Is this unusual valve behaviour?

[Pat Pending]

Being very rusty on valve theory and practice, I wondered if any of the "thermionic" savvy members could explain something I've noticed while setting up a test rig for a large quantity of valves that I recently rescued from being dumped, (see valves worth saving thread). With some examples, with indirectly heated cathodes, I've noticed that a small current flows if a multimeter is connected between the cathode and grid. With the meter set to a volts range this current has a pd of a few tens of mV. Connect cathode and grid together and measure between them and the anode and the voltage is then around 2V. The only external source of power applied being the heater current. I've only noticed it with small signal triode types so far, and haven't done any real world testing (ie. tried them in equipment) to see if they actually work. Could this be some form of contamination? (they've been unused for three decades or more), the getter silvering looks Ok though, and they read Ok on a valve tester. Anyone got any thoughts? Best regards. Andy.

[saxmaniac]

Could it be a tiny bit of induction assuming you are running the heater on ac? Change to dc and experiment with the polarity maybe? I'm thinking this is one of those spurious little things that means nothing in the real world

[Hartley118]

Looks like you're measuring the natural potential difference within the cloud of electrons emitted by the cathode. This is called the space charge and is always there, though normally a proportion of the electron cloud is drawn off when the anode is made positive by the HT, thus attracting the negatively charged electrons, which then flow as the valve's anode current.

It's worth taking a look at a good textbook from the valve era to understand more - such as Langford-Smith page 3.

Martin

[Pat Pending]

Thank for the suggestion but I've tried DC, in fact one of the 12AT7 types that showed this phenomenon, I series connected the heaters and ran them on a 12v SLA battery, still the same. My curiosity has been well and truly whet so I'll try one in my old valve tuner later and see if they actually work. If it doesn't affect them then that's Ok, but if it does then I'll toss them into the hazardous waste and miss a good research opportunity. Best regards. Tom Edison, sorry,I mean Andy!

[Pat Pending]

Martin, you posted before I got my reply in to saxmaniac. Your explanation seems most plausible but it seems to only effect maybe one in thirty or so of these triodes so far. Andy

[David Simpson]

As Martin says, many amplification valves exhibit the "space cloud" phenomenon. Several factors determine it, including the temperature of the heater/filament, and the coating eg. thoriated tungsten.
Temporary Cathode leakage is often experienced on AVO VCM's after the "Test" procedure is finished(Va & Vs switched off), particularly with pentodes. But it slowly decays as the valve cools down.
Incidentally, similar electron "clouding" could be experienced at the Anode, but to a lesser degree. However its effect is counteracted by G3 in pentodes, for example.

Regards, David

PS. Pat, you mentioned setting up a Test Rig - any chance of some details please? Always of interest in the valve testing fraternity.

[Pat Pending]

Thanks David. My idea behind the bake/test rig is to check if the valves are like the ones I described in the previous post can be revitalized with a half hour with their heaters on and some anode current flowing, (they were all apparently good when stored and hence my idea behind contamination/outgassing). I would also like to incorporate a gain test. Nothing set in stone yet but I've seen a circuit that was published by W.T.Cocking in the sixties that I might try, but I've still to receive the B7 & B9G valve holders that I ordered, I'm not holding out to much hope as the ceramic octals from the same supplier were glass filled noryl RS. relay chassis sockets. Oh well, I suppose they will do. I will post a diagram if get something that works. I would like to hear about anyone else's idea's on this one, I've got a lot to test! Andy.

[David Simpson]

Andy,
You've mentioned a "Large Quantity" of valves, but are we talking hundreds or thousands ?
To quickly discard dozens & dozens of valves with duff heaters - an initial continuity test on a valve tester would greatly help you. Then elimination due to very low Ia would be a further help.
Do you know any Forum guys or fellow vintage radio enthusiasts near you who have an AVO or Mullard or Taylor, etc. valve tester ? Perhaps they could be persuaded to help in return for a handful of decent valves for their own projects ?
The Forum's "Community" section should be of help. Or check the BVWS Membership list, or VMARS. Or just simply put out a request post.
The question of " instant soak testing" valves was raised a few months ago, (check "Search"). Some of us advised against it. Nothing wrong with a steady correct heater voltage & drawing Ia at a lowish point on the Ia/Vg curve.ie. by applying extra Vg. Stripping the cathode of it's coating, and distortion of the screen &/or Anode electrodes due to excessive heat dissipation must be avoided. Have you seen the prices asked for ECC83's, KT66's & 88's, etc. lately on internet sites ? Far too valuable to risk damaging.

