AVO VCM MK1? & meter FSD

[ianbatty311]

Hi, David,

Thank you for your comments.

I understand the issue of calibration valves.

Be aware that the AVO documentation is in error.

To quote from my extensive research:

"AVO recommend making up a calibration valve. You’ll find a description in AVO instructions and in other places. One description calls for the plotting of the characteristics of a 12AU7. Be aware that I could not get the specified anode current of 16 mA with only 200 V at the anode and AVO’s suggested –8 V bias: manufacturer’s charts suggest some –6.5 volts."

Thus, the manufacturer's documentation cannot be relied on to set the standard for calibration valves.

Let's assume you make up a calibration valve.

You can only do this if you have certified your AVO as accurate: its meter is in calibration (460 microamps for a MKi/II - check the specification, as there appears to be some confusion over the *exact* value), and the anode/screen/grid voltages have been measured and certified as accurate.

Let's say that you have certified your AVO, measured a test valve, and written down the gm value that your AVO returned, and that the bod you send your calibration valve to finds that their AVO repeats the value you maked on the valve. And that either (i) you specify the applied anode and grid voltages, as the AVO specification is wrong, or (ii) that you observe the AVO voltage specification and note the measured gm value, disregarding the AVO gm value.

Your mate gets the specified values, however they set up. So the AVO is in calibration.

All good, yeah?


Yes, and I am OK with this scenario.

But what you have done is to create a derived standard. If it is accurate, that's fine.

What if it does not read correctly?

Is it the fault of the user's instrument, or of yours?

Are you completely certain that your instrument is in calibration? If not, then your calibration valve is not reliable.

OK, so your AOV is in calibration.

Where does the user begin?

They start by checking the meter movement calibration, then moving on to the anode current reading, then checking the applied voltage readings....

Which they *could* have done themselves to begin with, and - in doing so - ended up with a wholly-qualified instrument *and* gotten first-hand, in-depth understanding of the instrument's design and elegance.

Me?

I'd go for the complete qualification as described.

As always, Your Mileage May Vary.

Ian.

[Station X]

I think you'll find that David uses a DC Tester and not an AVO Valve Tester to produce his standard valves.

Originally Posted by David Simpson

I've given away to Forum folk - well over 20 6AQ5's & 6AU5's, all DC standardised & accompanied with an A4 tabulation & graph of the Gm(mA/V) slope(curve).

Regards, David

[ianbatty311]

Hello, StationX (apologies for not knowing your real name).

Yes, a DC meter using a grid-shift method would be accurate and give a qualified reference valve.

That said, I hold to my proposition that:

(i) if a valve so qualified tests OK in an AVO, the job has been done, but

(ii) in the case where the AVO *does not* return the correct value, then the owner will need to follow the complete analysis/qualification process. The reference valve would then be a vital cross-check to confirm calibration.

Cheers,

Ian.

[ianbatty311]

Oops!

Apologies, Graham, I didn't pick up your personal name.

Ian.

[David Simpson]

Right enough Ian, AVO's VCM Manuals leave a lot to be desired. They draw their circuit diagrams like a big railway sidings. Re them recommending those wee double triodes - even strapped together one can only work at 20mA Ia max. Trouble is - get the -ve Vg wrong on your VCM, or there is a faulty pot - then "****" - horrendous Ia & the valve goes Pop! ECC82's these days can cost a few quid.
As Graham says, my Test Rig is DC & has pure real smooth Va,Vs & Vg. Monitored by a Marconi TF1041C built into it. Accurate Ia is measured externally with a trusty AVO8. Heater Voltage is supplied externally from a modern 21st century bench DC PSU with a digital display of Vh & Ih. The HT for Va & Vs comes from two Leybold PSU's - stripped down & built into my DC Tester. As a back up, I also use a modern Newlec DMM to double check all the electrode voltages. Cos my valve holder panel is off a VCM MK1(along with it's thumbwheel switch), and there is a wee socket strip on the tops of all AVO's VCM's.
Yep, its built out of other re-purposed items of vintage test equipment. Didn't cost me much more that a hundred quid. But, it works jolly well.
Both the DC Rig & my MK3 are jolly reliable & calibrated. As an example - have a shuftie at my thread post on 15th March - - "807 Project".
Me bets that no on else has had a go at a heffing big 807, DC Standardising-wise. At the same time my MK3 found 3 pairs of VT.60A's which "matched" jolly well, out of a box of 30. A mammoth task I can tell you.
Perhaps there is an ex "Gold Miner" in Australia who has coughed up a couple of $Grand for a SOTA RoeTest or a lovely old TEK570, and is now a VR enthusiast. Find someone like that to produce St'd valves cheaply & quickly.

My apologies for straying O/T, David

[ianbatty311]

Hi, David,

Thanks for the reply - you've certainly done a lot of work in creating a "test bench" that will create reference valves, and I'd be confident that what you offer would give a reliable indication.

