or does this C3 in my version also serve another purpose?

With meter protection installed (I hope) I did try to fire up the Mk4 and started to measure… as mentioned in OP there are some anomalies in there that I hope I’ll be able to troubleshoot with some help. To be continued…

[TomAW-60]

The C3 cap - may I just remove it now that the other security circuit is installed - anyone that would know?

Feeding 222V AC through a variac to the Mk4 tester, fuse placed in the 210-230 holder, my pointer is about 60% to the set mark (my meter measures 94/100 at 30 uA and 3390?) - so pointer far off the red mark.

With Ia set to 200V, I’m able to get (above/below) 222V AC measured between Va and C when adjusting the set~ knob at the front. So that’s a good sign.

Measuring Vs to C I got meaningless low figures, that be both at the AC and the DC measuring points on the plate (plate 6 - component board assembly).

Measuring the components on the plate 6 board (picture), the two resistors R1 and R40 (both 100k) show drift - 115? and 123? - the RV6 pot measured ok (0-4786?), the RV4 5k pot measured only (0-3935?), so should perhaps be replaced (or could be kept)?

The diodes MR3, MR4, MR5, MR6 on the board all measured OL/0.52 which seems correct. However, the M1 diode, obviously replaced at some point, measured 1.23/0.52 which seems odd? Replacing MR1 with a BYW96E, I then measured 0.47/0.52 when in circuit….while measured off circuit the diode shows OL/0.52? Is this correct, or a sign of something wrong?

The diodes on the part list - named SD94, SD91, 10D8E - can they all be replaced with the BYW96E in all places? (or other alternative?)

I’d be very happy to get feedback on the above

@Dekatron/Martin: I think along the way I read somewhere that you may have worked out a list of resistor combinations for the Mk III and Mk IV with values for replacement of the old resistors - would that be available? I’m unsure of what wattage to go look for the replacements. Thanks

[vintage port]

Hi Tom,

That old wiring looks pretty original to me? There's little reason to adjust it but I'm not sure which black and red wires you refer to?

Since you have fitted a new polyprop capacitor the original C3 can be simply cut out of the circuit from the circuit selector switch. You shouldn't need more capacitance than 8uF unless the meter is very twitchy?

222V into the meter and set as you indicate should allow the set AC switch to get the meter into the right position ie. on the red line. If not you have isues to fix. If the meter is moving to 60% you definitely have problems.

Sorry, I don't understand the "With Ia set to 200V" and the "Measuring Vs to C" paragraphs?

Those two 100k resistors are out of spec: +-10%. They need to be replaced. They are in loads of space so can be replaced with 5W or higher metal film 1% resistors. You need to check all the other carbon comp resistors (the ones that look like R1 and R40 - there aren't that many of them in the VCM - since they may all have drifted. I am coming round to Martin's suggestion that all the resistors should be replaced as a matter of course. There's only 48 of them, it removes one large obstacle from calibration. I was lucky first time, but had to replace about half a dozen when I went in this time including the critical R19 pairs. Strangely the two R47 3.9 Ohm carbon comp resistors were spot on.

Your diodes look like a mix of 1N4*** types from what I can see so have already been replaced at some point. I'd still replace the lot with BYW96E as Martin has suggested, they will work in all cases in the VCM. It doesn't cost too much and gives reassurance, wish I'd done it when I first went through my Mk4.

RV4 and 6 are carbon track pots I recall, but not certain. If so, these are always pretty low tolerance, I wouldn't fuss overly about these values at this point? There is no spec for them in the schematic suggesting (to me) that they are not critical values.

I have attached Martin's list of resistor combinations, be aware that a fair number of the right resistances are available off the peg from eg. Mouser anyway. Be careful about voltage ratings.

[TomAW-60]

Hi VP,

Thank you again for your follow-up! The resistor list (thanks for that work Martin) is very valuable and most helpful! Yes, the importance of replacing most resistors has grown on me - initially starting out hoping most parts were ok and the tester only needed some minor work. Not so, it seems like the only safe path to success is to first remove resistor uncertainty.

