Burn in for fresh builds?

Jump to Latest
#161
Markw4 said:
Phase noise.
Click to expand...
Fascinating.

I was the resident expert on crystal oscillators when I left TI. I spent seven years of my career there designing crystal oscillators for use in clock jitter cleaners. I designed the XO in the LMK040xx and LMK048xx series to name a few. I've spent many, many hours in front of the Agilent E5052B measuring phase noise and jitter. We're talking phase noise well below -160 dBc/Hz at 12.288 MHz and integrated jitter around 200 fs (100 Hz - 12 kHz BW if I recall correctly). The performance was industry-leading at the time and might still be.

Some of the tests would run for days. We had one or two tests that would run for weeks. In total our thorough test and characterization lasted a couple of months and I would get to test the XO at various times during that time. The testing was done at extreme temperatures (-40 to +125 ºC) and also at temperatures in between, including room temp, of course.
It was also common that I'd need to measure phase noise of a board that had spent a few months in storage.
I have never seen the phase noise or jitter performance change appreciably from test to test. What I saw was measurement-to-measurement repeatability. A part that measured, say, 210 fs on day one would measure 210 fs (±measurement tolerance) a year later.

If the phase noise changed significantly over time our cellphone network would break down frequently due to interference between adjacent channels. GPS would be grossly inaccurate. And that's not the case, is it?

Tom
 
  • Like
Reactions: Speedysteve7 and mdsimon2
#162
classicalfan said:
Audio burn-in, often referred to as the process by which audio equipment undergoes a series of played sounds to achieve optimal performance...
Click to expand...
This part is silly, IMHO. Its not a series of played sounds that does. It just time under actual operating conditions; sometimes is just power-on time. There is no voodoo to the sounds played. Only exception to that was one experienced designer told me playing pink noise works fine.
 
#163
Markw4 said:
It is that too. Accommodation to a particular sound can happen pretty quickly, IME. However, I believe a clever engineer can figure out how to disentangle the two effects, listener accommodation, and burn-in. Of course the problem with some of the so-called scientific studies is that the so-called scientists don't know how to measure real burn-in, and they they don't know how find skilled listeners, nor do they try to train people off the street as to what to listen for. Nor do they test people on a system that the listener is already very familiar with. There are many problems with studies that create an appearance of doing science without being actual science. This is an old problem, and one according to the philosphy of science is something that should work itself out over time. The bad science should eventually be discovered and corrected. That's because science is a process, not a particular result.
Click to expand...
Wow. So basically if the scientific result does not support your personal opinion, it must be junk science performed by "so-called scientists".

There are many clever engineers on this forum. We're often shouted down and dismissed as naysayers or ignoramus when we suggest that the scientific method be applied in audio.

Tom
 
  • Like
Reactions: Speedysteve7 and classicalfan
#164
Tom, I said phase noise because that's what Andrea told me. Recently my views on the subject have changed to account for some observations. A brief explanation of my current view can be found at: https://www.diyaudio.com/community/threads/general-purpose-dac-clock-board.413001/post-7753751 (item #5). I didn't say anything about it in that post, but the observation there is also what I suspect may be involved with the audible effects of ferrites on clock sound even though ferrites appear to reduce measured noise (according typical noise measurement techniques which omit phase information of the noise -- a potentially significant factor according to a modern understanding of random noise).
 
Last edited:
#165
Markw4 said:
This part is silly, IMHO. Its not a series of played sounds that does. It just time under actual operating conditions; sometimes is just power-on time. There is no voodoo to the sounds played.
Click to expand...
Mark Audio would disagree with you. They've argued several times that you must follow a prescribed sequence of loudness levels and music genres to break in their drivers. And who am I to argue with the manufacturer, right?

So who's right? You or the manufacturer who claims a certain program must be followed?

Tom
 
#173
"I guess gravity is not a thing then. After all, the replication crisis is real."

We certainly have no crisis of shortage of snarky remarks.
By the way, I am a physicist and electrical engineer.

The current best theory of gravity (general relativity) is known to be wrong,
and to be only an effective theory, since it is not compatible with quantum mechanics.
 
#174
100kHz? Okay. Here is a plot of my 24MHz clock oscillator in use now with my dsd only dac:

1723059342445.png


However as can be seen, down at .1Hz offset frequency, it is at -70dBc. What some people believe matters a lot for audio dacs is around 1Hz - 10Hz offset. Far-out phase noise (e.g. 100kHz, mostly just affects noise floor, whereas close-in phase noise is more convolved with -- or multiplied by in the time domain view -- with our audio band frequencies).
 
Last edited:
#175
[RANT-ON]

It seems that here we are again, dozens and dozens of posts in continuous opposition to those who do not think the same way, and to think that as I've already said I'm someone who is convinced that I've perceived improvements thanks to the burn-in, but if someone else does not agree I do not want to convince them otherwise at all costs: this is no longer an exchange of opinions, it is a kind of holy-war that involves aspects that have nothing to do with the audio or the title of this thread, and in my humble opinion they are not even interesting to me.

And please note that I know that I might not read certain comments, but I think also that overposting involves the whole thread in any case and at least it makes me lose the desire to converse with someone else.

I'd have gladly replied to @tombo56 who replied to me about a certain topic, but there is too much overexposure in my opinion of certain positions that we have read at least ten thousand times, and they are always the same.

One should come to terms with the fact that one has no obligation to convince anyone of a different positions and that overposting in continuous opposition is not a good idea. IMNSHO

[/RANT-ON]
 
Last edited:
  • Like
Reactions: rayma
You must log in or register to reply here.
Top Bottom