Testing my Repairs with Burn-In Test

[Rocks]

Hi.
Read through some of the posts on the subject and can’t find what I’m looking for. I just repaired a Sony STR-V3. I changed all the caps and transistors on the output amp board. I used what I think are reasonable equivalent transistors. Planning to put it up for sale and I always do a « burn in « test to get some idea that it will be reliiable.

What do you use for burn in? Sine wave, pink noise, music? I like to use music since it is most représentative of end use. I use 8 ohm dummy loads and raise the volume until i start to see clipping on the peaks, then back it off slightly. Is this too much? Too Little? On this amp I start to get clipping at volume level of 7 on 10. Does this make sense? I’ll run it for 5 or 6 hours like this. Before I do all of this I check performance with actual speakers.

Sine wave is good because you can calculate output power more easily but I find it severe and not representative.

All opinions are appreciated.

 

[DatunWalnut]

That sounds excessive to me, though I imagine everyone will have a different opinion. The amps I work on are around 40- 50 years old so I don't want to over-stress them. For class AB amp, I usually use a 1khz sine into an 8 ohm dummy load. Music is too full of quiet parts so tends to not work the amp as hard.
After it's gently warmed up and played a bit of music, I run the amp at about 70% max power for a maximum of ten minutes. Monitoring all the time with my IR gun. I feel this is way more than any normal music use will ever stress it. For class A I tend to be kinder as it's already going full bore so I run 50% for five mins.

 

[Bill Ferris]

FWIW, After a serious rework/rebuild/overhaul, I first certify the units factory`s specs. sine wave THD+N/IMD, and if it passes, I then continue to run it into my L/R dummy loads around depending on its rated wattage with at first pink noise around 1/4 of it`s 8 ohm sine wave power rating while monitoring the output transistor`s heatsinks & power transformers temperature as an initial stress test for maybe an hr. or more with the covers off.

If the unit still acts stable, I then feed it music at the same average output level for several days over night 48 hrs. continuous for further covers on thermal stress testing, then rerun the initial bench performance testing outlined above, if it still passes, I then do a shop speaker hook up listening test, and if it sounds good to my ears, then I just wait for the customer to show up and we take the audio device into my sunroom to connect it to the separates based HQ stereo system out there for their sonic evaluation/approval of whatever the audio device type is, be it a stand alone power amp, preamp, integrated amp, or the very rarely taken in receiver.

 

[Rocks]

FWIW, After a serious rework/rebuild/overhaul, I first certify the units factory`s specs. sine wave THD+N/IMD, and if it passes, I then continue to run it into my L/R dummy loads around depending on its rated wattage with at first pink noise around 1/4 of it`s 8 ohm sine wave power rating while monitoring the output transistor`s heatsinks & power transformers temperature as an initial stress test for maybe an hr. or more with the covers off.

If the unit still acts stable, I then feed it music at the same average output level for several days over night 48 hrs. continuous for further covers on thermal stress testing, then rerun the initial bench performance testing outlined above, if it still passes, I then do a shop speaker hook up listening test, and if it sounds good to my ears, then I just wait for the customer to show up and we take the audio device into my sunroom to connect it to the separates based HQ stereo system out there for their sonic evaluation/approval of whatever the audio device type is, be it a stand alone power amp, preamp, integrated amp, or the very rarely taken in receiver.

Thanks for the feedback. I’m not at your level yet. No distortion measurement capability. I just have a two channel scope.

 

[gammax]

FWIW, After a serious rework/rebuild/overhaul, I first certify the units factory`s specs. sine wave THD+N/IMD, and if it passes, I then continue to run it into my L/R dummy loads around depending on its rated wattage with at first pink noise around 1/4 of it`s 8 ohm sine wave power rating while monitoring the output transistor`s heatsinks & power transformers temperature as an initial stress test for maybe an hr. or more with the covers off.

If the unit still acts stable, I then feed it music at the same average output level for several days over night 48 hrs. continuous for further covers on thermal stress testing, then rerun the initial bench performance testing outlined above, if it still passes, I then do a shop speaker hook up listening test, and if it sounds good to my ears, then I just wait for the customer to show up and we take the audio device into my sunroom to connect it to the separates based HQ stereo system out there for their sonic evaluation/approval of whatever the audio device type is, be it a stand alone power amp, preamp, integrated amp, or the very rarely taken in receiver.

Excellent test and response

 

[RVT1K]

I've never worked on home audio for paying customers but, based on my varied electronics troubleshooting/repairing/testing experience, I would consider feeding the amp in question an input that sweeps between 20Hz and 20kHz.

