I would like help to work out a protocol for rapid stress testing and possibly bedding in a 24bit 384kbit R2R SS DAC.
In the rare event a DAC is going to fail I want it failing asap. I also want to test if playing in makes any difference to perceived sound for better or worse. Id appreciate this thread being a protocol thread. If you want to argue or speculate about the actual reality of burn-in please go to https://www.diyaudio.com/community/threads/burn-in.328225/page-9
Please share your ideas on how to maximise rapid burn-in:
1. Play for 11 hours on and 1 hour off for a good ten days.
2. Play a special white noise 24/384 file at -140dBFS to -0dBFS, doing 2Hz-96kHz, alternating with a 2Hz-96kHz sweep
3. use a low value resistor as a dummy load on the output, to stress the output stage and get the current flowing.
4. A/B with a known stable DAC before and after. Ideally have two of the same DACs, burn only one in and listen blindfolded and A/B before and after.
Thanks.
In the rare event a DAC is going to fail I want it failing asap. I also want to test if playing in makes any difference to perceived sound for better or worse. Id appreciate this thread being a protocol thread. If you want to argue or speculate about the actual reality of burn-in please go to https://www.diyaudio.com/community/threads/burn-in.328225/page-9
Please share your ideas on how to maximise rapid burn-in:
- Is it good to have a power off time to form caps and do expansion-contraction stress testing?
- Is white noise with high frequencies going to speed up processing?
- To activate all components is it necessary to play files with full volume range covering all the bit depths eg white noise from -140 to 0dBFS?
- Will a low resistance dummy load appropriate stress the output stage?
1. Play for 11 hours on and 1 hour off for a good ten days.
2. Play a special white noise 24/384 file at -140dBFS to -0dBFS, doing 2Hz-96kHz, alternating with a 2Hz-96kHz sweep
3. use a low value resistor as a dummy load on the output, to stress the output stage and get the current flowing.
4. A/B with a known stable DAC before and after. Ideally have two of the same DACs, burn only one in and listen blindfolded and A/B before and after.
Thanks.
Firework
Adept of dac tWeaking I did not found a better method than let playing the cd player one hundred hours. I just listen to it one hour to have a first opinion, then I let it alone to play the last 99 h alone till a second opinion. Some caps need this time, some others not.
Adept of dac tWeaking I did not found a better method than let playing the cd player one hundred hours. I just listen to it one hour to have a first opinion, then I let it alone to play the last 99 h alone till a second opinion. Some caps need this time, some others not.
No such a thing as burn-in a DAC ; all the "techniques" mentioned will do nothing, waste of time.
And no need to go anywhere else, burn-in is the key HERE so I answer HERE.
Just use your DAC to play and enjoy Music, instead of chasing ghosts and fairies.
Just play" Baby Shark Dance " for 100 hours.
whatever DAC turns itself off
is the winner.
whatever DAC turns itself off
is the winner.
With regard to accelerating any early component failures which I believe is the real thrust of your comment, sustained elevated operating temperatures will often help to weed them out.
It will definitely sound better - expectation bias is a powerful effect. 😉
Topping D90 took about 3-days of running before the clocks fully stabilized. Not unusual for crystal clock phase noise to decrease with time powered-on. Another guy I know with good ears said his D90 took 4-days to fully settle. After he was sure that was as good as it was going to get, he returned it for a refund.
Also, generally speaking the less feedback used in the Vref regulator and the analog output stage, the more sensitive the sound tends to be to individual component settling in that particular circuitry. When there is a lot of feedback it largely tends to correct for variations and settling of most components in the loop.
Also, generally speaking the less feedback used in the Vref regulator and the analog output stage, the more sensitive the sound tends to be to individual component settling in that particular circuitry. When there is a lot of feedback it largely tends to correct for variations and settling of most components in the loop.
Last edited:
The pacemakers I used to work on were run at (130 +/- 5) degrees Celsius for some time, I think it was one week, before being shipped to hospitals. Actually my circuit was never taken into production, but that was the standard procedure. Mind you, there are headers that melt at 125 degrees Celsius and electrolytic capacitors may also not survive.
They must be doing something VERY well:
- 11 BILLION views.
- 62 MILLION subscribers
- 37 MILLION likes
As of:
That does not prove anything physical happened, at most that he believed it did.
Maybe so, do not doubt the test.
But if a guy´s body temperature even approaches 130C he has worse problems than a non working pacemaker. 🙂
When I worked in telecom manufacturing we had large ovens that held about a dozen 6ft racks of equipment each, these ovens ran at 50c for 24hrs with all the equipment inside powered up, if there was going to be a premature component failure it happened there. This stress test had no effect on the equipments ability to crunch numbers... which is essentially what central office equipment does 24/7/365 for it's whole life.
It won't make any audible difference because DSP doesn't work that way... it's perfect every time or it's broken.. no exceptions. But there is always a possibility for infant mortality with electronics so there is some merit to doing a bit of a torture test with new equipment. I suggest you turn it on and leave it on for as long as you think is necessary, you could also feed it a signal(noise, music, whatever) and check on it occassionally to verify it's still working but I don't see this adding any meaningful stress as this should still be well within it's performance envelope. Your stress test is all about heat and any potential failure related to heat.
Ah... it doesn't matter what is under the hood of a DAC.. how many bits it has and what the sampling frequency is, an audio DAC only produces output in the audible spectrum below 20khz. You can try injecting a 96khz signal but you won't see any output.
