Valve Testing

[John12868]

I'm not sure if this is in the right place.......

Is it possible to obtain sensible data on a valve without testing with a valve tester?

I'm due to receive a job lot of used valves and am wondering if they can be tested without using a valve tester.

From the research I've done on the valve numbers it seems as though an awful lot are vintage television valves

Regards

John

[Skywave]

Well, the short answer - even if it sounds a little glib - is yes!

As it happens, this is something that I've been toying around with in my head for some time . . .

By making a suitable arrangement of milli-ammeters, voltmeters and potential dividers / rheostats in a test jig to monitor and control the voltages and currents at the various valve electrodes, you should be able to produce a set of static characteristic curves. You'll need a valve data book to tell you what the maximum ratings are and the pin-outs of the electrodes for the device/s under test. And a selection of valve bases. Plus, of course, a PSU to power the whole issue. With suitable switching arrangements, it could get rather ambitious!

And at that point, you've ended up with (almost) a fully-featured valve tester!

Al. / Skywave.

[Station X]

I tried that just to test a few valves of about four types. It got so complicated that I bought a CT160 Valve Tester. However it is my least used piece of test gear. It's much better to just try the valves in a real circut, check that they light up and look for strange voltages, which in many cases will be down to leaking capacitors. Actually some valves will test good, but not actually work as say an oscillator. Equally some valves with low emmission or low gm will work perfectly well in a circuit.

[Skywave]

Yes, Graham, I would certainly agree that for the purposes of testing a valve, plugging it into a "known good" circuit is the time-honoured method - mainly because it tests the valve/s under dynamic conditions.
However, the original Q. did specify obtaining "data" (on the valve/s) - which to me is not quite the same thing. It was this part of the OP that I was addressing.

Al.

[Station X]

You're right Al. The OP seems to be interested in measuring the characteristics of valves. I'm more interested in simple functionality tests ie heater OK, no inter-electrode shorts and some emission/gm. A GO/NO GO test is good enough for me.

[Mike Phelan]

I agree with Graham here - after all, the final result is to see if the valve works OK in whatever it is used in.

I worked at a shop once where the MD had a Mullard "one-arm bandit" tester which he used at night after fishing all our scrap valves from the bin and re-boxing them - including EY86, PY81 etc. TV re-calls were rife.

I only occasionally use my CT160 for real - my main reason for getting one was to see if the infamous O/C pot could be replaced in some way.

There are many valves that will test fine and only work in some applications.

[John12868]

I think I confused the issue by the use of the word data. I'm due to pick up 120ish used valves, the majority of which seem to be old TV valves, which are of no use to me. What I was looking for was an easy way of checking if the valves were any good, in respect to current/voltage. I had considered using a good known circuit, but wondered if there was another way of checking.

[Station X]

You'll need variable HT and LT supplies, valveholder, cathode bias resistor, screen dropper resistor (for tetrodes and pentodes), grid leak resistor and a screen decoupler capacitor. The voltage across the screen dropper can be used the caculate the screen current. The voltage across the cathode resistor is the bias voltage and can be used to calculate the combined anode and screen voltage.

Repeat for each different valve type.

[Boom]

Does a commercial valve tester give the frequency cut off limit for a valve? For some applications this would be essential to know.

[Station X]

No.

[Boom]

Thanks. In that case I can't see what cannot be achieved with resistors, caps and a meter.

[daviddeakin]

Originally Posted by Skywave

Well, the short answer - even if it sounds a little glib - is yes!
As it happens, this is something that I've been toying around with in my head for some time . . .

By making a suitable arrangement of milli-ammeters, voltmeters and potential dividers / rheostats in a test jig to monitor and control the voltages and currents at the various valve electrodes, you should be able to produce a set of static characteristic curves.

Lol, that's not so much testing without a tester as building a tester from scratch!

[Mike Phelan]

John, I'd say that the time spend doing such a large pile of valves to sell is hardly worth it, if they are all common-or-garden bottles.

I'd just check the heaters, but that is purely my opinion.

[Mike-repairman]

One thing to remember is that if the previous owner of the valves had a habit of saving ones that had half emission then many may not be worth keeping.

TV valves especially use to give up the ghost in circuit and got replaced with new ones; it was rarely worth fitting second hand valves in those days.

Mike.

[David Simpson]

John, I'm in the same boat as yourself. Have recently aquired a job lot of approx 400 valves. Even if only half of them pass AVO's test data(they should as most are still unused in their 1960's wee cardboard boxes), and are then flogged at less than or similar prices to those advertized on the internet, then they should cover the cost of my CT160.
To build a temp rig with B7G or B9A valve bases connected to wee fly leads everywhere & various bench supplies is asking for an accident to happen. Avo's data manual requires anode & screen voltages in the 150-250V range in many cases. Much better to buy an AVO 2 Panel tester in working order. They are approx 1/4 the cost of the 160 or Mk4. Unless you've already got them - to go out & purchase(even 2nd hand) HT & LT bench PSU's plus a couple of decent mulimeters - will cost you a lot more than a 2Panel Tester.
Another option would be to ask amongst other forum contributors near your part of the country who have a valve tester - if you could pop over some convenient time, and see if they would kindly test your valves for you. Prahaps a bottle of vino &/or a box of surplus electronic odds & ends might enhance the arrangement.

Regards, David

[Skywave]

Originally Posted by daviddeakin

Lol, that's not so much testing without a tester as building a tester from scratch!

Quite! That's the point I was trying to make!
It depends on just how far, in terms of measurement, you want to go.

On the one hand, you can just simply test for heater continuity with an ohm-meter. On the other, a full-featured valve tester is required. But this is for determining valve characteristics. For "go / no-go" testing, in the absence of a ready-built device (such as a radio or television) that uses the type/s needed to be tested, lash up a simple circuit to simulate the valves' behaviour under typical operating conditions. Some sort of signal source, load, oscilloscope & PSU will be needed for this type of assessment.

Al.