 

[Bill Ferris]

I've never worked on home audio for paying customers but, based on my varied electronics troubleshooting/repairing/testing experience, I would consider feeding the amp in question an input that sweeps between 20Hz and 20kHz.

That is usually part of the bench check out, manually swept (no computer based sweep automation capability), after zero reference setting my Tektronics audio analyzer to 0 DB, at first 1 watt power level, one channel at a time, then the same at it`s 8 ohm rated power level each channel.

Sometimes I have just select various spot frequencies via button punch ups from my Tek. sinewave generator watching for any significate changes in output level.

My IVEY brand class 1 IE-30 RTA set to its 1 DB display resolution feeding pink noise into the DUT can be, & has been used, especially to check any tone control, and or loudness switch controls for proper operation, if preamp, or mixer, receiver, or integrated is being tested.

Though if pressed for time, I have in the past just hooked up to shop`s test speakers with music fed to it and listened to the tone control, and or loudness`s properties while engaging them, and or varying the tone controls throughout their full range as a quick check.

 

[RVT1K]

That is usually part of the bench check out, manually swept (no computer based sweep automation capability), after zero reference setting my Tektronics audio analyzer to 0 DB, at first 1 watt power level, one channel at a time, then the same at it`s 8 ohm rated power level each channel.

Sometimes I have just select various spot frequencies via button punch ups from my Tek. sinewave generator watching for any significate changes in output level.

My IVEY brand class 1 IE-30 RTA set to its 1 DB display resolution feeding pink noise into the DUT can be, & has been used, especially to check any tone control, and or loudness switch controls for proper operation, if preamp, or mixer, receiver, or integrated is being tested.

Though if pressed for time, I have in the past just hooked up to shop`s test speakers with music fed to it and listened to the tone control, and or loudness`s properties while engaging them, and or varying the tone controls throughout their full range as a quick check.

That certainly sounds like a well thought-out and thorough plan.

My comments were based mostly on my experience where anything that was required to work over a range of frequencies was tested over that range of frequencies. But I also recall just doing spot checks over the range as well...low end-mid-high end.

 

[painttoad]

Music is more demanding than sine waves. When I test an amplifier, i beat the hell out of it. On dummy loads and speakers. Sine waves and music. Steady tones and sweeps. I measure output power and distortion and see what is going on at different freq. I am extremely hard on amps. I listen at high levels. I want to know that the amp can put up with my abuse before either returning them or selling them. It's not likely that anyone buying from me or having me repair their gear will roast an amp under heavy use after I'm done.

 

[tadawson]

My initial burn-in test of pro power amps I have repaired was to run tone to 100% power to verify full power out for a few minutes, and then the rather heavily compressed first CD by Madonna into it (this was in the 80's, and typical of what they would be playing when returned), both channels driven into resistive loads (6 ohm as I recall) for at least 12 hours continuous, at clipping.

If it survived that, then I'd check the details.

(I still do a burn in, but the program has changed . . . )

 

[Rocks]

Thanks for all the replies. Didn’t realize I was getting them. Some good ideas. Currently running Spotify for 6-8 hours at about 70% on the volume dial. I think I’ll add a short test at 1kHz at maximum power.

 

[DatunWalnut]

Your Spotify at 70% will certainly show most problems. And the full power for two or three mins should cement the deal. Personally, I've gotten sick of my music collection through using it to shake down amps in the background when working on others. To that end I've bought a Bluetooth DAC so that I can listen to podcasts and save my music for leisure.

 

[restorer-john]

@Rocks, You are on the right track. I do a full 8R and 4R 20-20k test and a five (5) minute 1kHz 8R at FRP (full rated power). Obviously dummy loads are required, but see below, they can be derated for the 24 hour test.

For burn-in/soak for restored gear, I do a EIA/EIAJ dynamic power test for 24 hours. The beauty of that test (in my opinion), is the amplifier is pushed to full power for 24 hours, but with a duty cycle that doesn't overheat anything competently designed. Duty cycle is such that vintage gear is perfectly happy and sees it as the equivalent to a robust weekend party with a lot of bourbon, some AC/DC, and maybe some ZZTop. /s

It's really easy with a standard industry CD player test disk (with EIA/EIAJ) toneburst tracks on repeat on a CD player.

Anyone who wants those files, PM me and I'll upload them to google drive.

You could also do it with an arbitrary function gen, if that gen has enough memory to create the waveform (as per the EIA/EIAJ/JIS specs) you want AND the ability to loop the arb file without delay (mine doesn't- it has a delay to 'repeat'). A CD player is instant on repeat.

You could potentially create a continuous 12/24 hour .wav file, but that's a whole lot of hardware running for a day when a CD player does it out the box...