Thanks for clarifying that the original C3 cap can be removed. The black and red wire refers to the two wires that are connected to - or end up at - the C3 cap mounted on the plate behind the meter. The new “Karsten” protection solution holds a 100uF cap and 4 diodes, so of a different order than the original/replaced 8uF. The “Karsten” solution really does wonders with the pointer movement - graceful behaviour - and is already adopted by a friend of mine for his Mk4 and his very nervous pointer, as I made three of them straight away.

Yes, the meter is usable (although not 100% on spec), so the 60%- position is caused by too little current fed to the meter at set~ session - the cause to hopefully be discovered, starting the troubleshooting at the “plate 6” -component board (ref service manual). The “with Ia set to 200V” refers to setting the anode voltage (so should have been Va, not Ia) to 200V on the front panel switch and then getting the 222V AC between Va and C at the plate (using a variac feeding 222V as a reference voltage) - all referring to (Martin’s documentation and) the Radio Bygones 2015 article by Euan MacKenzie with the picture on the first page indicating all measurement points at the “component board/plate 6” - showing AC and DC measurement points, respectively (- and now that made me get why I got the strange readings of Vs). This might perhaps also touch on your measurement thing, point 4), mentioned in your post 42 above. I attach a picture of my Mk4 component board/plate, with AC and DC measurement points copied in/over from the picture in the RB article (and hope they are correctly marked).

Now I also understand the RB article’s mentioning of “.. without anode rectification…” as my Mk3 - original with D77 tubes - will measure 444V at the top when set at “Test”, while the Mk4 only shows DC value of 200V (when anode voltage switch set at 400V as test point). Logically, I now can make sense of these numbers after the tour around DC mean, (True) RMS AC, AVO 8 and DMM discussions An additional point to the article is the interdependence between RV6 (Va) and RV4 (Vg), so (in theory at least) the calibration will be an iterative process tuning them both at the same time.

I’ll follow the advice to replace the diodes! And start on the list of resistors and values
I presume the HV means High Voltage and should exceed the max 444+V in there. I’ll definitely pay attention to the wattage ratings. Big thanks for your help again, VP!

[TomAW-60]

I see now that several of my has turned into in my above posts - sorry for that, as it looks like I'm putting in a sort of negative note

[Cobaltblue]

Originally Posted by TomAW-60

I see now that several of my has turned into in my above posts - sorry for that, as it looks like I'm putting in a sort of negative note

Smiles fixed

[vintage port]

Here's a link to Martin's original thread with the resistor value list and explanations in it. It is good reading. Not all of these resistors need a high voltage capability and I'd rather pay up for the right part from Mouser than string multiple resistors in a daisy chain. There is space inside a MK4 but not that much on the rotary switches so, as an amateur doing one machine, if necessary I prefer to parallel resistors rather than make a series chain. But each to their own ?

https://www.vintage-radio.net/forum/showthread.php?t=56207&highlight=avo+resistors

[vintage port]

My smilies turning into as well. Hey ho

[TomAW-60]

Hey ho!

A follow up question for my (better) working VCM Mk3: In the Radio Bygones article (2015) the metal rectifier I/6A (REC 1) in the backing off circuit is commented as "obsolete" and as such should be replaced. Just to be on the safe side: May I just remove the I/6A and solder in an BYW96E in its place, no other measures taken? Thanks

[stratfordade]

Originally Posted by TomAW-60

Hey ho!

A follow up question for my (better) working VCM Mk3: In the Radio Bygones article (2015) the metal rectifier I/6A (REC 1) in the backing off circuit is commented as "obsolete" and as such should be replaced. Just to be on the safe side: May I just remove the I/6A and solder in an BYW96E in its place, no other measures taken? Thanks

That’s what I did, and my VCM3 is functioning well.

[TomAW-60]

Thanks for the confirmation, Adrian!