Quite a set of claims:xiphmont said:
Brimming, teeming, overflowing with quacks, 99% of the people buy Auricaps for non sound quality reasons, knowledge that one item costs more convinces people it sounds better.
Certainly financial bias is an oft repeated refrain, always posted by those who demand documentable performance results. Yet I don't recall seeing a study that supports what you are quoted as saying above. Oh wait I forgot, the mere fact you object to subjectivists, allows you to make subjectivist style claims; personal beliefs presented as facts. Since I don't believe, I am therefore entitled to put forth wholly unsupported conclusions about other peoples decision making process.
One reason this forum exists is to help pursue or quantify what many audiophiles see as a disconnect between mass market performance and prices and the high end performance and prices. The gap (at least to me) initially seemed larger than common sense would imply. During my career as an EE I found most other EE's to reasonably competent. Sure there were those who were useless and a similar number of those who were brilliant. One would think that if the parts, of and by themselves, make no meaningful measurable difference and the design is at all competent the resulting product would sound fine.
As I said, an entire industry has sprung up to exploit (your frame of reference) this difference. One would assume that a competent engineer would quickly start a company to manufacture good sounding amplifiers say 100 watts for $200 retail. To the best of my knowledge this hasn't happened in the last 25 years, could it be that it's not possible? Or are you saying that a $6,000 Pass Labs amplifier sounds effectively identical to a 100 Watt $200 Pioneer amp?
If you honestly believe that last sentence, then we will have severe difficulty in reaching an agreement, because our starting positions are simply too far apart.
Due to my professional EE training I was a complete non-believer in this nonsense, particularly regarding cables. I knew for an absolute fact that the engineering science and math did not support those absurd claims about how wires sound. Right up the time my friend convinced me to take a pair home for a test drive. There were 5 people in the room: "What's that?", "Oh an overpriced piece of wire that's supposed to make my system sound better". "Let's try it!"
After a short interval, we all recovered from the clearly impossible improvement. Since then, I started paying real attention to component details. My posts her are simply intended for people to experiment for themselves. I was surprised, I think many others will be too.
For me the most relevant point (coming form the POV of a relative layman, but still, I have my view) is one that has been made several times here.. The differences in our hearing are far greater than the differences in circuits or components.. To me this means that ther can be no holy grail, just as there can be no universal "truth" in our subjective consciousness.
Any one for audiophole cranium corrective surgery to get rid of those resonances..???
Any one for audiophole cranium corrective surgery to get rid of those resonances..???
hermanv said:
You've gotten to the heart of it. You have it exactly. I'll go even farther. A $5 National chip sounds just as good as that Pass Labs amp (provided of course that it is applied competently).
We are too far apart and I politely invite you to disengage this argument and ignore me in the future, as I am obviously someone too deaf to be heeded.
Whoa there! None of that heresy around these parts!!!!xiphmont said:
Seriously, when applied within their sweet spots, I agree that most amps WILL sound exactly the same, whether AndrewT likes it or not! The differences in sound typically come at the extremes of the operating range--EG clipping or driving unusual loads, etc.
This goes out the window, of course, when we are talking about amps that intentionally introduce distortion (tube amps and the like), but that is an entirely different discussion.
"Just as good as" is a subjective term, wouldn't you agree? So your statement can't possibly be wrong (or falsifiable). Did you mean to say "indistinguishable from"? If so, I think I can convince you that you're wrong.
Is this based on listening tests or is it an opinion? I'm afraid I do not agree with your position.dfdye said:
What I mean is that when listening at "normal" levels, my quite expensive SS amplifier has a smoothness, a relaxed sound and a harmonic richness while it still appears to reproduce all high frequency information and impulses. I do not need to operate only at performance extremes for the amplifier to sound noticeably better than mass market equipment.
We have wandered a bit off topic, since I have no desire to Hijack the thread, I think I'll follow xiphmont's recommendation and quit this discussion.
xiphmont said:
The problem is that you would then be spreading misinformation unchallenged.
Here are some sites with articles and measurements.
http://waltjung.org/Classic_Articles.html
The classic Audio article that brought capacitors to general attention and the Audio Amateur article coauthored with John Curl with scope pictures.
From an IC mfr. , with curves.
http://www.maxim-ic.com/appnotes.cfm/appnote_number/3171
Note the THD.
Interesting test approach; see also the added sections.
http://members.aol.com/sbench102/caps.html
These tests show that the presupposition of linearity, and then arguing from that embedded assumption is specious. Note that hysteresis is very audible, out of all proportion to the measured THD; early transistor amplifiers demonstrated that all too well.
Subjective, correlates fairly well with my experience, to the extent that I’m acquainted with five or six of the tested models. As you’d expect in a competitive environment, quality roughly tracks price, (or vice versa) but with some exceptions. http://www.humblehomemadehifi.com/Cap.html
I've seen this lead to a self-reinforcing prophecy. A very poor system prevents any major improvement in any component from being truly heard; this is then offered as proof that (especially the more subtle) improvements are worthless.xiphmont said:
There are of course audio products that are sold (I believe) fraudulently; the Peter Belt products are the most extreme example. I continue to be resolutely agnostic with regard to power cords and cryogenic treatment.
The assertion that all tests needed to perfectly predict audible quality are known is an unprovable proposition.
And the idea that sound does not sell electronice equipment since it all sounds the same is silly. NAD has done well, in a very competitive environment as good sound/low cost products; other companies come to mind as well. If they don't have cost or obvious build quality to ratify them, they should have been out of business long ago if they were not competitive based on sound quality. Audio Research kept their selection of caps based on sonics something of a trade secret for a while. They were not able then to ride a false cap cachet to success; they sold on sound/dollar. Conversely I've heard expensive, strongly marketed gear that sounded lousy; and next year it was one with Nineveh and Tyre.
The notion that a substantial industry is supported soley by misled fools unheeding the unheard advice of their betters is not really credible.
xiphmont said:
I completely agree with this statement. Most of the people disagreeing are either the ones that will fear engaging themselves into a double blind test or the ones that won't trust the embarrassing results.
99% of the audio business is based on subjective non-demonstrable claims. That's human nature...
Quite right! There is another thread for the amp discussion, so let's limit this discussion to caps.hermanv said:
One quick response. I'll respond to the others later.
Please. I asked him a few times for data to back his claims, and his responses were "You're wrong because I know you're wrong."
You're arguing similar points, but you're also pulling out references and technical points for discussion. Huge difference. I'm happy to debate the value and content of the articles you mention.
...except that this is not showing a coupling application. It's showing a passive highpass where large voltage differences (well, we can't tell how large, let's assume line level) are appearing across the capacitor because the F3 is up high in the audible range. I should hope your hifi amp isn't highpassing at 100Hz.
They've also chosen to test X5R/X7R ceramics and tantalums, which I've already said are the two worst choices you can make to couple audio. Tantalum is such a bad choice you're risking damaging the capacitor because they're not designed for AC. Reverse biasing a tantalum causes it to break down chemically. Any capacitor test basing general conclusions on the results of reverse biasing a tantalum capacitor is a nothing more than a straw man (and yet-- the tantalums manage to do acceptably, at least for a little while until they start failing).
My test circuit is the same as the on in the articles but with the values chosen to couple not filter (10uF, 15kohm). Line level input. Like I said at the very beginning. When coupling, I can measure no THD down to -120dB from any film or electrolytic. Ceramic registers the tiniest blip at the very floor of detection; it's so low, it's hard to tell if it's a real reading or just occasional noise. When I turn on the noise floor sampling, I can't get any reading for X5R either.
[edit: I was using .1uF monolithic ceramics in the test which are obviously way smaller than the film caps. That should have degraded their performance, not improved it. Regardless-- no detectable THD]
Curmudgeon said:
Please. I asked him a few times for data to back his claims, and his responses were "You're wrong because I know you're wrong."
You're arguing similar points, but you're also pulling out references and technical points for discussion. Huge difference. I'm happy to debate the value and content of the articles you mention.
Curmudgeon said:
...except that this is not showing a coupling application. It's showing a passive highpass where large voltage differences (well, we can't tell how large, let's assume line level) are appearing across the capacitor because the F3 is up high in the audible range. I should hope your hifi amp isn't highpassing at 100Hz.
They've also chosen to test X5R/X7R ceramics and tantalums, which I've already said are the two worst choices you can make to couple audio. Tantalum is such a bad choice you're risking damaging the capacitor because they're not designed for AC. Reverse biasing a tantalum causes it to break down chemically. Any capacitor test basing general conclusions on the results of reverse biasing a tantalum capacitor is a nothing more than a straw man (and yet-- the tantalums manage to do acceptably, at least for a little while until they start failing).
My test circuit is the same as the on in the articles but with the values chosen to couple not filter (10uF, 15kohm). Line level input. Like I said at the very beginning. When coupling, I can measure no THD down to -120dB from any film or electrolytic. Ceramic registers the tiniest blip at the very floor of detection; it's so low, it's hard to tell if it's a real reading or just occasional noise. When I turn on the noise floor sampling, I can't get any reading for X5R either.
[edit: I was using .1uF monolithic ceramics in the test which are obviously way smaller than the film caps. That should have degraded their performance, not improved it. Regardless-- no detectable THD]
Curmudgeon said:
I read this one long ago. It's another straw man.
We're talking about coupling, a situation that places microvolts across the capacitor. Any nonlinearity of the capacitor is inconsequential in the context of the rest of the input impedence (unless we're talking about a very poor capacitor indeed). This test throws 70V across the cap in a test designed to show the capacitor's behavior in a situation that intentionally magnifies any defect and claims it means something in relation to audio.
My readings are being taken directly from the most dead-common audio application of a cap: line level coupling from a low impedence into a 10kohm input.
(That test is more relevant if you're talking about, eg, crossover design. Then we have a discussion about whether or not it matters in that context and you have a much better chance of winning the argument, or at least seeing the cap contribute in at least some detectable way)
SY said:
Well, yes, you tripped me up on a semantic wrinkle. Fair enough. And feel free to try to convince me wrong, so long as it's an argument we can test empirically. Especially if you're the one bringing the Pass Labs amp ;-)
[edit: aw drat, SY is left coast. I was hoping he was actually out here]
"The foil cap turns out to be coupling hum and hash noise coming from the PSU. "
Might it be that that cap is mounted with its outer electrode connected to the high-impedant point in the circuit? I would be surprised if the hum coupling wouldn't disappear or at least be reduced if you turned the cap around and resoldered it.
Might it be that that cap is mounted with its outer electrode connected to the high-impedant point in the circuit? I would be surprised if the hum coupling wouldn't disappear or at least be reduced if you turned the cap around and resoldered it.
Werner said:
Well, I did try that just to be complete. No difference whatsoever. The measurement was identical.
But why would that be expected to make any difference whatsoever? You'd just be reversing the polarity of the coupled noise voltage. Noise is noise in both directions 🙂
xiphont, although I'd be delighted to host you, I can convince you without it.
1. What is the source impedance of a $5 chip amp?
2. What is the source impedance of a First Watt Whatever?
3. Do speakers generally have flat impedance versus frequency?
I think you can guess question 4.😉
1. What is the source impedance of a $5 chip amp?
2. What is the source impedance of a First Watt Whatever?
3. Do speakers generally have flat impedance versus frequency?
I think you can guess question 4.😉
xiphmont said:
If you connect the outer foil to the lower impedance it acts as a shield, whereas then connected to the higher impedance of the circuit it acts as a garbage bin for radiated hum.
And yes, I have measured these differences. And yes, they were meaningful.
I've never heard of this before! Do you still have the data? I would greatly appreciate seeing it.Werner said:
Actually, I don't much care, x. Once I had satisfied myself that there were differences, and that I was taking some care to avoid preconceptions, I don't terribly care what the cause is. I'm interested in finding out, but I was educated in the sciences, so observation first, then hypothesis, then theory. There is a tendency to dismiss observations if there is no theory; most famously perhaps the Wegener deniers. I hear what I hear, others hear the same, and if you cannot, well, I suppose you'll just have to listen to the musick. 🙂
But I'm not at all comfortable with the need you seem to feel to correct others... mission from God? And I am concerned that people who can hear will be misled by either those supposedly all-powerful preconceptions, or whatever into never trying for themselves.
