AndrR
pre-prepared responce, as I didn't say I didnn't believe, I asked questions, and for the record I dont believe in anything, life is but an illusion, here the reply:
If I am to learn anything, then why can I not question. Anecdotal information is of no use to me, I would like empirical data so that I can understand what is causing the change, and what effect electrically are taking place. This data can be collated with other experimental data and I can make a decision based on facts. That is what engineering is, and audio reproduction is engineering based, when I am trying to improve my system I am not creating some art work or musical instrument, I am building or re-engineering a tool, that's function is to reproduce music with the maximum fidelity, only by a combination of measurements and listening can an engineering decision be formed, and only with hard empirical data can you reproduce and determine how the change is made and what path to improve it more (if a change does in fact make an improvement).
pre-prepared responce, as I didn't say I didnn't believe, I asked questions, and for the record I dont believe in anything, life is but an illusion, here the reply:
If I am to learn anything, then why can I not question. Anecdotal information is of no use to me, I would like empirical data so that I can understand what is causing the change, and what effect electrically are taking place. This data can be collated with other experimental data and I can make a decision based on facts. That is what engineering is, and audio reproduction is engineering based, when I am trying to improve my system I am not creating some art work or musical instrument, I am building or re-engineering a tool, that's function is to reproduce music with the maximum fidelity, only by a combination of measurements and listening can an engineering decision be formed, and only with hard empirical data can you reproduce and determine how the change is made and what path to improve it more (if a change does in fact make an improvement).
Last edited:
Any plastic or non hermetic capacitor will be affected by moisture content. This decreases from use. As water is slightly polar there is voltage induced movement of the molecules. This may show up as an increase in distortion peaking at some frequency that could be influenced by film type and thickness. As the capacitor is used or even placed in a warm environment the humidity level in the film drops.
I haven't tried all of the types you mention. I think we agree the mylar has that problem. But I have also seen to a lesser extent the issue in polyprop. I don't know if it had to do with winding gaps or resin used. It was an "Audio" cap so quality of winding could be the reason for moisture issues.
Polystyrene and teflon are considered the golden boys of caps and are well known as very stable designs. But there is a cheap brand of polystyrene caps that do not seem to perform as well as others. As they even have a bit of different optical quality from the better ones I suspect either regrind or just a lower quality of resin.
Then there is always the issue of corrosion at the connections in extreme cases of moisture exposure, although I haven't seen that in any in use capacitors.
Far more likely that than the film- an example of why I avoid "audiophile" caps.
When I get some 'first off' plastic parts (ABS, PC generally), the moulders put them in a sealed bag with a drop of water for 24 hrs, this improves the materials properties by a small amount...I was surprised when I first learnt this as I would not have imagined that plastics can absorb water.
As I dont use audiophile caps, just standard available caps from Farnell etc I dont supose I'll worry about any humidity caused changes.
As I dont use audiophile caps, just standard available caps from Farnell etc I dont supose I'll worry about any humidity caused changes.
)
Magnetic saturation of the transformers of course! shifts in the magnetic pole, American High Frequency Active Aural Research Project! HAARP , little green men!
Actually nutty suggestions aside I also notice chances in audio output of coupling caps after running for some time. To my ears they smoothen out. I just did the .47 coupling in my C22 preamp with some NOS sprague black beauties and my initial impression was WHAT HAVE I DONE!! when given a few hours run in they sound just lovely. My opinion only, after observation of course.
Even new electrolytics continue forming for some time after being put in operation. The leakage measurably drops, and levels off after many hours.
I used to design and build very high order active filters for some tone signalling projects. They had to have some dielectric temperature coefficients balanced to work over temperature range, but never drifted off with time, so there was no sign of "burn in", so capacitors must "know" when they are in a hifi circuit rather than an audio comms circuit
Last edited:
I used to design and build very high order active filters for some tone signalling projects. They had to have some dielectric temperature coefficients balanced to work over temperature range, but never drifted off with time, so there was no sign of "burn in", so capacitors must "know" when they are in a hifi circuit rather than an audio comms circuit
Small wonder. They can sense the pace, rhythm and timing.