DO NOT under any circumstances connect a dummy load to the output of a DAC, it isn't designed to produce any current it's intended to feed a high impedance preamp input circuit which responds to voltage.. not current.
Expensive but it would prove all of the above.
the sound of the dac will not change unless you listening to it becaucse this is not an issue of electrons or current but one of psychoacoustics, free your mind and the dac might follow
The first time I saw a maximum operating temperature of 135 degrees Celsius in the specification list, I also wondered what it was good for - it won't even get that hot when you end up getting cooked by a cannibal. It only started to make sense when they explained it was for a burn-in test to eliminate early failures.
Last edited:
It's not a DSP but a multibit DAC. If the resistors age at different rates, an excessive burn-in test could very well worsen the DAC's non-linearity due to worsened mismatch - hopefully not to an audible extent. Besides, if the resistor network is driven from CMOS gates or flip-flops, the output resistance of the gates will worsen somewhat due to various MOS ageing effects, like hot carrier degradation and negative bias temperature instability.
Last edited:
Hi, for me it is not a surprise that new equipment needs some time to stabilize of whatever you may call it. I also noticed the past 40 years that DIY electronics need some time to perform best. What is "best" is open for debate but let's just say it is better and since time is a variable one can't control it will happen anyway 😀 so discussion, heated debate and "explanations" with a lot of trigger words are a waste of time. It is like discussing if water is really wet while standing with your feet in the water.
FWIW I did the AB test with 2 identical factory new DACs. It was clear that the one that had run some hours was the better choice. Wasting time and OCD like behavior on items that will happen anyway without any required effort should IMO be suppressed so I just leave the DAC on playing random music.
It is also no surprise that new electronics either fail shortly after first power on and then at EOL. This widely known fact is the reason for stress testing. So it is a good thing to test equipment and especially consumer stuff that is not meant to last longer than 2 years. A guy I know had a factory new Sony device that developed a defective display after just 5 hours. Planned obsolescence is laughed at as being non existing and a fairy tale but please do test for instance Braun/OralB devices or audio devices with a monochrome OLED display.
FWIW I did the AB test with 2 identical factory new DACs. It was clear that the one that had run some hours was the better choice. Wasting time and OCD like behavior on items that will happen anyway without any required effort should IMO be suppressed so I just leave the DAC on playing random music.
It is also no surprise that new electronics either fail shortly after first power on and then at EOL. This widely known fact is the reason for stress testing. So it is a good thing to test equipment and especially consumer stuff that is not meant to last longer than 2 years. A guy I know had a factory new Sony device that developed a defective display after just 5 hours. Planned obsolescence is laughed at as being non existing and a fairy tale but please do test for instance Braun/OralB devices or audio devices with a monochrome OLED display.
Last edited:
Regarding planned obsolence I have a Philips SL18 still working after daily use of over 30 years. It was one of the first compact fluorescent energy saving bulbs. I guess later Philips (and other manufacturers) decided that such long life is not good for business 🙂
http://www.lamptech.co.uk/Spec Sheets/D FLCi Philips SL18 Mk2 Prismatic.htm
http://www.lamptech.co.uk/Spec Sheets/D FLCi Philips SL18 Mk2 Prismatic.htm
https://en.wikipedia.org/wiki/Phoebus_cartel
Kazap, if you want to put so much effort in testing/burning in maybe your warranty/guarantee is not adequate enough?! Or did you experience unreliability? If so no protocol will prevent that as the device has "failure by design". Like ventilation slots in the upper cover of a discretely built R2R DAC.
Kazap, if you want to put so much effort in testing/burning in maybe your warranty/guarantee is not adequate enough?! Or did you experience unreliability? If so no protocol will prevent that as the device has "failure by design". Like ventilation slots in the upper cover of a discretely built R2R DAC.
Last edited:
I noticed changes in same songs over period of time in new devices, before i knew what burn in was. So yes, changes do happen, at least to my hearing. Dac is a device that can be on 24/7h, just let some signal trough.
Absolutely the greatest 'burn in' change is that of the brain getting used to the different sound of new equipment or source material. I usually find that a brand new CD sounds fantastic, but as the brain learns these new songs and sounds, the sparkle disappears. Similarly with lovingly-built DIY speakers.
I'm not sure i agree on this 100%. Yes, psihoacoustics are there as a factor, but changes at least to me are not linear which is the curve you are explaining. Yes we get accustomed to things rather quickly.
As for speakers, it's not much of a burn in, as it is break in, surrounds, as well cones made of paper and other slightly deformative materials need time to break in and thats a known fact.
After decades the pattern still is the same: items like these are often denied by naysayers that haven't put energy in even one test. After experiencing powerlessness/incomprehension denial is simplest and the last thing that is left then is ridiculing, opening different subjects, wrong comparisons etc. Of course that won't be of any influence to those that simply accept self experienced items. For a long time I see green and grey mixed up and call certain grey tones green and vice versa. Having been corrected many times I reply that for me their dark grey is my dark green and vice versa and nothing will change that. Now try religion to an atheist......
But when discussing loudspeakers ... most will agree that suspension of the units is stiff at first and will become more flexible because of use.....
But when discussing loudspeakers ... most will agree that suspension of the units is stiff at first and will become more flexible because of use.....
Last edited: