The "Sussex" Homebrew Valve Tester.

[Top Cap]

Hi Mike, sorry for that error and due corrections made. It arose because that 'physical' sketch was made very early, before my Sussex was made, with a view to obtaining some feel of how it would all go together. As you no doubt realise, the original Sussex circuit had no 50V position and the switch had 11 positions. I added the extra position which meant the selectors now used all 12 positions. The 'Physical' sketch, still stored here as XP (experimental) did not get the necessary update. With respect to the RMS Meter, well I am fortunate in having one, albeit a rather ancient HP3400A but still doing sterling work. I have often wondered if it is possible to measure the 100mV using the existing 200mV AC panel meter so that you can set the panel meter to read 100mV. At least any error will be common to both your signal drive setting and that obtained when the meter reads the Anode circuit signal. Edited: Because I still keep thinking the drive is 1V (silly) and not 100mV. Have 1V on the brain, Geez

Just added a pic to show you my idea but would it work?

Probably best to set Grid Volts to zero first

[adriaanhendrik]

Hello to all,

I have finally started to build my version of the Sussex.
Since my last post I have been abroad, so I could not start earlier.
As you can see, I have a lot of wiring still to do, so I am happy with all your remarks about the problems each of you encountered with the build..
It sure helps me to avoid mistakes.

No doubt about the Frontpanel. I had that one made by Scheaffer. As you can see I made a mistake with the voltages near the switches. I have to figure out how to redo the 120 V position. Maybe painting black the 5 and use a white 'rubbing-off' digit 0.. Any idea's?

While building the PB, I could not resist to test the -47 V. Using a small chinese Oscilloscope and the panel meter. I have a nice waveform (with a little hum, but that aside.)

The switches are army surplus from The Sovjet Union that i obtained from a guy in Romania. They are very, very sturdy.. I even had to loosen one of the springs that hold the ball, to make it click, with the relatively small knobs that I use. (I had to repeat that 12 times..

The relatively small cooling-blocks are attached to the underlying metal grid-plate, that houses all the trannies and boards, so I do not expect overheating of the BUZ80's.

When i have progressed a little more I'll give an update on my adventures...

Happy hollidays...

[Mike Brett]

Hi Les
Interesting idea, looks simple enough,I will wait and see what other more experienced minds think about this one.
Mike

[Mike Brett]

Richard
Many thanks for the info.
Cheers Mike

[John Coles]

Hi everyone,

I have been following this thread for a while now and have decided to take the plunge and have a go myself, I'm a bit of a newbie to this stuff but I feel confident that with all the information in this thread and all the very helpful people here I can get it done (and learn a good bit along the way).

So far all I have done is order the panel meters.

Could anyone tell me if the transformer for this project is still available and if so how to get one.

Thanks.

[Top Cap]

Hi Mike - Placing a single pole (biased to OFF) toggle switch alongside the Cath/Htr LED might do the trick, after all the CT160 does not permanently measure this parameter. The biased toggle will make sure the -45V line never gets to the heater supplies except when activated during indirectly heated valve tests. At any time during the tests of such valves, the toggle switch can be operated to make sure insulation is still OK, I found testing some KT88's, that it is very desirable to be able to do this. One of the valves being tested was fine until it had run for a few seconds when the C/H insulation failed. This was as good as a thermal switch so I reckon it was due to thermal expansion of the internal connections. I am hoping to add this mod to my tester a.s.a.p. and I may also experiment with the AC meter too. I think that a small slide switch may do for that (if the idea is sound) and could be hidden inside for calibration purposes only.

[Mike Brett]

Hi Les
I was going to fit a toggle switch to protect my 20v heater meter when using the jump lead between the two blue terminals. I am also waiting for another 48v relay I have ordered to protect the meter when turning the Function switch to the 0 position. I was going to power this extra relay from RL2. When the Function switch 0 position is selected RL2 is energized and this would also energize the extra relay thus cutting the 45v to my 20v heater meter.
I thought at least this way I could not forget to operate a switch and destroy another meter. Can you see any problem with this idea.
Mike

[John Coles]

Originally Posted by G8UWM-MildMartin

Yes, I finally got round to ordering 30 of Mike Rowe's original design this week.

Hi Martin,

I would like to order one of these boards if they are still available could you let me know about ordering details.

Thanks

John.

[Mike Brett]

I have just tried my first valves in my Sussex. I did not have any EL34 so I used the 3 ECC83 valves for my first test, results are as follows.
G.M. Anode Current
Valve one 0.7 1.6
Valve two 0.6 1.2
Valve three 0.5 1.2
As you can see apart from the first valve Anode current was about right.But the G. M. meter readings are a bit low. Do you think this indicates poor valves or is the meter reading wrong. As this is my first test I am not really sure what to expect so would like a
little help if possible.
Cheers Mike

[crackle]

Hi John
Ed Dinning can wind the transformers, he is contactable through this forum.
Mike

[John Coles]

Thanks Mike,

I've sent you a PM.

[mole42uk]

Originally Posted by Mike Brett

As this is my first test I am not really sure what to expect so would like a little help if possible.

Mike,

I would grab the valve data sheet and look at the chart of grid voltage versus anode current. You'll see a slope which you can test on the Sussex by charting the Ia against a number of specified Vg points. This slope should be quite close to the 'book' value, and should be indicated quite accurately on the Gm meter, which reads mA/V.

I have tried several EF86 valves, most of which perform below the 'book' values. I have also tested a valve of known characteristics and found that my Sussex gave me the correct reading on the Gm meter, and when I plotted the characteristic mA/V slope it was exactly the same as one plotted on a CT160.

