New 6F4P build, good for first 2 hours, now sounds like long string of fircrackers

[Mick Bailey]

R.G. suggested an idea a while back for testing and measuring microphonics by vibrating the tube with a pager/phone vibrator and measuring the output. Whilst not an absolute measurement, it would give a relative comparison between tubes. The more recent devices could be swept for frequency.

I also like this test setup;
https://www.youtube.com/watch?v=205__gxWJc8

[g1]

Seems there is difference of opinion on what constitutes 'good' in a tube and whether the tester shows it.
For me, it must include noise, microphonics, and actual working voltages that tube will be subjected to in the amp it is going into.
By those requirements, there is no tester that checks all those things (aside from the amp itself).

[Jon Snell]

R.G. suggested an idea a while back for testing and measuring microphonics by vibrating the tube with a pager/phone vibrator and measuring the output. Whilst not an absolute measurement, it would give a relative comparison between tubes. The more recent devices could be swept for frequency.

I also like this test setup;
https://www.youtube.com/watch?v=205__gxWJc8

Should be easy enough to couple an amplifier to the tester but it will have to be a bespoke valve tester as most are chopped AC derived, not smooth DC on the anodes.
I built one a few years ago, here is a copy of my design. I hardly ever use it as most faults occur in a working amplifier so as long as the valves are discarded when noisy, I have no need for it.
All component valuse are nominal; 100k anode load, 1k8 cathode load, 10k grid leak and 10n coupling capacitor.
Main HT smoothing was 100u, heater smoothing 6k8u and thats is about it.
BNC sockets give trouble!

[Helmholtz]

A tube tester is nice to pre-select tubes or even match tubes.
A preamp tube may work fine as a PI but show excessive heater hum or microphonics as input tube.
All tubes are more or less microphonic, because the internal structure can vibrate and thus modulate inter-electrode distances.
The problem increases with number of thermal cycles, i.e. with age.

And which tester measures plate "saturation" voltage of power tubes, which determines available output power?

So the final test is always in the application/amp.

[Mark37]

Thanks everybody for the assistance.

Received the Sylvania 6DX8. No more firecrackers, steady B+ readings, and the pentode plate, screen, and cathode voltages are pretty close to those reported by the designer. It sounds more or less like it did earlier, maybe a little more ooomph, but that could be me being on the learning curve of how the gain and master volume interact.

Next, I'll see if anything changes with a few hours of playing time.

Speaking of tube testers and their capabilities & limitations, what I recently acquired is a tube "checker" --Commercial Trades Institute TC-20. I assume this is a basic unit used by vocational school students, not a pro unit, maybe there's a reason they don't call it a "tester." I have the charts of test settings for various tubes, but no instruction manual. Not really sure how the "line" control is supposed to be adjusted. I have it set so the needle is in the middle of the dial when the "test" switch is not actuated.
For the 6DX8 it lists 2 tests, presumably 1 for the triode and 1 for the pentode. On the first test, both my supposedly (according to the EBAY seller) NOS tubes read in the "?" zone, one at 46, one at 47. On the other test, both read "good" one at 78, one at 83. So they are consistent, but should I be concerned about the "?" zone readings? Did I get hosed on these tubes?

[Mick Bailey]

I think your tester is mainly an emission and short tester. I have one of this type of unit, though mine has a 'calibration' tube which is just an Octal base with a resistor inside. The results are pretty meaningless with mine and even some good new tubes will read '?' or low.

[Enzo]

There are fancy tube testers and there are simple ones. And they may be operated by a very experienced pro or they may be used by someone with limited experience. That said, when someone writes to say they know a tube is good because their tester said so, I generally assume they have a lower level of experience, and the tester is likely basic. And their trust in the tester is misplaced. Of all the tube testers out in the world, how many are of advanced types? Not so many.

[Mick Bailey]

Agreed. Many of these testers strap the tube to operate as a diode and (usually) operate at a fairly low voltage compared with the operating voltages found in a tube amp, and almost impossible to derive any meaningful comparisons between tubes. That being said, if the tester shows zero emission or a short, the tube is most likely dead. In my own opinion, this type of device is like measuring muscular effort on a funfair grip strength tester. I have an acquaintance who has a large 'proper' industrial tester sitting on a dedicated cabinet full of punched cards. I asked him how often it got used (as it was covered in dust) and he said "Never". He went on to say if he plugs a tube into an amp and that tube works fine, that's all he needed to know.

[Mark37]

Thanks, this is really helpful to me. Before this, in my hobbyist experience, I have only had one tube failure, which was obviously due to vacuum loss--white cloudy appearance.

If I may summarize and possibly oversimplify, questionable tubes are best checked in a known good amp, questionable amps are best checked with known good tubes.

That being the case, I should order up some spare 6DX8 and verify them while my amp is +/- "known good."

But in this case the tube checker did help me see that the 6F4P had become faulty. The speed at which they did so was a surprise to this hobbyist/amateur, as was the 4 out of 4 failure rate.

[Mick Bailey]

I have a few bench sets of tubes for testing amps, and many boxes of known-good 'pulls' that I also use. Check by substitution is my primary method of eliminating tube issues. If I suspect a shorted power tube though I wouldn't plug it into another amp to avoid damage to that amp.