Valve tester advice

[stuart_morgan_64]

I want a valve tester, many in the UK seem expensive. I would of thought most are in need of refurbishment. I have seen a Military tube tester TV-2C/U but seem very involved.

Is it better to build one, like the SUSSEX? any advice please.

Thanks

[GrimJosef]

It would be useful to know what you want it for. I've got one and it's lovely to have. It's a thing of beauty. It pleases me a good deal, which is mostly why I keep it. But if I just had a few valves to test then I'm not sure that its cost would be justified. If I had a lot of valves to test but they were all of much the same kind then I'd probably be better off putting together a tester optimised for them. Where a general-purpose wide-ranging tester comes in useful is if you have plenty of valves to test and they're all different. But they'd better not be too different - few testers will cope with a 6C33C-B, or a DM70, or an XG1-2500 !

Cheers,

GJ

[stuart_morgan_64]

I have a palet size box of a wide range of valves to.test. many from medical equipment. Also mainly for domestic radio and communication equipment. Large range of valves.

[Radio Wrangler]

That sounds like a suitable job for a general purpose valve tester.

Valve testers split into two groups:

Emission testers.... with these you get the heater going and bias the valve, then see if the right current flows. You also get tests for gas contamination etc.

Mutual conductance testers.... these do the above, but also perturb the grid bias and measure the change in anode current, so their meter reads the slope of the anode current versus grid bias curve at that point.

(There is a third type, the curve tracer, but these are rather rare)

Emission testers are perfectly adequate for fixing radios etc. America survived quite well wit these things as their dominant type. They did make some mutual conductance testers, but most repairmen used emission testers.

Mutual conductance testers are definitely posher and their advantage lies in grouping valves into matched pairs. Most British valve testers ar mutual conductance jobs.

For some reason, high-end audio aficionados want mutual conductance testers, specifically ones made by AVO. Guitarists are right behind them. Both groups will spend spend spend to get an AVO, especially one of the later models. This has pushed their prices up to levels that rational people could not justify to themselves.

The other British maker of mutual conductance testers was Taylor instruments. They were eventually bought out by AVO and their last model, 45D was made after the take over. Whatever the magic of the AVO is, the Taylor doesn't seem to have it... at least there isn't the same demand and the same price hype.

The sussex is an excellent mutual conductance tester, and easier to calibrate than an AVO. You can put one together for the cost of parts, no hype gets involved.

If you just want to check whether valves are dead or worth keeping, then almost any one would do for you.

If you want to test the valves for sale (eg internet auction of some sort) then what the punters want to pay money for are AVO tested valves. Tested on a Taylor doesn't have the same magic effect. Tested on a home made valve tester will get you laughed at, no matter how good it is. They just want the AVO name.

If your pallet load of valves is going for sale, then you could buy an AVO, and sell it when you've tested and photographed the valves.

Be aware that valve testers are fragile and are easily damaged by incautious setting of the controls and switching meter ranges in the wrong order.

Look on this site at the threads on repairing the things to see what's involved.

David

[stuart_morgan_64]

Thank you for the information. Have read it all and taken it in. I did watch a YouTube video this evening, American which said many of the same things. Bit more reading too. I have noticed Avos seem to go for more money than I paid for my van.

[Argus25]

I would go for a military type Hickok tester, TV10, TV7 or TV2C/U.

The TV7 is not that appealing though because the meter is calibrated in a relative scale, though you can convert that to transconductance fairly easily. The TV2C/U has a lot more metering than you need and it is a little complicated in my view, but there is nothing wrong with one, if you can get a good one at a reasonable price.

Like the bowls of porridge, the TV-10 is just right in my view.

A good physical size. Precision crafted. A well calibrated meter scale, but more to the point; it uses a unique system for dynamic transconductance testing which trumps all other designs, including the AVO's or any other valve tester for that matter. They patented it at the time, it was a revolutionary concept.

The design revolves around a precision transformer, where if the line voltages are set correctly, all other voltages in the unit become correct and the measurement involves unbalancing a unique bridge circuit.

The bottom line is these units are virtually "self calibrated" a calibration valve is never required. They can be set to accurate calibration by nothing more than a single AC voltage reference, or there is a semiconductor tool you can read about here I designed, and the article also how these testers operate, if you are interested. Like many clever designs of yesteryear, this design remains generally ill understood in the valve tester community in many cases. There is also a picture of the TV10 in this article:

http://worldphaco.com/uploads/Hickok...d_the_ACS..pdf

[Radio Wrangler]

The Hickock bridge is indeed very elegant. These and the AVOs and Taylors have one factor in common, they are all designed around the idea of powering the valve under test from unrectified AC supplies from tapped transformers. Having proper DC supplies to supply all the screen and anode voltages needed for the gamut of valves required was seen as something to be avoided.

The Sussex is a true DC tester. Nowadays those DC supplies aren't quite so difficult. It attacks the job head on and does exactly what you'd expect. If you set a grid, screen or anode voltage, you can check it with your multimeter... same with anode and screen currents. They are real DC and easy to measure.