Regards, David

[Pat Pending]

Thanks for the comments and info David. If you read the thread I started in this section on the 30th November, it will give you the background info. I saved these valves originally because (1) I could see an opportunity to make some donations for a charity whose services my family may shortly have to call upon. (2) I hate seeing things that are still potentially of use, being thrown away. Firstly the number of valves..I have to say thousands. I have only got one box full at home at the moment, it's contents fill a 24"×18"×10" drawer, and there are maybe 20 times more I've had to store at work. I have a valve tester as I stated in the other thread, but what I've been hoping to achieve with a test rig is speed! Say the initial testing as you have advised on a patch board setup that could be configured for a particular type of valve quickly, then tests for heater continuity, cold and hot electrode shorts, and a "gentle" warm through with heaters lit and a current limited anode/cathode bias. I don't think I would feel comfortable passing cathode boosted valves on to anyone as good workers even if its possible. Also don't get too excited, nearly all of these valves are used, and are"commodity" types, but in the sample I've looked at so far there are a few "interesting" examples that I think should be saved for the likes of "us"! I will have a dig round in a while and see if there are any old bulbs in amongst them and will post a list/photo to see if the membership think they are interesting to the community here. B.T.W. nearly forgot, the 12at7 in place of the Lo valve in the FM tuner worked, and was also Ok in the phase splitter stage of the amp. So we will put that one down to (a very strong) space charge as you and Martin said?
Regards Andy.

[Pat Pending]

Hello all, Iv'e just revisited the measurements that I was making when I "discovered" this space charge phenomenon. I went through most of the valves that I had before, and guess what? They all have the same thing going on, I just hadn't noticed. Well it brings it home to you, you are given a day and a bit training on the theory of thermionic emission for you to eventualy become qualified to fix broken telly's,(yes that's all I had, when I started work the last valve equiped TV had been out of production for a couple of years.(ITT CVC 5 I think..) We were told that we should give it a couple of years then tell customers with the valve sets to bin them and buy/rent new! The rotters!) Here we are in the second decade of a new century and I'm still learning about valves. I bet those amongst you who paid attention in class, or are turned on to valves, had a good giggle at this thread! Anyway to correct what I said before, and examining a very used 12AT7/ECC81 as an example, with the heaters lit, and measuring on a voltage range between the cathode and grid, on section one of the valve, a P.D. of 1volt is present with + on the cathode. The short circuit current being 230 microamps. on section two of the valve the readings were 1.2 volts and 882 microamps. The old valve tester confirmed that although servicable, section one of this example was quite a lot down on emission. As I said, I'm not the guy with the experience to say if this is a good way to determine if a valve is Ok or not, but I would like to hear from others if they can duplicate the test on known good and bad valves to see if they concur using decent test equipment. Getting back to the subject of the "valve mountain" I have taken a representative sample of the old timers that are amongst the ones I have at home, and taken a (bad) picture to illustrate their existance! From left to right top row: UNMARKED, EL38, KT61, UU5, TDD4, KTZ63?, 6J7G, UNMARKED, second row, EZ??, 35Z4, 35L6, EBC33, 6SN7, 25L6, UNMARKED, 12Q7, 6A8GT, third row, UNMARKED, TH41, UNMARKED,KT6?,EL36, EF37A, 7B7, and for scale only a "modern ECC82". These are in various physical conditions and unknown electrical health. ironicaly the two apparently oldest look the best. the question is are these old things any use to the fraternity? More where they came from. Twenty times more of the newer ones! I have just to test them before finding them new homes. Best regards. Andy.

[Diabolical Artificer]

Could it be that what your reading has something to do with the space charge stored in inter electrode capacitance? Hmm, maybe not, the voltage would drop. I've read a few books on deep thermionic valve theory but end up falling asleep before I get too far - too many big equations and very drily written text. I'll do some experiments today if I get a mo to see if I can replicate your findings.



Andy

[kalee20]

The voltage you measure sounds like the normal 'contact potential'.

Usually, in indirectly heated valves, this is of the order of a volt - and a current of a faction of a microamp is available. The grid is negative.

It can even be used to give grid bias - connect cathode to chassis, and use a 10MΩ grid leak resistor, and you'll find that the anode current is equivalent to what it would be with about 1 volt bias. This is in fact a recommended way of using the triode in the EABC80, ECL82, etc.

[Hartley118]

Well Andy, you've successfully demonstrated one of the great things about valve electronics - with a few measurements, you can get a good feel for how stuff works. The textbooks always told us that the electron carries a negative charge and now you've actually seen it on your meter. Interesting observations.

That cloud of electrons that's always present around a hot cathode is an essential 'reservoir' of charge that 'feeds' the anode current. Actually, if the anode were to pluck its electrons directly from the cathode surface rather than from the space charge reservoir, the cathode coating would soon get damaged by all that electronic violence!

Whether you could measure the emission quality of a valve by the space charge potential measured on the grid, I rather doubt. I suspect that even a defective low emission cathode would be able to build up a fair space charge when there's no anode current flowing to drain it away - a dripping tap will eventually fill a basin if there's no outflow.

Martin

[Brian R Pateman]

Please keep this thread on topic.

This section is not intended for sales / wants discussions.

[Pat Pending]

First of all sorry for leading the membership off topic. I will post in the relevant section when its disposal time. The meter measurements that I took were done with the only connection's to the valve being the heater current and the meter leads. The examples I've tried so far have all had a substantial (in valve terms), current flowing, ie. Prime mover the valve cathode and grid, and the load being the meters internal shunt. In "battery" terms the + being the cathode and the - being the grid. I was suggesting that there may be a relationship between this current and the health of the cathode, accepting of course that there would be big variations between differing types of valve. I would speculate that it would be possible by just using a multimeter and a predetermined level for this current for a given valve type, to give a reasonably accurate stop/go type test of valve health. Thanks to all who have contributed so far, and some independent tests would either prove or disprove my sanity! Regards Andy.