It would be nice if I had a proper valve tester, even one of those things that Tandy had in all their stores would be nice.

So, instead I have to test valves in radios and hope for the best.

A long time ago I had a large box full of 6V6 valves. I haven't seen this box for some years and I went looking for it (again), but still it cannot be found. However I did find a box of octal-based rectifiers, so I thought I might as well test them all. Firstly I emptied the box onto the floor, then put back the TV valves (1B3GT, 6AU4GTA, 6AX4GT) since I have nothing to test them with.

Then I found a few valves that shouldn't be there: an unlabelled 6SN7 equivalent, and 2 unlabelled large 4-pin rectifiers, like 80 but bigger.

Now I could get started - I used a Hotpoint V55ME standard radio that was missing its back panel.

Tested the dozen or so 5Y3GT's - all worked except one. Then the 5Y3G's - all ok. One of these was an 80 that had been given an octal base. Then into the TV valves: about 30 5AS4's, and a bunch of 5V4G, 5Z4G, GZ32, U50, U52, and 5R4GY. About a 20% failure rate with these - mostly wouldn't light up, 2 with broken glass, and a few lit up but then the purple fireworks started.

Still, in the end I have quite a pile of spare rectifiers. I wouldn't really recommend any of the TV valves as permanent replacements for 5Y3 - they work, but draw more filament current. The GZ32 has a 3 amp filament for example, that's an extra 10 watts wasted for nothing.

Lastly I noticed a pair of AZ31's - one was a real AZ31, the other was an AZ1 with an octal base added. The real AZ31 works, but the AZ1 became wreathed in purple in the filament-anode space, so into the bin along with the other pile of duds.

Oh, forgot to mention the 3 Russian 5u4M odd-looking rectifiers that Carl gave me. These work fine as 5Y3GT replacements - and they all appeared to be brand new.

Interesting, I shifted a rudimentry tube tester, to a radioclub member here as I really did not need three.

I have Russian 5Y3's. They actually "ruggedised" them and put in a cathode sleeve. These have a large single tubular filament and instead of the ribbon in an air conditioner duct. These have plates like inverted tank stands like 6V4 etc. Meaning they cannot short like the RCA type design and they do not produce a surge on startup.

5U4 risks cooking an #80 / 5Y3 transformer winding. If the same 5U4 filament draws 3A ; #80, 2A.

5Y3 is an #80 with a changed base. If you rest tubes in a radio the pin outs & parameters must match or you will destroy the tube & perhaps the set rectifier, or transformer,

Hi Robbbert,
I have a Tester surplus to my needs ... Email sent to you.

Email was answered. I made some mistakes above - the 5Y3 has a 2amp filament (I was thinking 1 amp for some reason), and so some of the TV rectifiers will be perfectly fine. In fact the ones with a cathode sleeve will be preferred since they will have better regulation.

This means that the GZ32 wastes an extra 5 watts rather than 10.

The Russian ones - the u symbol is actually an uppercase micro shape (like in uF), pronounced "ts" as in bits. Whatever valve equivalent that is, is anyone's guess. But they work very well indeed.

EDIT: I found a picture and datasheet, it compares very favourably with 5Y3.
http://www.r-type.org/exhib/aaa1820.htm

As for the pin connections, fortunately octal rectifiers are 99% all the same connections, and I don't have any of the "wrong" ones (such as 5Y4). If you're worried, use your ohmmeter between pins 8 and 2, it should measure about 1 ohm. These are the filament pins. Pins 4 and 6 are the anodes. Most of these have a dummy pin 1 and omit pins 3,5,7.

More valve testing today. The first thing was finding the box that contained the 6V6 collection, although there's not so many left, but there's a bunch of other octal valves in there so that's some consolation. There were 4 rectifiers, one usable (U50 - tested ok), so they went to the rectifier box mentioned in post #1.

So, time to test some output valves. Fortunately, again, most octal ones have the same pin connections. So I was able to test maybe 100 valves of type 6V6, 6F6, 6L6, EL33(A), one KT66, one EL37, 3x EL3NG, and one 6Y6. The vast majority were 6V6 of course. There were even a few rusty metal ones (2x 6F6 and 3x 6L6) and they all worked. In the end only 6 failures were found: a EL33 that wouldn't light up, and 5x 6V6GT with various issues (3 had sparks flying around inside).

After that I tested a bunch of ECH33, ECH33B and ECH35 - all good apart from a couple of ECH35's. The ECH33B is an unusual variant - it's a ECH33 with an extra fat base attached (why?) and a larger top cap.

In the previous post I said I didn't have any 5Y4 rectifiers. Now I do - one was found in this new box. Couldn't test it though.

Guys, keep an eye on Silicon Chip mag.
They are running an excellent 2 part article by Ian Batty on valve testers.
Ian goes into the design of valve testers and how they work.
1st part in August 2022.

Fred.

Bit more valve testing. The 2 large 80s did turn out to be 5Z3's, when tested one worked and one didn't. The faulty one appears to have been constructed badly and failure was only a matter of time.

Tested 5 EM34's - one worked, 3 are dull, and one does nothing.

