Stress testing the power amp

[GainFreak]

Hi,

You've built a tube power amp, everything checks OK however you would like to stress test your build and all it's components. How do you do that?
1/ Do you feed it a specified frequency sine wave at full power for certain period of time?
2/ Do you feed it a square wave?
3/ Do you use pink/white noise and at what levels?
4/ Do you reamp with certain track/s or what?
I'm assuming the amp is connected to a dummy load.
So what would you recommend for a "stress test" and for how long?

[eschertron]

Throw it in the back of a station wagon, drive across railroad tracks all afternoon, then take it apart and look for loose wires and broken solder traces.

The only other stress test I can think of is to pour a pint glass of beer onto it. Oh, wait. That would stress test ME.

edit: but seriously - a full power timed test is a good idea. Probably the standard is a sine wave at just clipping; measuring power output is easiest that way. for max clipping I'm not sure if sine wave vs pink noise would make that much difference. Me, I'd think pink/white noise would approximate the strummed guitar better, but it's all distortion product on the output by then anyway, innit?

[J M Fahey]

The standard burn in test "before packing and delivering any amp" is to leave it at least overnight on a shelf (allow adequate ventilation) blasting full power.

Sinewave at clipping or equivalent level pink noise as suggested above is perfect , I personally play AC/DC discography (no kidding) full blast and set to loop and repeat so it does not stop at, say, 5 a.m. or something.

Mind you, next day you may need cotton gloves to handle the chassis, it can get that hot.

This applies both to tube or SS transistors but the tube ones have the extra step of looking at tubes themselves from close up, preferrably under dimmed room light, to check that screens do not glow white hot in rhythm with the music.

Redplating is bad, faint blue glow nothing to worry about.

[GainFreak]

I don't feel comfortable leaving the amp while I'm not around it but maybe several hours during the day will be OK.
I'm having second thoughts about the pink noise as it contains all frequencies from 20hz to 20kHz but in a guitar amp we don't need to go that low (or at least at full power) since most guitar OTs are cut below 50-70Hz.

[Chuck H]

The standard burn in test "before packing and delivering any amp" is to leave it at least overnight on a shelf (allow adequate ventilation) blasting full power.

Sinewave at clipping or equivalent level pink noise as suggested above is perfect , I personally play AC/DC discography (no kidding) full blast and set to loop and repeat so it does not stop at, say, 5 a.m. or something.

Mind you, next day you may need cotton gloves to handle the chassis, it can get that hot.

This applies both to tube or SS transistors but the tube ones have the extra step of looking at tubes themselves from close up, preferrably under dimmed room light, to check that screens do not glow white hot in rhythm with the music.

Redplating is bad, faint blue glow nothing to worry about.

Holy hell! That's some kind of a test for every amp you ship! One thing I totally agree with is the use of a music program. It gives the amp a breather in between peaks. A continuous sine wave is only one frequency so that seems lacking and continuous white noise would constitute continuous power and NOT musical power, which SHOULD take into account the notion that the amp will be used dynamically and not as a siren.

[GainFreak]

That's some kindl of a test for every amp you ship!

It's doable if you make couple of amps a month but I'm wondering how they do it on a larger scale in a factory.
Concerning the musical program or a guitar track it's quite different if you're playing blues compared to a high gain stuff where you are closer to a square wave.

[Chuck H]

Agreed. But for the purposes of amplifier testing I would go with max power only. Not "cranked" into output device square wave mayhem. That's a different story. Of course you need to test for both if that's how the amp will be used. I really doubt any amp should be expected to withstand clipping it's power devices into a square wave with a white noise input for twenty four hours. Not that some won't. In fact mine probably will. I'm not going to put it to the test, but I've seen my amps cranked into heavy output clipping for a couple of hours at a time without any problems. That was still a music program though. With peaks and lulls. Testing for that sort of abuse is unprecedented. It's up to you I suppose.

[Enzo]

AMps are not designed to put out sine waves at full tilt, they ARE designed to put music out at full tilt.

If you rebuild a car engine, do you rev it up to the red line and put a brick on the throttle pedal?

I don't see an amp test as running it to its limits to see if it cracks. I apply a music signal and run it at about 3/4 power.