The AC type valve testers have half-wave rectified currents (rectified by the valve under test) Multimeters are scaled to read the RMS equivalent of full sinewaves. So they read wrong on valve tester waveforms for voltages and currents. There are various papers written on this issue. Not easy reading!

The Sussex is a star. It will test valves and you can easily check it.

The AVO is the one that will boost valve sale prices. It is purely a factor of market acceptance. People mortgage their children to buy them then most just trust them blindly.

The Taylors are a bit less versatile than the AVOs (fewer socket types) they work OK but don't have the street cred.

The Hickock is probably the best of the AC testers, but also lacks the street cred. Betamax was better than VHS, but Joe Public went for VHS. (someone will mention V2000 at this point!)

You have to be careful buying AVOs. They have fragile meter movements and are easily damaged. Also, they are the ones sought after by hifi and guitar people, some of whom have little understanding of what they are doing. There is a steady supply of damaged AVOs. Some people even try to use them to burn-in valves (to make them sound better, I suppose). but they manage to burn out the transformers in the valve tester instead. The testers were designed for just a quick test, just long enough to get a reading.

If you want to test valves for your own use, the pecking order is:

Sussex
Hickock (If you can find one in the UK)
AVO
Taylor
Mullard High Speed Valve Tester (It's not a mutual conductance tester)

If you're testing valves for sale, the pecking order becomes:

AVO VCM163
AVO Mk IV
AVO Mk III/ AVO CT160
AVO MkII
Taylor 45D
Taylor other models/Mullard HSVT
Various Emission testers.
The UK market doesn't know of the Hickock models.
and rock bottom, anything you made yourself.


Don't expect logic in the market acceptance order!

Cheers,
David

[Argus25]

Originally Posted by Radio Wrangler

Nowadays those DC supplies aren't quite so difficult.
David

yes I agree with this. When Valve based "Valve Testers" were designed they had to work within the limitations of the valves that created the tester. They did not have the ease of things like transistor based super regulated feedback based DC power supplies. This is where necessity was the mother of invention and how it inspired the genius designs of the bridge circuit and AC transconductance methods which Hickok perfected.

I mentioned this very fact in he article I cited about the calculation of transconductance (blue text below).

Nowadays you have IC's that easily perform mathematical functions, like division and multiplication of signals. Of course when the Hickok or AVO testers were designed, these parts did not exist.

"It could be tempting to assume, that to satisfy the definition of gm, that the tester would measure the alternating plate current and the alternating grid voltage and divide the grid voltage into the plate current and compute and display that value of on the instrument’s meter. This is not the case. Hickok’s designers realised that such a computation notion, involving a mathematical division of data, was completely unnecessary. (A division of digital data or analog voltage is easy with modern electronics, but it is more difficult and sometimes cumbersome with vintage electronics)"

So when it comes to many vintage valve testers, you have to judge them in the context of the time in history they were created.

Also, with the vintage commercial valve testers, a lot of effort was put into the mechanical engineering side of things too, multiple switches, valve sockets and switching options, which some of the modern semiconductor based testers have basically ignored at the peril of errors testing the valve. Because its just too much effort to wire that up in a professional manner.

[M0SOE_Bruce]

I currently have 4 valve testers, a TV-2/U, TV-7/CU, Hickok 539C and a Stark 10-44 (rebadged Hickok). I use both the TV7 and the 539C. The TV7 is my daily driver because most of the stuff I work on is older military radios. The 539C covers newer valve types. When shopping have a look at the roll charts to see if the tester covers the valves you want to test.

The TV2 is a bit of overkill. It's a nice tester but time consuming to set up especially when going through a large batch. For all the variable adjustments it's still only a quality measurement (go/no go).

[stuart_morgan_64]

The reason i have asked is there is a TV2 for £150. Bit far away. Although as comments above seems a little complicated.

[Craig Sawyers]

FWIW I have

1. AVO MkIII. FOC not working from a friend. Vg pot, simple 10k wirewound open circuit. Easy fix.

2. CT160 for £100 with a burnt out meter, from another friend. Got another piece of AVO kit (forget which - I think it was a VVM) FOC with the same meter style but different sensitivity. Used the kind of standard op-amp solution to match sensitivity.

3. Taylor 45D. Restored it with new hammerite. Selector switches are a weakness on this with soggy detentes.

4. Part way though building one of these https://www.dos4ever.com/uTracer3/uTracer3_pag0.html

5. There is an adaptor board that sits in a valve tester, and converts a semiconductor curve tracer into a valve curve tracer. I have a number of Tek curve tracers, the best of which, a 576, I got FOC not working (and fixed). I also have an AVO MkII (got it FOC) with a burnt out meter that will be the valve tester mule. This was developed by a US friend. Circuit board is on its way. http://www.ke5fx.com/An%20Inexpensiv...%20Tracers.pdf

So somewhat spoiled for choice, and nothing has cost me much at all.

