Pooge and Syn08, you have found a thread to 'backbite' me, on. However, I am the 'expert' here, because I did my homework on these caps, and I have had bad experiences using them in professional and consumer designs.
Anyone here can learn about and characterize caps, if you want to.
First, you should have some way of measuring harmonic or IM distortion down to .01% or better. All you have to do is make the cap start rolling off some signal, either high or low.
Second, you can contrive some differential subtraction test to measure DA easily and effectively. Perhaps not the exact defined amount, but you can derive the RELATIVE magnitude compared to 'better' caps, like polypropylene or polystyrene.
The best ceramics like COG or NPO are pretty good in the distortion department, however they still have 10 times the DA compared to polystyrene. If that is OK with you, then go right ahead and use them.
However, the COG caps seem to blend into the more compromised types of ceramic material almost seamlessly, (as far as the exterior case is concerned) and it is difficult to tell at a glance what the material composing the cap is about 1000 pf or so. Then the distortion begins and hysteresis, and even higher DA.
What is it worth to you? Are you making 1000 units, where the cost of caps is critical, and you will get a dozen free lunches from the savings? Many real engineers are guilty of this.
Or are you making a unit or 2 for yourself, and a few dollars invested, maybe will sound a little better when actually using the audio equipment. It is your choice.
Anyone here can learn about and characterize caps, if you want to.
First, you should have some way of measuring harmonic or IM distortion down to .01% or better. All you have to do is make the cap start rolling off some signal, either high or low.
Second, you can contrive some differential subtraction test to measure DA easily and effectively. Perhaps not the exact defined amount, but you can derive the RELATIVE magnitude compared to 'better' caps, like polypropylene or polystyrene.
The best ceramics like COG or NPO are pretty good in the distortion department, however they still have 10 times the DA compared to polystyrene. If that is OK with you, then go right ahead and use them.
However, the COG caps seem to blend into the more compromised types of ceramic material almost seamlessly, (as far as the exterior case is concerned) and it is difficult to tell at a glance what the material composing the cap is about 1000 pf or so. Then the distortion begins and hysteresis, and even higher DA.
What is it worth to you? Are you making 1000 units, where the cost of caps is critical, and you will get a dozen free lunches from the savings? Many real engineers are guilty of this.
Or are you making a unit or 2 for yourself, and a few dollars invested, maybe will sound a little better when actually using the audio equipment. It is your choice.
Re: that is a very nice argument
Conclusion:
A microscope is bad tool to knock nails. Hammer is better.
sakis said:
Conclusion:
A microscope is bad tool to knock nails. Hammer is better.
hello.
there is a great variety of different ceramic caps....
in my experience greater values colour the sound (e.g. i have changed ceramic caps in receivers to plastic caps ( polystyrene or so...) to enhance the sound..........)
greater values are used for sound design, e. g. in guitar amps to give the sound a specific color................
the small one's ( npo....) are a good but not optimal choice for hf equalisation or so...........
greetings..............
there is a great variety of different ceramic caps....
in my experience greater values colour the sound (e.g. i have changed ceramic caps in receivers to plastic caps ( polystyrene or so...) to enhance the sound..........)
greater values are used for sound design, e. g. in guitar amps to give the sound a specific color................
the small one's ( npo....) are a good but not optimal choice for hf equalisation or so...........
greetings..............
Pease published some DA measurements:
http://www.national.com/rap/Application/0,1570,28,00.html
Iv'e seen ~ -115 db 1KHz IMD with npo - dissapeared when polystyrene was swapped in
npo looks better than silver mica on DA - but there can be manufacturer/formulation differences
for instance Pease' DA plot shows the better mylar as better than silver mica too
DIY discrete PA don't have a big need for smt - but commercial stuff or portable headphone amps are often built with smt components - polystyrene, polyethylene are not options - npo/cog can be used - anyone have the distortion and DA # for PEN or PPS smt caps?
http://www.national.com/rap/Application/0,1570,28,00.html
Iv'e seen ~ -115 db 1KHz IMD with npo - dissapeared when polystyrene was swapped in
npo looks better than silver mica on DA - but there can be manufacturer/formulation differences
for instance Pease' DA plot shows the better mylar as better than silver mica too
DIY discrete PA don't have a big need for smt - but commercial stuff or portable headphone amps are often built with smt components - polystyrene, polyethylene are not options - npo/cog can be used - anyone have the distortion and DA # for PEN or PPS smt caps?
john curl said:
You are again, and I suspect on purpose, mixing up things and facts.
Let's get at least this one straight; you are talking like about DA and how poor ceramics are in this respect.
- Mica has more dielectric absorbtion than NP0.
- There is absolutely no reason why silver mica caps should have lower distortion than NP0 ceramic caps, and in fact they don't.
The only problem that you as a "high end designer" have is the desperate need to perpetuate another myth supported by, and for the use of, the ignorant crowd. As much as the negative feedback bad rap, the 7th harmonic, silver wire, directional wires, Bybee quantum devices, etc... etc... etc...
