suspected to reflect mechanical movement within the capacitor - it's slower than the sort of acoustic resonance that's being talked about here, but re latively easy to demonstrate in the right circuit - which in my case was a quad-ramp integrating A/D converter ...
charge again. The effect is very obvious with polyester film and several o thers but is largely absent with polypropylene film.
** Shame I only measured a few dozen caps and found ones sold as being poly propylene were in fact polyester.
** Cos it is inconsistent with the facts.
audio, which doesn't stop the crooks who exploit audiophools from selling expensive polypropylene capacitors that aren't appropriate for most audio a pplications.
** "Capacitor sound" is one of the longest running, biggest myths in the wo rld of audiophoolery. There are thousands of articles on the topic publishe d on the net.
s suspected to reflect mechanical movement within the capacitor - it's slow er than the sort of acoustic resonance that's being talked about here, but relatively easy to demonstrate in the right circuit - which in my case was a quad-ramp integrating A/D converter ...
ll charge again. The effect is very obvious with polyester film and several others but is largely absent with polypropylene film.
lypropylene were in fact polyester.
That's the audiophool market for you. Buy your capacitors from Farnell/Elem ent 14 and you'll get what you pay for.
What facts? I adduced a few which are consistent with my point of view and you didn't argue with them.
or audio, which doesn't stop the crooks who exploit audiophools from sellin g expensive polypropylene capacitors that aren't appropriate for most audio applications.
world of audiophoolery. There are thousands of articles on the topic publis hed on the net.
One of my audio-freak friends wanted a voltage divider once, and I gave him one made from 0.1% 15ppm precision resistors, together with the calculatio ns that showed that it was exactly the same as the same thing made with muc h cheaper, lower tolerance 100ppm metal film resistors. He was still happie r to have the ritzy resistors - they probably impressed one of his more gul lible colleagues who thought that tube amplifiers sounds "better" and spent a lot of money on this delusion.
Audiophoolery is about concocting a story to sell to the marks. Capacitor s ound is almost certainly nonsense (once you get away from microphonic ceram ic capacitors) but dielectric absorbtion isn't.
The company concerned is WES Components here in Sydney - who supply the repair trade. I bought some 4.7uF & 10uF, 250V caps to use in passive x-overs for about $2 or $3, claimed to be polyprop but showing clear evidence of DA.
Didn't matter a hoot me.
** FFS, the fact the effect depends on the di-electric used.
** The "marks" are only too willing to believe that if a component costs more it will sound better and since they only ever do subjective, fully sighted listening tests, that is just what they hear after fitting such parts.
This kind of self delusion is very powerful and affects most people, whether they are heavily into hi-fi sound or not.
epair trade. I bought some 4.7uF & 10uF, 250V caps to use in passive x-over s for about $2 or $3, claimed to be polyprop but showing clear evidence of DA.
Even polypropylene shows some dielectric absorbtion.
If you had wanted to work out which dielectric was used in the capacitors, you might have measured the glass transition temperature (if you could fin d someone with the measuring gear)
formatting link
Polypropylene comes in a -20C, polyethylene terephthalate (PET) - which we call polyester - comes in at 70C and polycarbonate at 145C.
You might be able to do something with the infra-red absorbtion spectrum - solid state samples have horribly broad lines, but polypropylene won't have the oxygen-carbon stretch modes that you'd see in polyester or polycarbona te.
Different dielectrics don't have different mechanical properties?
The fact that polypropylene has a glass transition temperature below room t emperature, while the other two are well above room temperature strikes me as compatible with a mechanical explanation of dielectric absorbtion.
more it will sound better and since they only ever do subjective, fully sig hted listening tests, that is just what they hear after fitting such parts.
her they are heavily into hi-fi sound or not.
Peter Baxandall and his mate from Quad were full bottles on the subject. I never heard them talk about the subject, but their opinions did get ventila ted in HiFi News and Record Review, back when it was worth reading. Double blind comparisons were the only ones that meant anything.
Brian Moore at Cambridge
formatting link
could also occasionally be heard saying a few words about that kind of self
-deception when you ran into him socially. He owned the only ionophone twee ter I've ever heard ...
** Measuring the tempco of a capacitor is simple. Polyprop has a near zero or negative one, while polyester is strongly positive.
A DMM with capacitance rage and source of heat are all you need.
A 555 oscillator and frequency counter can make a very sensitive tempco tester too. I tried them both.