Don't forget that the 'book' test involves the ECC83 anodes being shorted together and the valve being tested as one unit. Maybe time for another toggle switch?

The will also be some variations in Gm readings if the Vh is slightly off 'optimal'. My Sussex has a Vh on the 6.3V setting of 6.74V so that will account for some of the differences I get in my readings. I don't see this as an issue, however, because once I have 'calibrated' the Sussex I will always know how much it over- or under-reads.

I am setting out on a scheme to map all the valves I have on the shelf (not too many, I service vintage instrument and PA amps mostly and my stocks are a quite limited range) against the data sheets, so I know which of my valves are close to spec, and how far out the Sussex really is. So far, it has been remarkably close to the text book!

Hope this helps,
Richard

[Mike Brett]

Richard
Thanks very much for explaining it all to me. Assuming all is well with the circuit then, its just a case of testing various valves and keeping a mental note of the expected differences
for future reference. I will go through the rest of my valve stocks and check the results
to get an overall picture of the accuracy of my Sussex.
Cheers Mike

[mole42uk]

Hi Mike,

It sounds to me as though you have the main Sussex working according to plan. It is, of course, open to modification, as Les has done and others have contributed to. We have a DIY valve tester that is potentially as good as an AVO, without the price tag that those invariably command.

At the moment, I am very pleased with the Sussex. It is so much better than my Mullard tester, which is really a go/no-go device having no real indication of measurement.

One plan that has been forming in my mind is to regulate the heater supply. David Simpson, on another thread and in a PM, told me that the heater supply is hugely significant. He always uses a regulated DC heater on his CT160. I have a plan to revise the heater supply with a DC regulator and set up another meter on my Sussex to indicate the voltage and the current. Then I can set up the test situation as close to the data sheet spec as possible.

Richard

[Ed_Dinning]

Hi Gents, valve emission is roughly proportional to heater power to the power of 1.5, so that heater voltage is important to obtaining correct results. Most commercial valve testers that use AC heater connections have a setting to "standardise" the main voltage with a multitude of primary taps on the transformer.
My transformer will allow matching to the mains in 10V steps from 200 to 250v. It is worth checking what the voltage is at your location, and if it varies.
There is also the question of transformer regulation.
Given the same mains input, there will be a slightly lower voltage on the heater of a 6.3v valve that consumes 1A, to that on a valve consuming 100mA.
The use of DC, regulated heaters would be an option to increase accuracy.
The 12.6v heater tapping could be used at a max of 1.5A to power an LM317 type of regulator, with either a switch and presets or a pot and voltmeter to obtain the desired heater voltage with greater accuracy.
I can arrange to wind a single, 15V, 3A heater winding in place of the present tapped winding on these transformers at time of manufacture if required.

Ed

[Radio Wrangler]

Why the need for great precision, anyway?

It's always good to be on the more accurate side, but if the aim is to check whether a valve is still worth using, then the bounds needed are quite wide.

For a valve manufacturer experimenting with different constructions and materials, accuracy is important.

For checking that a pair are well enough matched, absolute accuracy is less important than repeatability.

For deciding which valve is best for pace, authority and rhythm, consult an astrologer. They've been in the business far longer

Just for the hell of it, I like the idea of carefully regulated supplies, heater and all, for decent repeatability of measured results free from reliance on mains voltage stability, but I'm honestly not sure what actual difference it would make to the things I do. But then I don't flog valves on ebay.

David

[Ed_Dinning]

Hi Gents, I've just discovered a valve tester in my collection that was home-brew but by a professional engineer. This has a pot on the heater supply and a thermal ammeter ti set the heater current to its precise value.
As I should have said, it is the current to the power 3/2 that is proportional to the emission.

Ed

[Dekatron]

Thank you David (Radio Wrangler) for bringing up the repeatability as that is an issue with many valve testers that people often forget.

Even with computerized valve testers there is a problem with repeatability in the tests, component temperature affects the measurement circuits and even though the voltages and currents are measured and controlled with precision A/D & D/A ic's the components in the measurement circuits vary due to the temperature and that results in measurement errors. Modern computerized valve testers also needs warming up to get stable readings, and more so to get repeatable readings. Don't confuse this with the time you need to heat up some valves properly as that is something else, but it sure helps some testers if you heat up the valve in the tester as that also heats up some or all of the measurement circuits too.

One more very important thing is to make sure that the valve you test have clean pins and that the socket you put it in also is clean as you will otherwise get somewhat different readings every time you test a valve, even with the same valve. High heater currents and low grid currents are affected by dirty pins and sockets more than people realise and that spoils repeatability and measurements in general. High contact resistance (or poor contact) is not only a problem by the voltage drop it incurs but also by its changing nature, it usually varies with the current flowing through it.

So using an old grubby valve tester or a new one with grubby sockets or grubby valves will not result in good, and neither repeatable, readings.

/Martin

[mole42uk]

Originally Posted by Ed_Dinning

The use of DC, regulated heaters would be an option to increase accuracy.

Hi Ed,

I'm scheming a DC heater supply around your transformer. Would it be possible to use the 6.3v tap as a centre tap on the 12.6v winding at 1.5A? I want to build a supply fat enough to run an EL34 valve which needs 1500mA on the heater, so a full-wave LC filtered rectifier will get me close enough to that for test purposes.

Best,
Richard

[Ed_Dinning]

Hi Richard, for short times only.
6.3 and below are rated at 3A, 12.6 is rated at 1A.
It would probably be better to use another small transformer and wire the secondary's series aiding.

Ed