Linear DA would not create harmonic distortion, but non-linear DA would. Has this been already brought up, or is there maybe a study even?
At least for electrolytic capacitors I would not assume DA being strictly linear/stable/time-invariant (linear resistors and linear capacitors in the equivalent circuit) given their actual construction.
Basically, increased DA for e-caps means the electrolyte is not conducting equally well everywhere, up to the point where it starts to dry up partially.... which is why increased DA appears to be a better indicator of capacitor health as it rises much earlier than bulk ESR.
At least for electrolytic capacitors I would not assume DA being strictly linear/stable/time-invariant (linear resistors and linear capacitors in the equivalent circuit) given their actual construction.
Basically, increased DA for e-caps means the electrolyte is not conducting equally well everywhere, up to the point where it starts to dry up partially.... which is why increased DA appears to be a better indicator of capacitor health as it rises much earlier than bulk ESR.
Agree - SID is a solved problem, but I still see power amps with undegenerated LTP’s and folks claiming dominant pole comp’d amps are slow, and feedback goes ‘round and round’ 😳🤦♂️
You can say ‘I hear something’ (hell, we all hear stuff) but you can’t then extrapolate that and make a general engineering/science claim unless it is supported by hard evidence. That’s all the engineering side is saying. The two best ways to do that at present are measurements and DBT. If those methodologies are rejected, there’s no factual basis for the general claim (only for your personal experience) and what you then have is opinion and you have to recognize it as such. You are entitled to that, but it remains a non-factual claim.
Last edited:
Could be said to be an illusion - one that might or might not be a reflection of reality. It may differ between individuals. Sometimes you hear things that don't exits in reality.
I measure - is most probably much closer to physical reality - if not equal. But...
Don't immediately exchange one for the other - there are ions between the two.
However - I don't think anyone objects to your experienced illusion - that is most probably very true.
//
We have been through all these arguments before here in the forum. When people make arguments like the above quote and like Tom is making, they show a lack of understanding of the history of science, and of observational science. I have explained exactly why before, so there should be no need for me to do it again.
Moreover if I take the bait from you and from Tom then I will be blamed for making a diversion, when actually you two are the ones inviting diversion.
Very well said Bonsai, in both your above posts.
Changing memory is responsible for most component "burn-in" claims. As humans, we do not have hard, locked in and error corrected memories. We do not have built-in references. That is why we use test equipment for starters, and why we designed that equipment to have built-in references and record the data. We can only compare, and even that is suspect due to our wonderful minds that affect our perceptions.
We run into trouble when we reject the validity of the one thing we created to record and show us the truth, and instead depend on the most variable thing known ... ourselves.
Changing memory is responsible for most component "burn-in" claims. As humans, we do not have hard, locked in and error corrected memories. We do not have built-in references. That is why we use test equipment for starters, and why we designed that equipment to have built-in references and record the data. We can only compare, and even that is suspect due to our wonderful minds that affect our perceptions.
We run into trouble when we reject the validity of the one thing we created to record and show us the truth, and instead depend on the most variable thing known ... ourselves.
A few years ago there was a set of tests done to test the proposition that amplifiers components 'burn-in'. The stuff was measured before and after and no changes were measurable. In another test, they looked at just a woofer to see if the rubber surround changed significantly in the first few weeks of use, but I can't remember the result of that unfortunately. The 'burn-in' thing is recycled ad infinitum and its a staple of just about any hifi magazine you care to name.
It may be that what's actually happening in most cases (I won't say all) is that it's the ear<>brain system that is tuning into the new sound. At first, the setup may sound a bit jarring and not what you are used to, but after a few weeks, your brain warms to it and suddenly it sounds great. A bit like that girl in class who used to wear thick horn-rimmed glasses and then one day she took them off and all the boys said 'Wow!'
I guess Tom will have more on the psychology bit.
It may be that what's actually happening in most cases (I won't say all) is that it's the ear<>brain system that is tuning into the new sound. At first, the setup may sound a bit jarring and not what you are used to, but after a few weeks, your brain warms to it and suddenly it sounds great. A bit like that girl in class who used to wear thick horn-rimmed glasses and then one day she took them off and all the boys said 'Wow!'
I guess Tom will have more on the psychology bit.
As always interesting discussions by esteemed members 🙂
I always get a laugh about folks going on about capacitor burn-in wrt sound. Know of any cap mfr datasheet or even system mfr saying anything about cap burn-in? Sure they say things about leakage current but does that have an effect on sound quality or distortion?
From my experience, Burn-in is a testing criteria wrt infant morality or wear-in phase. We used HALT and HAS testing criteria in the past for system testing
I always get a laugh about folks going on about capacitor burn-in wrt sound. Know of any cap mfr datasheet or even system mfr saying anything about cap burn-in? Sure they say things about leakage current but does that have an effect on sound quality or distortion?
From my experience, Burn-in is a testing criteria wrt infant morality or wear-in phase. We used HALT and HAS testing criteria in the past for system testing
Last edited:
New parts do drift in value a little when first put into operation. It also depends on stresses and ambient temperature. However, everything is pretty stable these days. Test equipment, for example, can hold calibration for decades. My HP 34401A meters have never been out of specification, nor my 974A meters. The readings will drift slightly with temperature and time. However, these drifts are way less than characteristics needed for audio equipment.
Audio performance drifts with temperature too. Leaving it on does not help one bit, the changing loads will shift temperatures. You don't hear it but it is easily measured.
Audio performance drifts with temperature too. Leaving it on does not help one bit, the changing loads will shift temperatures. You don't hear it but it is easily measured.
One has to wonder how long current mfrs run test cycles. Is burn-in a thing of the past?
Test = time= $
I would think not very long, definitely not very much along the bathtub curve. I run my tests on new assemblies for at least 12 hours but I can afford to do it whereas a large mfr can not afford to do so for obvious reasons. Even if I measure thd on a new amp, the specs do not change very much after a burn-in cycle
Test = time= $
I would think not very long, definitely not very much along the bathtub curve. I run my tests on new assemblies for at least 12 hours but I can afford to do it whereas a large mfr can not afford to do so for obvious reasons. Even if I measure thd on a new amp, the specs do not change very much after a burn-in cycle
Last edited:
It should also be noted that drift over time correlates closely with the overall quality of the component. A large part of the extra expense of top-tier components is in the R&D and QC procedures in place to guarantee repeatable and stable performance. And, of course, by "top-tier" I do not mean boutique. As @anatech mentioned above, most components will have various specifications drift over time - the most initially - but the amount of change is much smaller today than historically - even with middle of the road components. That stability does not, however, often apply to components made of Chinesium. As with most things in life you tend get what you pay for. Again, though, that does not really apply to boutique parts - that tends to be the opposite.
Hal
This is an interesting thread .... when it sticks to presenting objective analysis , data with context, and linkable references. Expert opinion is also useful, when presented as expert opinion and not as anything more. Too many good threads are being wrecked by bickering about anecdotal stuff. The AFOM thread is another that died from this stuff.
end of rant
end of rant
Hi Rick,
Component manufacturers do accelerated life tests, the good ones anyway. These are done on samples from production and a reject can kill the batch.
Completed consumer equipment is run, possibly the odd sample. But in contract manufacturing it is checked for function. Passed? Out the door. They do not do any life or reliability testing at all. No one is responsible for the design or equipment except the buyer (and that is whatever nameplate is stuck on). Then the end user owns it along with any fault it may have.
For industrial-commercial, medical or military use you do see responsible testing. But these days we also see "we'll fix it in the field" far too often.
Component manufacturers do accelerated life tests, the good ones anyway. These are done on samples from production and a reject can kill the batch.
Completed consumer equipment is run, possibly the odd sample. But in contract manufacturing it is checked for function. Passed? Out the door. They do not do any life or reliability testing at all. No one is responsible for the design or equipment except the buyer (and that is whatever nameplate is stuck on). Then the end user owns it along with any fault it may have.
For industrial-commercial, medical or military use you do see responsible testing. But these days we also see "we'll fix it in the field" far too often.
In the end though it really doesn't matter how good and how well tested the components are if the base design is poor. Every day I see items that cost well into the 5 digits and made with the absolute best components... that don't work for spit because the design is bad. There is so much FUD thrown around on these forums regarding parts and the effects - or lack of - they have and far too infrequently does anyone consider that the design itself may just be a steaming pile.
Exactly. Design and reasonable part quality go hand in hand. I'm talking about real part quality, not "audiophile approved" junk.