But I'm not at all comfortable with the need you seem to feel to correct others... mission from God? And I am concerned that people who can hear will be misled by either those supposedly all-powerful preconceptions, or whatever into never trying for themselves.
Unless the cap is in a situation where one terminal is grounded, i.e., has no signal on it, noise pickup won't be helped much by polarity. A properly used coupling cap will have identical voltages on both terminals, including the noise signal. My guess is, if you're going to shield a coupling cap, it would be best to put a low dielectric constant spacer around it, or maybe just a few bands of such, then a grounded copper foil shield or similar. Should give good noise performance and low stray capacitance, as opposed to just wrapping a shield directly around the cap.
As for the Pass vs. $5 chip, the results of blind testing to date say if the amps are operated within their capabilities, if the frequency response is the same, levels matched, etc., you shouldn't be able to tell the difference. My guess is that level matching alone wouldn't satisfy the necessary conditions- you'd probably need some equalization, and maybe need to look at damping factor as well. Then, a huge amount of evidence says they'd be indistinguishable.
As for the Pass vs. $5 chip, the results of blind testing to date say if the amps are operated within their capabilities, if the frequency response is the same, levels matched, etc., you shouldn't be able to tell the difference. My guess is that level matching alone wouldn't satisfy the necessary conditions- you'd probably need some equalization, and maybe need to look at damping factor as well. Then, a huge amount of evidence says they'd be indistinguishable.
SY said:
Are you trying to suggest the open loop drive impedence is somehow relevant? I appreciate the attempt at a trick question 🙂
Actually, I suspect I misunderstand where you're going. So, sadly, I can't guess #4.
Werner said:
Hm, no it doesn't work that way. the induced noise doesn't drain to only the low impedence side. You get a voltage across the component and the impedences matter in terms of the circuit design but it doesn't matter which side the 'shield' is on. It's only a shield if it's electrically uncoupled and connected to chassis ground.
I suspect the differences people notice have more to do with the physical orientation of the innards coupling to noise sources slightly differently in different positions (like pulling a car forward a few inches to get better radio reception at a stoplight, or turning a TV antenna to a different direction). I don't deny that you were able to measure a difference.
Curmudgeon said:
The people who uncritically believe what they're told aren't going to do any testing anyway, or they're going to do it the lazy way and be mislead by power of suggestion. Is it better they believe the quacks or believe the scientists and engineers?
Since we're being careful with semantics around SY, I couldn't be on a mission from God being an athiest. But yes, the same concept. The same mission as Carl Sagan and Mythbusters. I'm a professional in the field and irrational beliefs and outright misinformation have hurt my professional goals badly at times in the past. A single sentence from an 'authority' who is sincere but talking out his *** can ruin aspriations. I'm mostly bitter about patent law, but the Green Marker crowd have caused me professional pain in the pages of the NYT and WSJ.
So I've got a chip on my shoulder. Guilty as charged. I'm not ashamed of it. It's a valuable tool for improving the ststus quo. I do try not to be rude. I also own up whem I'm proven wrong. Sometimes I talk out my *** too. I don't do it on purpose. We're all human.
Conrad Hoffman said:
Re: shielding the cap. I agree wholeheartedly. A shield wrapped around the cap will itself couple with to the cap to some extent and you end up with a voltage being induced across the shield and the cap. It's a smaller voltage and preferable to having the noise voltage appear across the cap itself, but it doesn't eliminate the problem. A low-k spacer will help and having the spacer and the shield helps more (even air space between the shield and cap improves the shielding by dropping the coupled capacitace). Using a physically smaller cap helps more yet. Eliminating the cap entirely is the only way to kill it off completely of course, but that's a pretty deep design constraint and probably thoroughly unneccessary.
Simply managing the problem is good for 10-20dB with little effort on the part of a DIYer. "Don't use an unshielded grapefruit sized cap with the mistaken belief it is somehow practically superior " is the upshot of the advice. And we've come full circle! :-D
- Status
- Not open for further replies.