** Do drowning men clutch at straws ?
Time for you to stop.
** Wireless World published numerous articles and letters from both. Prompted me to devise a listening test procedure that eliminated the main causes of imaginary differences and hence false positives.
formatting link
** But to whom ?
Audiophiles and audiophools were left quite unimpressed after hearing about some rather tedious listening test done somewhere, set up by people with a vested interest, like Quad.
A conclusion that flies directly in the face of common experience and deeply held prejudices associated with the cost of an item and brand credibility is hard for non scientific folk to swallow.
The test procedure detailed in the link gives an instant result when set up correctly. I tried it on quite a few people and all made the same remark - " the switch is not working".
Turning one amp off proved it was.
I am not sure anyone was 100% convinced, but it worried them and the topic of mysterious "amplifier sound" was dropped.
he repair trade. I bought some 4.7uF & 10uF, 250V caps to use in passive x- overs for about $2 or $3, claimed to be polyprop but showing clear evidence of DA.
rs, you might have measured the glass transition temperature (if you could find someone with the measuring gear)
o or negative one, while polyester is strongly positive.
My capacitance measuring tool is the Almost All Digital Electronics L/C met er IIB. It's probably accurate enough to capacitor tempco on polyester, but I'd never though of doing it. You've taught me something useful.
ester too. I tried them both.
The 555 has it own tempco. An emitter-coupled multivibrator can have a much less temperature-sensitive threshold - still not perfect (nor good enough for what we were trying to do). We wrote it off to the temperature dependen ce of the Early voltage without actually doing the work that might have pro ved the point.
They might, but polycarbonate, PET (polyester) and polypropylene definitely do have different mechanical properties.
You'd like me to, but can't find a convincing counter-argument.
sts more it will sound better and since they only ever do subjective, fully sighted listening tests, that is just what they hear after fitting such pa rts.
whether they are heavily into hi-fi sound or not.
.
pted me to devise a listening test procedure that eliminated the main cause s of imaginary differences and hence false positives.
To people who can manage critical thinking.
ut some rather tedious listening test done somewhere, set up by people with a vested interest, like Quad.
Audiophiles make quite a lot of money out of audiophools by talking nonsens e. They have a vested interest in being taken seriously.
ply held prejudices associated with the cost of an item and brand credibili ty is hard for non scientific folk to swallow.
We are talking about another religion here ...
up correctly. I tried it on quite a few people and all made the same remark - " the switch is not working".
No surprises there. Peter Blaxandall's contention was that all "good" ampli fiers were indistinguishable for human listeners. Quite a few over-priced a mplifiers aren't "good" in that sense - tube-based amplifiers have been kno wn to introduce odd harmonic distortion, which some listeners like ...
Among them, that the amps must be used within their power and current outpu t limits at all times - no clipping allowed. Also, bandwidths must be simil ar so that sub sonic or supersonic input signals are not rejected by one ye t amplified by another. Most turntables deliver considerable sub-sonic signals ( particularly with warped LPs ) and so do a few CDs - making woofer cones flap about and produ ce audible amplitude modulation.
More subtle effects can occur when an electrostatic speaker is being used ( like the Quad ESL57) - subsonic energy can saturate the input transformer c ore and induce current limiting ( hence severe distortion) in the output st ages of SS and valve amplifiers.
Knowing about this problem, Quad built subsonic filters into most of their SS amplifiers. Practically no one else did.
Baxandall also devised a test rig that allowed the error signal of an ampli fier to be continuously extracted while playing music programme into a spea ker - IOW a wideband, input/output subtractor. It was then applied to the Q uad 303 amp.
This error signal was fed to another amplifier and reproduced via a speaker in a quiet room at the actual level. It was invariably *non audible* until the Quad 303 was driven into clipping, whence it became very loud.
The set up was demonstrated by Quad to numerous audio reviewers and other a udiophools none of whom could see the point.
put limits at all times - no clipping allowed. Also, bandwidths must be sim ilar so that sub sonic or supersonic input signals are not rejected by one yet amplified by another.
h warped LPs ) and so do a few CDs - making woofer cones flap about and pro duce audible amplitude modulation.
(like the Quad ESL57) - subsonic energy can saturate the input transformer core and induce current limiting ( hence severe distortion) in the output stages of SS and valve amplifiers.
r SS amplifiers. Practically no one else did.
lifier to be continuously extracted while playing music programme into a sp eaker - IOW a wideband, input/output subtractor. It was then applied to the Quad 303 amp.