But if I was starting out with nothing, I'd definitely go for a Sussex build, or for a uTracer.

Craig

Craig

[Craig Sawyers]

Just for clarity, none of the above are for sale!

With the possible exception of the Taylor 45D that is. I'll think about that and post onto the relevant part of the forum if I decide to part with it.

[David G4EBT]

Originally Posted by Craig Sawyers

Just for clarity, none of the above are for sale!

With the possible exception of the Taylor 45D that is. I'll think about that and post onto the relevant part of the forum if I decide to part with it.

I've got a Taylor 45D which I bought more than thirty years ago when there was little interest in valve testers of any make or model. It had been on a Bring & Buy stand at a rally ignored all day and was an impulse buy as I left, around 3pm. At the time, it cost me £20.00 - equivalent to £60.00 today. Over the years it's proved its worth many times over.

One limitation (which applies to other valve testers in varying degrees), is that it has ten valve-holders, with no space for any more. Originally, adaptors were available to extend the range, but are long since unobtainable, so I made seven adaptors to cover valves which were of interest to me - B4, B5, B7, UX5, UX6. M.O and Side Contact. The all plug into the I.O. socket of the 45D tester.

If I didn't have a valve tester I don't think I'd bother to buy one, given that all you need to do is to plug a valve into your radio or whatever, and it will test it for you, but I understand the desire to test valves, if only out of curiosity. It would be well down my list of test gear after multi-meters, signal generator, waveform generator, 'scope, signal tracer/injector.

That said, for homebrewers such as myself, the Sussex would prove an almost irresistible project, particularly given the improvements over time and the huge body of information about it. It would be interesting to know how many have built, and how many have been started but have become 'stalled' projects.

Hope that's of interest.

[Argus25]

Originally Posted by M0SOE_Bruce

The TV2 is a bit of overkill. It's a nice tester but time consuming to set up especially when going through a large batch.

I agree with that.

Of the mil style Hickok testers the TV-10 is ideal. It has all the features of the TV-7 but with a much better meter calibrated in micromhos and an internal scroll for all the switch settings/valve types . It is a tad bigger than the TV-7, but not cumbersome like the TV2.

I'm not much of a valve tester though and I believe in checking the valve out in its intended application. However, I know a well made piece of laboratory apparatus when I study it and the TV10 is one of those things. So I could recommend it to anyone without thinking I might have dished out poor advice.

[ms660]

I had a 1-177 type, nice and compact and excellent build quality, also it didn't have a "fart and it'll bust" meter...

Lawrence.

[Ed_Dinning]

Hi David, in answer to your question, I have supplied over 100 transformers for the Sussex, so probably at least 300 have been started.
Judging by the correspondence I think a good proportion of then have been finished.

It may lack kudos, but it is simple to use and works well.

Ed

[G6Tanuki]

The "testing" of valves will always lead to controversy, partly because different people want different things from the valves they will be testing!

For me, with power-valves the #1 thing is ability-to-deliver-the-rated-emission - but that's because most of my 'serious' valve-stuff has been for RF linear-amps, where loss-of-cathode-emission leads to distortion/intermodulation-products that could be annoying to users of adjacent frequencies. The sorts of valves I used typically could deliver something between 1 and 2 Amps of cathode-emission - not something your average valve-tester can competently explore.

Again, for small-signal valves a tester is ill-equipped to analyse microphony - which is a significant issue in valves used in tunable-oscillator service [one application I dealt with used a 6AK5 as an oscillator: there was a massive difference in the anti-microphonic performance between RCA, National Union and Tung-Sol valves - no valve-tester could ever determine this].

Test, by all means - test to extremes - but the only 'real' test of a valve is how well it performs in its intended application.

[David G4EBT]

Thanks Ed - there are some excellent examples in the Sussex thread, very nicely constructed and finished off.

I haven't studied the design, but I assume that the transformer has multiple secondaries, so just winding one will keep you busy, let alone more than 100!

[Argus25]

Originally Posted by G6Tanuki

Again, for small-signal valves a tester is ill-equipped to analyse microphony - which is a significant issue in valves used in tunable-oscillator service [one application I dealt with used a 6AK5 as an oscillator: there was a massive difference in the anti-microphonic performance between RCA, National Union and Tung-Sol valves - no valve-tester could ever determine this].

I agree, a valve tester has no hope in this area. Ideally you would need the valve to be set up to detect amplitude irregularities with vibration or possibly amplitude modulation of an HF carrier.

I think this was one reason why Teledyne created and targeted the Fetron version of the 6AK5, the TS6AK5, with zero microphony. Because 6AK5's were notorious for microphony.

I used these in my radio, there is zero microphony of course:

http://worldphaco.com/uploads/WORLDFETRON.pdf

[Craig Sawyers]

Just shows how long it is since I used the Taylor in my list of valve testers - mine is a 45C, not a 45D.

Craig