Today I pulled out a Hotpoint L34ME to play with as I wasn't satisfied with its performance, even though it's a 4-valve set. According to both the internal sticker and the official diagram, the valves are 6BE6, 6AR7GT, N78 and 5Y3GT, however the rectifier is actually a 6X5GT and it seems it came from the factory like that. There's no reflex either. I replaced the last few black mud caps with new ones and tuned the 2nd IF and it seems much better. Then I tested some valves in it.

6AR7GT - out of 8 spares, 3 failed (2 didn't light up, 1 did but produced no output). A unique valve in every sense of the word, with its integrated lead shield on the outside of the glass.

N78 - My only spare worked.

6X5GT - out of 14 spares, there was one failure. It started off ok, but then there was a flash and loss of sound, and the valve started getting blue around the cathode. A check showed that the cathode had shorted to the heater. There's 2 main types of 6X5GT - the original one with 2 overlapping plates, and the newer one which is simply a 6X4 in the octal shell. I used to have a rare 6X5G valve, but the heater failed on it, so out it went.

6BE6 - out of approx 30 spares, one failed. It wouldn't light up.

My only EL91 was found to have got a crack in the glass and the inside was white. Sad, so out it went too.

I think the glass in the older valves is getting more fragile as the years pass by, just like old-fashioned delicate vases and the like.

A check showed that the cathode had shorted to the heater.

A very common fault with original 6X5 valves. Always struck me as odd for a valve that was designed for car radios and therefore ought have been rugged.

I wasn't intending to do valve testing today, but that changed. I found 2 more boxes of valves, and I'm pretty sure that's all there is. One box was filled with transmitting valves (Philips QQ-series and the like), and gigantic things such as 815, 808, 810 and so on. Not much to work with there.

The other box had 13 KT66's, and a bunch of 6J7, 6J7G and 6J7GT valves. The KT66's were tested and they all worked. There was also a Cossor 2-volt, 5-pin pentode. It looks very ancient but perfectly clean.

While looking at the 6J7, I suddenly realised that both it and the 6K7 could be used as a replacement for the troublesome 6U7G, because they have a proper internal shield - so long as pin 1 of the valve socket is earthed. I tried both a 6J7G and a 6K7G, and they worked fine. I couldn't test them all because the radio on the bench has a faulty socket, so it can wait until I find another 6U7G-equipped set to play with.

So, unless I find another box of valves, I think I can say that my meagre collection of P-socket valves (and many others) are just gone.

6U7 and 6K7 are not electrically the same they have different cut offs, for one.

Bit more testing... firstly my small collection of EM84, EM840, EM87 and EM85, which fit into my 10-valve set. The first 3 are for practical purposes identical, with a white stripe down the glass and giving a bright cyan glow. They all worked, although some of the EM84's from tape recorders are a bit faint. Some of these don't even have getters.

The EM85 is more like a magic eye of old, with its elliptical target sitting inside the glass and with the traditional green. My one example of this works fine. My other 9-pin magic eyes (EM80, EM81, 6DA5, EAM86) have different pin connections so couldn't be tested.

Now, back to the octal valves. The Pye PE-34 which had a bad socket for the EF22 has had the socket swapped out, fixing the issue of continually cutting out or squealing. This meant I could test more frequency converters. The original was a ECH33, but I found that ECH35, 6A8G, 6J8G, 6K8G, EK2G, VR57, X61 and X61M all worked in the socket. The first 6J8G that I tried started flashing over and I thought I must have pushed my luck too far, but then another 4 of them worked fine so that one must have just been faulty. There was also one faulty 6A8G and a ECH35 - they lit up but nothing happened.

I found a few WW2 army valves, so I looked up to see what they were equivalent to. VR55 = EBC33, so that got tested and worked. VT52 = EL32, couldn't test that. VR57 = EK32, that was tested ok. And lastly the CV1281, listed as restricted info at radiomuseum, showed as being a KTW61 at other sites - couldn't test it.

A brand new (NOS) metal 6L6 in the box was found, this worked fine (as you'd expect). Not a scratch on it.

Rob, fascinating reading!
I get the impression you have a shed full of valves.
Then I realised so do I!
Well not quite, but there are boxes of valves grouped mainly by base size, 4,5 and 6 pins here, 7 and 9 pins there, big fat valves in there, and so on.
Not very scientific but ok for early ones when there was not too many types. You just pick them by sight.
When you get to the miniatures there are so many types, I just have them all chucked into boxes and sort through them when I want something.
Then there are valves still in their original cartons and they are easier to sort through and find.
I guess there must have been Billions of valves made in Australia alone, no wonder they still hang around in boxes!
Fred.

Furthering Fred's comments on Ian Batty and valve testers. Some excellent content by Ian on the AVOs recently recorded at a HRSA meeting in Melbourne.

Part 1, https://youtu.be/ITJE4f-oVfc

Part 2, https://youtu.be/wqQV-yFLUd8

Part 3, https://youtu.be/scEQgrcmZJg

In Part 3 video at about 8mins, he talks about characterising mystery valves ... then refers to a Handout for further information on this subject.
Does anyone have that Handout? Can you email it to me please? I have quite a few mystery valves that I would like to characterise.

Does anyone have that Handout? Can you email it to me please?

Pretty sure that Ian himself would send it to you. His email address is on the inside front cover of Radio Waves.