AndrR
pre-prepared responce, as I didn't say I didnn't believe, I asked questions, and for the record I dont believe in anything, life is but an illusion, here the reply:
If I am to learn anything, then why can I not question. Anecdotal information is of no use to me, I would like empirical data so that I can understand what is causing the change, and what effect electrically are taking place. This data can be collated with other experimental data and I can make a decision based on facts. That is what engineering is, and audio reproduction is engineering based, when I am trying to improve my system I am not creating some art work or musical instrument, I am building or re-engineering a tool, that's function is to reproduce music with the maximum fidelity, only by a combination of measurements and listening can an engineering decision be formed, and only with hard empirical data can you reproduce and determine how the change is made and what path to improve it more (if a change does in fact make an improvement).
Weeeell, yes ... sort of!
Here's what I agree with you on:
"only by a combination of measurements and listening can an engineering decision be formed"
NB: "fundamentalist" engineers would say that measurements and theory are all there is to it - listening has got nothing to do with it!
these are the people who have us boring 70s & 80s amps with their vanishingly small THD (achieved by lashings of GNFB ).Here's what I don't entirely agree with you on:
"and only with hard empirical data can you reproduce and determine how the change is made and what path to improve it more"
Sometimes, you don't know what it is you need to be measuring, to get empirical data on why something sounds different/better. Case in point - years ago, I was one of several people listening to some amps being swapped in a system - so everything else was the same. A particular, single-ended design (not a "SET" - not wholly "tube") by Hugh "AKSA" Dean called the "Glass Harmony" made the singer jump forward from the plane of the speakers by about 6'. This effect was unanimously deemed to be a terrific plus-point for that amp (more soundstage depth) ... so what measurements do you think would explain this?
And here's another point on which I don't entirely agree with you:
"when I am trying to improve my system I am not creating some art work or musical instrument, I am building or re-engineering a tool, that's function is to reproduce music with the maximum fidelity"
From over a decade of discussing Hugh Dean's amplifier design philosophies with him, and seeing him come out with better & better-sounding designs, I believe there's a lot of "musical instrument" concepts in the art of designing a great-sounding amplifier. Of relevance here are:
* employing judicial amounts of NFB - but no GNFB. (With the emphasis on "judicial"!)
* ensuring distortion harmonics are monotonically decreasing.
So, in effect, you are Christian ... and I am Muslim ... and we will never agree on the right way to get to heaven - but we can both do our best to get there.
Regards,
Andy
It was careful listening tests that originally established which measurements are important. If, say, 1% low order distortion is about the limit of audibility then aiming for 0.001% is unnecessary, but being happy with 10% is strange. Amps are supposed to be boring, if hi-fi is the aim. It is the music which should be exciting.andyr said:
People who struggle with theory are more likely to design an amp with significant technical shortcomings. Curiously, such amps are often expensive and get glowing reviews from journalists who can't read circuit diagrams but can read the price tag and the accompanying 'story'.
Amplifiers should not "sound great", they should not sound at all.
There is nothing intrinsically wrong with GNFB, and nothing fundamentally different between GNFB and other NFB - they are governed by exactly the same physics and maths. The issue is over how many stages is the feedback applied, and what that means for stability and peak signal levels before any integrator (or low pass filter) in the forward path. Slapping lots of feedback around a bad design is not a good way to get a good design; neither is using insufficient feedback around a weak design.
Apart from a well-balanced push-pull amp (where 3rd may exceed 2nd) this is nearly always true, however much or little feedback is used.
A particular, single-ended design (not a "SET" - not wholly "tube") by Hugh "AKSA" Dean called the "Glass Harmony" made the singer jump forward from the plane of the speakers by about 6'. This effect was unanimously deemed to be a terrific plus-point for that amp (more soundstage depth) ... so what measurements do you think would explain this?
Some BBC research long ago found that frequency balance and volume gave image depth clues
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.