See PMA, this is why we don't bother with really low level tests. As I said, this stuff happens at a very low level, the ear can hear it, but the test equipment has problems measuring it. While I commend you for putting out the effort measure it, it is generally easier to just listen to it, for differences. Look at poor Clarity caps. They made good and interesting measurements. They presented a paper. Where did it get them? Maybe somewhere with listeners, but 'engineers' won't have it.
Actually I did an article on an inexpensive accelerometer. Among other uses I showed putting capacitors on a shaker table, biasing then with a resistor and 9v battery. It ran in Elektor this past summer.
No big surprises. All capacitors had some level of being microphonic. The axis of acceleration matters, and some capacitors were resonant in the mid range audio band on some axis. Liquid filled capacitors were less microphonic (Duh!).
I also measured a speaker case.
The conclusion if you put your phono preamp on top of your loudspeakers, some film capacitors will induce extra signal as high as -50db re the program material. Also some ceramic capacitors had 20db less output than a reference guitar pickup!
From my reading I used to prefer round capacitors as their values were supposed to be more stable as the squared capacitors were wound round and then squished. This was supposed to cause them to drift in value with time. Turns out the squishing reduces the sensitivity to vibration, particularly on the narrow axis or in simple terms the way they are intended to be mounted.
Also not surprising I did not get results to confirm what was shown in that paper and would expect two or three resonances based on the axis of propagation differences.
So paying attention to mechanical movement in the equipment housing and mechanical mounting methods can actually yield a measurable and listening difference.
Now lets talk about capacitors in crossovers inside loudspeaker cases!
One does not follow from the other, you can almost always measure a difference. Remember Fudd's Law?
One does not follow from the other, you can almost always measure a difference. Remember Fudd's Law?
Scott,
The example of -50db phono preamp is what I meant by measure and hear. Yes there are some who could not, but it really does stick out to my ears at that level.
>We have known about this since the vacuum tube days
>and unpolarized AC line plugs (2 prong).
>Why this happened was always a mystery to me.
It's something to do with the interaction of the chassis
with the ' ground field ' (for lack of a better term)
If a little of the hot side of the mains is on the chassis
it causes hum (as well as some potential for hazard)
Guitar amps actually have a switch to reverse polarity.
>and unpolarized AC line plugs (2 prong).
>Why this happened was always a mystery to me.
It's something to do with the interaction of the chassis
with the ' ground field ' (for lack of a better term)
If a little of the hot side of the mains is on the chassis
it causes hum (as well as some potential for hazard)
Guitar amps actually have a switch to reverse polarity.
Attachments
Last edited:
Scott is correct about the frequency doubling. B&K described this multi decades ago in terms of microphone capsule calibration.
However, when it comes to caps, a ringy-dingy cap may not be completely neutral.
Going further, some ceramic caps have piezoelectric properties and at this time, I am listening though about 3 of them, per channel. I know, I know, it is just my imagination, but I want my tube tuner back, if replacing these caps does not fix the rough but subtle 'edge' on my favorite FM programs.
However, when it comes to caps, a ringy-dingy cap may not be completely neutral.
Going further, some ceramic caps have piezoelectric properties and at this time, I am listening though about 3 of them, per channel. I know, I know, it is just my imagination, but I want my tube tuner back, if replacing these caps does not fix the rough but subtle 'edge' on my favorite FM programs.
I don't mean to give you a hard time but you have to take speculative conclusions to the next level. A speaker playing that loud would be lucky to have .3% distortions.
Well men, I guess our job is done, and we should change our occupations to making widgets, or military weapons. ;-) Of course, Scott, loudspeakers have a significant amount of distortion, especially direct radiators. How can we stand them? Why do we press on to find better amplifier designs? We asked this question almost 40 years ago, when my friends and I went to see Richard Heyser, to ask him, as resident guru at the time, the answer to this very question. His response was: It is global negative feedback that is the culprit. Get rid of global negative feedback and improvement will be made. He found this out by accident, listening to an open loop amp on his K-horn, just for the heck of it. His amp design is in the JAES. It measured lousy, YET sounded good. What a concept!
Scott,
There is a fine line between rigor and a hard time.
But a loudspeaker producing say 70dba slow weighting ANSI/IEC metered would peak around 90 to 100 playing a record. So 100-50 would put the capacitor echo down by only 20db. Pretty sure you would hear that.
Of course you understand that there is a difference between THD and this effect.
Had a Disco once where there were trolley tracks in the road outside. When a trolley went by the turntable had lots of problems. Putting in a single notch filter solved that problem. Building resonance plus tone arm plus speaker/room added up to something big.
When a live sound system is near feedback the resonances behave differently. The amount they can stretch out an excited mode would surprise some. That is what I think happens with the phono preamp on the loudspeaker.