 

[tadawson]

Hi.
Read through some of the posts on the subject and can’t find what I’m looking for. I just repaired a Sony STR-V3. I changed all the caps and transistors on the output amp board. I used what I think are reasonable equivalent transistors. Planning to put it up for sale and I always do a « burn in « test to get some idea that it will be reliiable.

What do you use for burn in? Sine wave, pink noise, music? I like to use music since it is most représentative of end use. I use 8 ohm dummy loads and raise the volume until i start to see clipping on the peaks, then back it off slightly. Is this too much? Too Little? On this amp I start to get clipping at volume level of 7 on 10. Does this make sense? I’ll run it for 5 or 6 hours like this. Before I do all of this I check performance with actual speakers.

Sine wave is good because you can calculate output power more easily but I find it severe and not representative.

All opinions are appreciated.

A little light compared to my burn in. I tend to use a pretty "thumpy" CD on loop, driven to the threshold of clip (with intermittent peaking still visible into dummy loads for 8 hours minimum. Never had a come-back . . .

 

[Bill Ferris]

A little light compared to my burn in. I tend to use a pretty "thumpy" CD on loop, driven to the threshold of clip (with intermittent peaking still visible into dummy loads for 8 hours minimum. Never had a come-back . . .

Especially important stress testing for the local rock bands multi-hundreds of watts per channel PA power amps, where on stage reliability is very, very important after repairing/rebuilding the PA amp`s output section !

High signal input levels of Pink noise with its full bandwidth fluctuating random with it`s fairly high crest factor signal quickly stresses the output transistors, promptly routing out any boarder line thermally sensitive weak ones that`s best discovered on the bench, along with close multi strapped output transistors even load sharing.

My very helpful B & K transistor checker, while able to test all the tested transistor up to 150 collector volts which is very important for weeding out voltage sensitive leakers, but some prior stressed output transistors become unhappily leaky problematic when elevated in real world operating temperatures carrying high C/E currents.

Doing ever increasing output wattage PA power amps for 5 local music stores workload in the early nineties pushed me to purchase the Sencore brand supplemental PA-82 mate to their PA-81 250 WPC audio analyzer to the uprated 2,500 WPC dummy load capacity, so I wouldn`t stress the lower 250 WPC capacity while burning in higher wattage power amp rework/rebuilds.

Both Sencore brand model dummy loads have a thermostatically controlled fan for each channel`s internal resistors load bank, plus a global thermal cutout to protect the units if over heating either of the load resistor banks, also if the amp under test develops a 1 volt or higher DC output offset, they promptly internal relay disconnect themselves from the amp being tested.

Both purchases were very pricy, but greatly sped up my repair work/test time, so in short order paid for themselves.

 

[tadawson]

Especially important stress testing for the local rock bands multi-hundreds of watts per channel PA power amps, where on stage reliability is very, very important after repairing/rebuilding the PA amp`s output section !

High signal input levels of Pink noise with its full bandwidth fluctuating random with it`s fairly high crest factor signal quickly stresses the output transistors, promptly routing out any boarder line thermally sensitive weak ones that`s best discovered on the bench, along with close multi strapped output transistors even load sharing.

My very helpful B & K transistor checker, while able to test all the tested transistor up to 150 collector volts which is very important for weeding out voltage sensitive leakers, but some prior stressed output transistors become unhappily leaky problematic when elevated in real world operating temperatures carrying high C/E currents.

Doing ever increasing output wattage PA power amps for 5 local music stores workload in the early nineties pushed me to purchase the Sencore brand supplemental PA-82 mate to their PA-81 250 WPC audio analyzer to the uprated 2,500 WPC dummy load capacity, so I wouldn`t stress the lower 250 WPC capacity while burning in higher wattage power amp rework/rebuilds.

Both Sencore brand model dummy loads have a thermostatically controlled fan for each channel`s internal resistors load bank, plus a global thermal cutout to protect the units if over heating either of the load resistor banks, also if the amp under test develops a 1 volt or higher DC output offset, they promptly internal relay disconnect themselves from the amp being tested.

Both purchases were very pricy, but greatly sped up my repair work/test time, so in short order paid for themselves.

My most memorable ones were back in the 80's . . . Tapco CP-500's for an Indiana disco/dive bar . . . The biggest question was whether to hand clean, or basically boil it out to start . . (Main PCB's typically did a soak in the sink with soap). Cooling was the second worst enemy here, right after abuse. (Oh course, the one that came in for work after half the bar burned down next to/around it was a special kind of nasty. No fire damage, but I still shudder at the stink . . . )