Valve testing - what do the readings mean?

[hificricketboy]

Can someone explain what the readings listed on some of the valve sites mean please? Here's an example for a "matched pair" but what does it actually mean? (I get the bit about the numbers being similar hence matched but do the numbers indicate remaining taube/valve life?)

Hickok TV-7B/U checked@ 57/55 58/61
min: 38/38(Triode#1 / Triode#2)

[guypettigrew]

It could be the emission reading for each "side" of the valve.

In other words, valve 1 shows emission readings of 57 and 55 for each half of the valve. These will be double triode valves, ie two triodes in each glass envelope, such as an ECC82.

The reason I say "may be" is because the values seem low. My tester measures emissions on a scale from 0-100. Figures like these would encourage me to get new valves! The Hickok TV-7B/U may measure differently.

Guy

[Tel]

Those values don't really mean anything unless you have the same tester. You have no idea if the tester was zeroed properly or what the numbers mean.

[guypettigrew]

Spot on, Tel.

My guess is the true meaning of the values is that the seller is saying the valves test better than the minimum value, so are worth buying.

Are they a matched pair? No idea. If you take the figures at face value, one triode in one valve measures 55, the equivalent triode in the other valve measures 61. A worrying 11% discrepancy.

I wouldn't buy them!

Guy

[hificricketboy]

Cheers guys. I see that some sellers say "tests strong" but it means bugger all tbh without real data. But the data seems a bit iffy.

I'll have to look out for "new" valves!

[Valvebloke]

Well it doesn't necessarily mean that there's anything wrong with them. At least it's clear that they're not stone dead. The point is that without more information it's pretty difficult to tell just how good they are. The more complex testers will give you two numbers for each valve, or section within a valve - the anode current (labelled Ia) and the mutual conductance (labelled gm). You can compare these numbers with the datasheet published by the valve manufacturer and find out whether the valve is working properly at least at one operating point. Of course you'd really like to know if it's working properly at the operating point used in your amp. Realistically the easiest way to discover that is to put the valve into your amp and see if it measures correctly there. Unfortunately that usually involves buying it first .

VB

[i_should_coco]

From one of my first posts in the tent:

The test numbers that you often see quoted on eBay and such is that they are specific to a model of tester. Unfortunately there are many types of tester and not all give true Gm readings, but indicate the results as an "english" figure of merit value. The most common testers (at least in the US) are the Hickok types such as a TV-7 and it's variants and these do exactly this.

Testers generally have a minimum good value specified for a particular tube type which relates to the minimum Gm where a tube can still be considered good. Typically, tubes are generally considered used-up when the transconductance (Gm) falls below 2/3 (or so) of the new value.

So for example, if the tube tests at "59/38" then "38" is the minimum value and "59" is the test reading. In the absence of any other information, this would indicate that the tube is good, but possibly used. Sometimes you see the seller specify what a new tube should test at, e.g. "65/38". Unfortunately, this not very reliable as tube tends to vary all over the shop (even new ones) so these figures should only be used a a rough guide and not taken as gospel. They'll give you an idea if the tube is good, but no more.

A very high test result (much more than expected) could indicate other problems such as gas, and should be treated with suspicion.

Note, some testers such as the English AVO series give true transconductance (Gm) readings, along with emission measurements, so these results are not comparable to others.

Another class of tester is the "emission" tester - these are inferior to the transconductance type testers and again the results are not comparable.

Hope this helps....

[hificricketboy]

Thanks Valvebloke and Coco-san.

It sort of suggests that tester readings are actually very much open to interpretation and to the extent that they don't mean too much...

[i_should_coco]

I wouldn't necessarily agree that the don't mean much, done properly, they do. The difficulty is understanding what they mean, but also the limitations.

[Jazid]

If the valve tester is either calibrated (unlikely these days) or cross referenced to known standards (ie valves tested on calibrated machines) then you get a yardstick of what the valve's strwngth and life expectancy may be. Trouble is that new valves exhibit wide variance but manufacturers let anything that meets standards through. Thus one fleabay valve that tests 100% may be nearly exhausted if it originally tested at 150%of standard, whilst another showing at 95% may be new. Yes they vary this much, and more. To be certain the valves need to be tested when new and the results noted, along with the tester, such that at some later point the valve may be individually assessed against its own original performance. This is frankly unlikely, so caveat emptor is my best advice.
New valves from the far east are often rather mediocre, and fade faster than the premium Philips/Mullard/Telefunken/Rca/Sylvania crowd, hence the value in sourcing them.
I hope this helps!
Sent from my GT-I5800 using Tapatalk 2

[hificricketboy]

Thanks James. I think it is all a little clearer.

From what Coco has said the example I pulled off an eBay listing...

...the valves are dual triode so test 57/55 each side with an anticipated minimum of 38. The other measures 58/61 each side with an anticipated minimum of 38. I guess this means they have reasonable life left in them, but as I don't know the "new" reading it is difficult to ascertain their remaining life.

Does that sound about right?

[i_should_coco]

I'd say they're good. Normally the minimum reading is 2/3 of the nominal value, so I'd expect if the minimum is 38, the nominal would be 57. Due to significant sample variation, anything in that ballpark (say 54-61) would indicate a strong or even new tube.

Be wary of very high readings as that can indicate gas.

[hificricketboy]

OK, thanks Coco. I think that lot makes sense now.

[Jazid]

I would add that IF the readings are honest and correct then reasonably closely matching sections is another good sign.


Sent from my GT-I5800 using Tapatalk 2