I wonder which came first? That or the Quad 405?
er in a quiet room at the actual level. It was invariably *non audible* unt il the Quad 303 was driven into clipping, whence it became very loud.
audiophools none of whom could see the point.
The Quad 405 made the point very clear, and it was well received.
Being one cylinder within another, I think a coaxial resonator best describes it.
I suppose my question relates to how the resonant frequency is affected by the 1) proportions of the cap, and 2) the refractive index of the insulator.
For example, if I have a 30mm tube inside a 40mm tube, both 100mm long. How do I calculate the resonant frequency if the insulator in BaTiO3 (2.4272)?
Is there any chance it could be engineered to fall below 100KHz?
utput limits at all times - no clipping allowed. Also, bandwidths must be s imilar so that sub sonic or supersonic input signals are not rejected by on e yet amplified by another.
ith warped LPs ) and so do a few CDs - making woofer cones flap about and p roduce audible amplitude modulation.
ed (like the Quad ESL57) - subsonic energy can saturate the input transform er core and induce current limiting ( hence severe distortion) in the outpu t stages of SS and valve amplifiers.
eir SS amplifiers. Practically no one else did.
mplifier to be continuously extracted while playing music programme into a speaker - IOW a wideband, input/output subtractor. It was then applied to t he Quad 303 amp.
** The 303 was released in 1968.
aker in a quiet room at the actual level. It was invariably *non audible* u ntil the Quad 303 was driven into clipping, whence it became very loud.
er audiophools none of whom could see the point.
** Err, what "poit " was that?
** Not by the UK hi-fi press it wasn't - it got canned. The unfortunate tit le "current dumping" turned off audiophools in droves.
The Quad 405 ( mk 1 & 2) are now the subject of countless pages on the net detailing the most elaborate "modifications" intended to make it sound good . Fancy film and electro capacitors, high speed op-amps and all kinds of wh acko ideas.
output limits at all times - no clipping allowed. Also, bandwidths must be similar so that sub sonic or supersonic input signals are not rejected by one yet amplified by another.
with warped LPs ) and so do a few CDs - making woofer cones flap about and produce audible amplitude modulation.
used (like the Quad ESL57) - subsonic energy can saturate the input transfo rmer core and induce current limiting ( hence severe distortion) in the out put stages of SS and valve amplifiers.
their SS amplifiers. Practically no one else did.
amplifier to be continuously extracted while playing music programme into a speaker - IOW a wideband, input/output subtractor. It was then applied to the Quad 303 amp.
Not the 303 but the error signal extraction.
peaker in a quiet room at the actual level. It was invariably *non audible* until the Quad 303 was driven into clipping, whence it became very loud.
ther audiophools none of whom could see the point.
That one could extract an error signal from the amplifier and amplify it se parately
itle "current dumping" turned off audiophools in droves.
HiFi News and Record Review were decidedly positive about it. I never paid any attention to the down-market UK audio magazines - Quad wouldn't bribe t hem so they never had anything nice to say about Quad, which didn't do anyt hing for their already negligible credibility. I stopped buying HiFi News a nd Record Review when their opinions became more purchasable.
Peter Walker had retired from Quad when I got my job interview there. Ross Walker, his son - whom I did talk to - didn't seem to know much about the t echnology. I blew the part of the interview with the technical director by claiming to know a bit about transformers (by which I meant ferrite-cored s ignal transfomers). The technical director was a whizz with iron-cored powe r transformers - as I would have expected, if I'd thought about it - and ga ve me a very hard time.
t detailing the most elaborate "modifications" intended to make it sound go od. Fancy film and electro capacitors, high speed op-amps and all kinds of whacko ideas.
In other words it sold well (and for a long time - mark 1 from 1975 to 1982 , and Mark 2 from 1982 to 1993), and there are plenty of used examples arou nd to play with.
an amplifier to be continuously extracted while playing music programme int o a speaker - IOW a wideband, input/output subtractor. It was then applied to the Quad 303 amp.
** The idea of extracting the error signal for separate reproduction dates from about 1970. It was then applied by Quad to their 303 and 405 models an d demonstrated to various people who visited the factory.
Interestingly, final testing of production units of the ESL63 speaker inclu ded each being placed alongside a thoroughly tested reference example - bot h speakers facing a test mic, centrally positioned 2 metres away. The outpu t of the mic was amplified and fed to a scope.
The speakers were fed with a square wave of about 1kHz and the operator cou ld swap the phase of one of them with a switch. The tests was passed if the signal from the mic dropped by 20dB or more at all frequencies.
separately
** The 405 "current dumping" amplifier did something like that internally - if that is what you are alluding to and not relevant at all.
s Walker, his son - whom I did talk to - didn't seem to know much about the technology. I blew the part of the interview with the technical director b y claiming to know a bit about transformers (by which I meant ferrite-cored signal transfomers). The technical director was a whizz with iron-cored po wer transformers - as I would have expected, if I'd thought about it - and gave me a very hard time.
** The audio step up transformers in Quad's electrostatic speakers were exc eptional. The ESL63 uses 2 identical C-cores with the secondaries in series to generate up to 5kVrms at any audio frequency.
net detailing the most elaborate "modifications" intended to make it sound good. Fancy film and electro capacitors, high speed op-amps and all kinds o f whacko ideas.
82, and Mark 2 from 1982 to 1993), and there are plenty of used examples ar ound to play with.
** Yep - sold in large numbers almost exclusively to *non audiophools* all over the world. Mainly because it was one of the cheapest, famous brand 100 wpc amplifiers you could buy. Over 200,000 were made.
f an amplifier to be continuously extracted while playing music programme i nto a speaker - IOW a wideband, input/output subtractor. It was then applie d to the Quad 303 amp.
s from about 1970. It was then applied by Quad to their 303 and 405 models and demonstrated to various people who visited the factory.
luded each being placed alongside a thoroughly tested reference example - b oth speakers facing a test mic, centrally positioned 2 metres away. The out put of the mic was amplified and fed to a scope.
ould swap the phase of one of them with a switch. The tests was passed if t he signal from the mic dropped by 20dB or more at all frequencies.
One of colleagues from Cambridge Instruments went from there to be a senior quality control guy at Quad (or so he told me). He did like Quad's approac h to quality, though some of their choices were a bit idiosyncratic. They w ere very late moving to plated through-holes, relying on metal inserts in d ouble-side boards long after everybody else had made the switch.
t separately
- if that is what you are alluding to and not relevant at all.
That was exactly what I was referring to, and it strikes me as obviously re levant.
xceptional. The ESL63 uses 2 identical C-cores with the secondaries in seri es to generate up to 5kVrms at any audio frequency.
Weren't they part of the delay-line structure that fed the concentric eleme nts of the EL-63 radiator?
e net detailing the most elaborate "modifications" intended to make it soun d good. Fancy film and electro capacitors, high speed op-amps and all kinds of whacko ideas.
1982, and Mark 2 from 1982 to 1993), and there are plenty of used examples around to play with.
l over the world. Mainly because it was one of the cheapest, famous brand 1
00wpc amplifiers you could buy. Over 200,000 were made.
As far as I know, it also sounded okay. I never felt the need for that much power. Built my own amplifier early on (1975) then replaced it with a NAD about a decade or so laterwhen my wife got picky about the aesthetics. Even tually needed a second amplifier (for the dining-room and kitchen area) and bought a Sony TA-F361R after an hour so browsing around downtown Nijmegen
- not a particulary exhaustive process, but we had guests coming.
I'm clearly not reading your posts as you expect them to be read, but I suspect the problem here is more your unrealistic expectations than my reading comprehension.
You've made a significant snip here which you haven't marked. The next sentence - which you've posted as if it referred to the EL-63 - came several lines later, and referred to the Quad 405, which you don't seem to think all that highly of.
I'm reading it all, but clearly not comprehending it in quite the way that you do.
Following your line of thought would take telepathy, rather than superior r eading comprehension. The sort of argument that goes "** Cos it is inconsis tent with the facts" and then fails to identify the "facts" that are incons istent with the hypothesis is usually labelled fatuous, rather than logical .
I'ts a minor point, but indicative of your capacity to lay out a coherent a rgument.
That's an opinion - from somebody who sounds tolerably disturbed at the moment, and is clearly not receptive to my suggestion (snipped) that what he's posting doesn't look much like logical argument.
And it would be even more pedantic to point out that you've both snipped the comment to the effect that while it might have been a minor point, it was a revealing one, and failed to mark the snip.
It's a misapprehension to make a distinction between intent and meaning in this context. The cognitive failure seems to be more yours than mine.
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.