That said, some folks on "chat" sites have challenged the AVO half-wave method as not giving a truly accurate measure of gm.

That is certainly the case for the Hikock line, where they do not use manufacturers voltages or currents, and you have to rely on their own roll-chart values to understand their meter readings.

Have you found that your DC-calibrated valves measure correctly on an AVO? I *expect* that they do - it was AVO's claim from their release of the VCM MkI and onwards.

Your thoughts?

Ian.

[vintage_8bit]

I have just recently received one of David's charictorised valves, a 6AQ5. I was pleased to see the mA & mA/V results on my MK3 VCM matched his AVO MK3 results almost exactly. It's a great idea to see that us enthusiasts are all in the same ball park. I used to work in a calibration department before retirement, but I no longer have test equipment, calibrated to national standards. This exercise does give me confidence that my test equipment is fit for purpose. Colin

[ianbatty311]

Hi, Colin,

Thanks for confirming that the AVO method is accurate - I can "politely educate" those occasional folks who have not put in the work that you, David and others have.

Has anyone caught my presentation yet?

Hi, you might like to check out my three-part presentation on the Historical Radio Society of Australia's website:

https://hrsa.org.au/

and look for our Youtube Channel.

I'd be interested in feedback, comments, etc.

[David Simpson]

Yep Ian,
I've had a good shufti at the HRSA link & your presentation. Its great to see a society with kindred spirits to our BVWS & VMARS. Perhaps one or two of your members might take a leap of faith & build a RoeTest from it's kit form, or acquire a decent working TEK570. Thus becoming a national standard in Oz for valve testing & collect enthusiasts. If you could find 2000 members to cough up a couple of dollars each - that should cover it.
Like Colin, I have a background(many years ago in the RAF) in test eq./lab eq. servicing(quite a few CT160's & CT446's). Nowadays, VR pursuits-wise, relying on pensions income, I cant afford extravagant purchases. Hence I've adopted a Sgt Bilko attitude to acquiring a lot of surplus t/eq. for as little as possible & re-purposing some of it.
Like Colin, a number of years back I picked up a phooked VCM MK1(cost me a hundred quid I think), but sadly(thanks to a tit of a previous owner with a soldering iron & screwdriver) it was BER. So re-purposed it.
A few years ago, I scrounged, cheaply, off Crowthorne Tubes - a couple of tatty CT160's in need of TLC, which I used for a while, then flogged on several years later when that silly VCM price rise started. I certainly couldn't afford the current internet prices for MK3's/MK4's/CT160's/ VCM 163's. However, I reckon that a Sussex kit build is the way forward in the 21st century, at a modest price.
I'm not belittling MK1's or MK2's, or "2 Panels", or HSVT's or Taylors. Kept in good nick & calibrated - a good companion for valve testing/collecting folk.
Hikocks - have worked on one - lot of fannying about with their charts - but they seem jolly accurate. AVO's Valve Data Books - can contain some errors. I also use an Iliffe VDB, and the "Valve Museum" website for confirmation if I have any doubts.
Once again Moderators, I apologise for straying O/T. David

[ianbatty311]

Hi, David,

THank you for the reply and for looking at my AVO presentation. You might also like to check out my two-part "Malicious Mantel", where you'll discover (i) some tips on fauly-finding and (ii) an Australian mantel radio from the mid-1930s that is still one of the *best* five-valve domestic sets of all time.

On valve testers, there are a few HRSA members who might be interested in following up your suggestions.

One is an audio/hi-fi enthusiast who needs to test transmitter-class valves. My experience with the CT160 showed filament voltage sag on high heater current types (EL34/6DQ6), and I reckon that the commonly used valves from the 800-series (805/211) with heater/filament ratings of 10 V at 3.25 A (and higher in others) would either overload an AVO, or just not get a high enough LT voltage for accurate testing, let alone getting anywhere near the 1 kilovolt-plus at 100 mA-plus needed to put the valve into its design operating range.

I'll pass this link on, if it's OK, and maybe they will come on board.

I have another thread going at the moment - I have picked up a Tektronix 524AD oscilloscope that was used as the reference instrument for one of our local TV stations - it's got *the lot*, 10 MHz bandwidth, delayed timebase, etc, etc, so I'll be focusing on that, but I will keep an eye on this thread.

Note to moderators: us researchers like to hoover up *everything*, so we regard all contributions as potentially useful, then sort them in order of value.

Yes, I do get "opinion masquerading as fact", but even that is often useful, as it prompts me to go back and examine *my* theoretical and practical foundations.

Cheers,
Ian.

.

[vintage_8bit]

After a few more wiring errors corrected, I have found that VR2, ganged with VR1 has gone open circuit. It is part of the back off circuit. I have managed to repair this with the aid of a microscope. What I have noticed is, that the wound track assembly seems to just sit in the pot body & so moves slightly as the pot is rotated. Question, is this normally glued in place.VR1 is the same.

I have ordered a multi turn pot to replace, if my repair fails. Colin