As to measuring anything audio, if you can't measure a difference you just ain't try'n.
Yes, we found that the differing capacitor types and designs could be heard as much as -50db and more (lower) in the mix, in speaker crossovers.
So, the idea that one can use electrolytics (or even PET) in zobels and such that may be acting, seemingly over 50db down in the mix ....ends up being totally off.
The ear is not a nice little set of engineering numbers, it is a totally different animal than math.
The problem is the same as it has always been. and that is getting the engineering mind off the numbers game and into a reality that is not based on all reality being that of prior art or prior knowledge paths, in totality. New stuff is exactly that- new stuff. Even if it is simply new to the given person.
Like any sensible human being, I do what I can to speak on the subject, but stop short of hammering my self to pieces over that which others have mental issues on 'going clear' on.
It does not mean I feel superior, no that would be someone else's mental conception of my attitude. Perhaps they feel some sort of denigration is intended. Not at all. But I don't have the time to pussyfoot around their issues on the subject. I have done my best to point out some areas where things can be pushed forward and if they can't or won't invest in it or attempt to understand differently,as that is what it takes..Then I simply walk away , far ahead of them, in that given subject area.
I have done what conscience demands of me, with regard to trying, and then I simply have to let it go.
It's a tough lesson but one does tire of being stabbed by those who refuse to see and refuse to listen. If it is simply outside of the given capacity - then so be it. All I can do is hope that it is a temporary block in their makeup. Which is why I don't spend much time, if at all, on these forums anymore. I have my own issues. Like most, I can see that of others. I cannot address the given person's blind areas, but I can point to them....
Last edited:
We seem to have switched from the original discussion on the capacitor generating SOUND due to magneto- and electrostriction, to a capictor generating a SIGNAL due to microphonics. A different kettle of fish.
As to the original issue, where say the cap body vibrates at less than a material molecule diameter, I wonder what the generated SPL level would be, ballpark number for a 1 in dia, 2 in length wound foil cap?
jan didden
As to the original issue, where say the cap body vibrates at less than a material molecule diameter, I wonder what the generated SPL level would be, ballpark number for a 1 in dia, 2 in length wound foil cap?
jan didden
I put the fix for that.... in your hands, John. You didn't hear me.
I even went through the logic of it on the phone with you. I gave you the best you can imagine. I gave you what you wanted with none of the negatives. The best of both worlds. And.... like all good solutions, it is dead simple.And that particular solution can be applied to a complex design and it can distort till the cows come home and it will sound the same at any level, no change. Damping factor of 1000 and the sound of a single ended zero-feedback class A amp, with none of the weaknesses. I did the experiments, I heard the differences -I know what it does.
Last edited:
I put the fix for that.... in your hands, John. You didn't hear me.
And that particular solution can be applied to a complex design and it can distort till the cows come home and it will sound the same at any level, no change. Damping factor of 1000 and the sound of a single ended zero-feedback class A amp, with none of the weaknesses. I did the experiments, I heard the differences -I know what it does.
Are you willing to share it with all of us?
After taking 20 years of prior thought ---to get there? No.
Apologies. I did not give the solution to John to have him Vet it, to get some sort of 'approval', per se,...not at all. I gave it to him as I like him.
That is a valid point or consideration, Jan, but the noise is always heard by the ear as the sum total Peak level.
So it is always irrelevant if any noise of any sort is a fraction of 1% or less.
This never concerns the ear, which only hears the sum total, so the noise is an additive thing... heard as a change in the peak values of the given harmonic content in the given signal.....and thus is heard by the ear as an obvious sonic change.
It's like the point that we don't swim on some fractional aspect of water, we swim at the top and thus all that is below it -- affects the peak surface of the water that is - our sole concern. Every fraction of water below the peak must be 'just so' --for us to swim the surface. Every molecule of water below the surface affects our swimming and perception of the swim.
It is only a human illogical process of applying mental weighting to the math, an internally created mental idealistic weighting to the numbers involved ----that is what causes the disconnect in logic.
Last edited:
I should say I gave it to him to play with and see if it was useful in his work. I already know what it does and do plan to use it commercially, some day. If I ever get ahead enough to use it in some contracted design.
I can't design my way out of a paper bag,and i've said that before. But that does not say that I do not know where some of the critical problems lie on/in design work for most audio circuits.
I spoke frankly (on critical aspects) with another audio engineer the other day and they told me I was one of maybe 5 people (and less) on the planet they knew of that had figured a particular thing out, about audio design.
The obvious hint in that statement is that---- a large number of the remaining problems lie outside of basic engineering issues.
Otherwise all the circuit designers would have defeated them by now, hhhmmm?
That much is obvious.
- Status
- Not open for further replies.