Tech question - testing procedure

[wayward bluesman]

For all of you experienced techs, out of curiosity, how do you typically go about testing a piece of equipment after you've finished the repair? And for how long?

I understand the answer will always depend on the type of unit and what the problem was, but one of my biggest fears is sending a piece of equipment back out to the customer and, after having tested fine in the shop, have it fail the first time he uses it.

[52 Bill]

I don't think that there are any hard and fast rules for that. It is a great question though.

[Enzo]

First the equilibrium dynamics must be calibrated to the bandwidth as well as the soundstaging. The 3db down points at either end of the response curve must be...


Oh crap, I haven;t had lunch yet, and I really can;t come up with any good bullshit.


It's a good question, and points for thinking there is more to it than just fixing the fault.

Really, you need to decide what the thing is and what it ought to be able to do. For me, this has become a sort of zen thing, I don;t really think about it. WHatever it is, does it power up? Doesn;t stink? No lights where ther ought not be any?

Does it do the basic function? Does it well? DO the options work? In other words, if it is a guitar amp, does it make sound come out the speaker? I mean sound you'd want. Does it not make noise beyond what is expected? That is the sound you don;t want. DO all the inputs work? DO the controls control what they sposed'ta? Channels switch, FX work? FX sends and returns work? FS work? I know, they didn;t bring their FS, so you got something to try it with?

All working, sounds OK? If it is a mixer, this can be laborious, going into each channel, both line and mic, exercising all the EQ and AUX sends and returns, and channel assigns and subs, etc etc. If all the crap is OK< connect a load to the amplifier, or loads to the amplifiers. APply a test signal to the input, run it up to clip a moment, scoping the output. Look OK, waveform OK under load? CLip symmetrically, and reasonably close to the rails? If it has a limiter like the PV DDT, does that kick in?

A sine wave is really harder for an amp than music. So don;t send the amp to the burn in table with a huge sine wave at the edge of clip. OK briefly, but not really fair to the amp. HAlf power maybe for extended time. I prefer music, myself. I donl;t want to listen all afternoon to 100 watts of the radio or a CD, but into a load, you bet. Let it go for an hour or more at 2/3 of max. Nothing special about that amount, I just made it up. I am not trying to destroy the amp, it is not a stress test, I am trying to see if it works under good load, doing real work, for more than minutes.

Powr amps like for PA systems can have a lot more going on than a guitar amp. The QSC adjustment has things like current limits to set, and short protection. SOme amps want you to short dead across the output and make sure it thermals out within a specified time.


AMplifiers by their nature need to be exercised. But things like preamps or processors don't really work hard. An active crossover or a graphic EQ probably won;t be overheating and melting down. Signal circuits don;t really care much about signal level, in terms of stress. Let something sit there for half an hour turned on, and that ouoght to be enough generall.y to find out if an op amp is going noisy. That is enopugh that the internal temperature is probably up to as warm as it gets.

[jrfrond]

Everything that is serviced needs to be tested, especially high-power amps. In fact, if I had the room, I'd establish a whole separate QC department just for this purpose.

[Billy R aka DynaFreak]

One of my rules is that I always do a final test after the amp is COMPLETELY re-assembled, just like it is going back to the customer. It is too easy to knock something loose, forget to re-attach the reverb cables, etc., during final assembly. I don't want to get a bad rep for something stupid after doing a stellar job troubleshooting and fixing the amp. My load test procedure is basically the same as Enzo's.

[wayward bluesman]

+1 on testing after complete reassembly. And thank you guys for the replies. In my limited experience, power amps and mixers definitely take the longest to test, but testing time is usually proportional to the complexity of the original issue or repair time. Obviously I'm not going to spend 45 minutes testing an amp that only needed a five minute jack replacement. But amps that came in with intermittents or would only exhibit issues after they'd warmed up tend to make me nervous.

I once had Yorkville perform an in-house repair on one of their AP series amps, and their tech had mentioned they ran the amp for several hours after repairing one of the output boards. At the time, I thought this was a great practice for amp repairs of that sort, but how practical is it to run an amp at 3/4 output for hours at a time when you're in a shop with other people who need to be able hear themselves think? Do you run into a dummy load? Obviously this isn't always representative of an actual inductive, speaker load. But if you let it run for a couple of hours while you work on other repairs, how will you hear it fail or malfunction? Every once in a while I'll parallel a simple voltage divider with the dummy load and then run the attenuated signal into a cheap stereo speaker so I can hear something, but this is rarely representative of the actual signal I want to hear.

Like Enzo, I prefer actual music for final testing, oftentimes something classical because of the dynamic and frequency range. But do you guys typically run amps at 90 dB for literally hours before deeming them good to go?