Unfortunately, you don't realize the ROW beyond your "associates" are laughing their ***** reading about.
er that should be polyproplylene - not polyethylene in my post above (not that polyethylene isn't an excellent dielectric - just not commonly used in caps)
I hope its the Linear Polyethylene, vacuum melt extruded for high orientation/crystallinity and to prevent oxidation (in fact the feed pellet processing should have been inerted with N2 to prevent oxidation of the feedstock)
OK, could you point out a randomized blind listening test with enough statistical power to prove that an expensive nice looking plastic cap sounds better than a common looking ceramic NPO as compensating cap, or X7R as filtering cap not in the audio path. If not, it's anecdotal. If there are such studies, then there's no reason for these kind of arguments.Lumba Ogir said:
nelsonvandal said:
We need no stinkin' blind test. WE KNOW WHAT WE HEAR, and anybody debating our knowledge, in particular based on the physics as we know it, is a ******* ignorant.
Do I need to spell it for you?
Just when I thought this thread was getting on track and starting to distinguish between C0G/NP0s and lessor ceramics, it has disintegrated into a generalization war again.
JC indicated that C0Gs have 10 times the DA as polystyrene. Well this may be true, but polystyrene film is being phased out, and they are getting harder to find; and in the future, maybe impossible unless you happen to have bought a large stock.
On the other hand, C0Gs have the reputation of being very stable with voltage and temperature, maybe the best.
Further, I thought one of the issues was the supposed superiority of micas over C0Gs, and that DA was the factor of primary importance for this opinion. Now if C0Gs have superior DA than mica, as well as other superior characteristics over mica, what might be the basis for the opinion that micas are better?
Note that I did not ask for a generalization of ALL ceramics here.
JC indicated that C0Gs have 10 times the DA as polystyrene. Well this may be true, but polystyrene film is being phased out, and they are getting harder to find; and in the future, maybe impossible unless you happen to have bought a large stock.
On the other hand, C0Gs have the reputation of being very stable with voltage and temperature, maybe the best.
Further, I thought one of the issues was the supposed superiority of micas over C0Gs, and that DA was the factor of primary importance for this opinion. Now if C0Gs have superior DA than mica, as well as other superior characteristics over mica, what might be the basis for the opinion that micas are better?
Note that I did not ask for a generalization of ALL ceramics here.
jcx said:
And for what most people around here (especially the OP) build, -115dB caused by an NPO cap is the least of their problems.
Considering that the highly voltage dependent 2-3pF Cob of the transistor that it's in parallel with will add a whole lot more nonlinearity.
pooge said:
Well, jcx already posted a link to this above, but it's worth repeating. This article by Bob Pease tests DA for a number of capacitor types including C0G and mica. Figure 7 shows the graph of voltage error using a sample-and-hold circuit. Mica does not have good DA at all.
Yes, I think you have to. I think it's ignorant to be afraid of such tests. It's like being afraid of falling off the edge of a flat earth.syn08 said:
andy_c said:
Thanks. Forgot to check that link. Looks familiar, but I haven't seen it in a while. The interesting thing to me is how the same material can be quite different depending on the cap used, and that even the best NP0 is better than some polystyrenes in DA. It's too bad we can't get model numbers out of such papers for those who choose a capacitor that acts like a capacitor. However, this may not be as helpful at first blush, since Bateman found quite a difference between caps on the same tape. This ialone is fodder against those who make generalizations about capacitors. Other than designing caps out of the circuit altogether, it appears to be a difficult issue, without good test equipment or extensive trial and error, to know what you are getting.
One disconcerting aspect of Bob Pease's graph, (for our purpose) is the statistical distribution of the samples. He only used one or two samples of each and the distribution showed significant random overlap.
It makes me wonder about a couple things: where the bell curve for each dialectric really is, given a statistically significant number of samples; and more to the point, assuming both sides of these arguments are correct, how worthwhile is DA as a measure of sound quality. I can't draw any conclusions on the latter issue, based on the arguments here. Both the science and the art seem to be all over the map here.
And why are Teflon caps so freakin expensive? Teflon is used extensively for many cheap products and is not costly to manufacture (though there are environmental concerns with the manufacturing process that will undoubtedly affect prices). Can we make our own caps using plumber's sealing tape?
..Todd
It makes me wonder about a couple things: where the bell curve for each dialectric really is, given a statistically significant number of samples; and more to the point, assuming both sides of these arguments are correct, how worthwhile is DA as a measure of sound quality. I can't draw any conclusions on the latter issue, based on the arguments here. Both the science and the art seem to be all over the map here.
And why are Teflon caps so freakin expensive? Teflon is used extensively for many cheap products and is not costly to manufacture (though there are environmental concerns with the manufacturing process that will undoubtedly affect prices). Can we make our own caps using plumber's sealing tape?
..Todd
Anyone have any references linking DA and distortion? The Pease model of R/C extention ladder has no distortion and precious little frequency response deviations in many cases. It becomes easy to get into a bait(man)🙂 and switch situation where the measurements are taken in an inappropriate context for your application.
- Status
- Not open for further replies.