I discovered using side-by-side electros for tants to reduce distortion when operating at low levels and 0 DC Volts. This was used by some designers in the early 80's. Now we AVOID tants completely. Series connection did not help for DA or non-linear distortion because the distortion mechanism is different from just reverse leakage.
Nelson Pass said:
Not to mention the electromechanical devices (including eardrums) used to evaluate these distortions.
It is my opinion that the role of capacitor is overstated
as a source of distortion and sonic degradation. [/QUOTE]
Apparently not to people who can hear type and directivity of
the wire used as a bypass on the PCB, orientation of a single
resistor or difference between square and round feet on a
power amp.
There are many things in the "high end" audio world that are bogus.
However I will confidently say that some capacitors will make a negative difference in audio, therefore I think it is best to stick with certain types.
Please understand that caps ARE sometimes necessary unless you want added complexity in a circuit.
If you want to go completely capacitorless, then go for it. But the addition of servos and trimming might be avoided if you bow down to cap coupling.
There are many caps out there that really do not color the sound at all... and they don't necessarily cost a lot. Rodersteins from Mouser are great. Nichicon Muse make excellent output coupling for the single-ended folks out there. Rubycon Blackgates are nice for stuff like feedback decoupling, etc.
However, if you honestly cannot tell the difference between a Teapo electrolytic and a Nichicon Muse electrolytic, then I submit that you either aren't as careful a listener as you may think, or that your equipment is somehow dithering so much that such detail is inaudible.
However I will confidently say that some capacitors will make a negative difference in audio, therefore I think it is best to stick with certain types.
Please understand that caps ARE sometimes necessary unless you want added complexity in a circuit.
If you want to go completely capacitorless, then go for it. But the addition of servos and trimming might be avoided if you bow down to cap coupling.
There are many caps out there that really do not color the sound at all... and they don't necessarily cost a lot. Rodersteins from Mouser are great. Nichicon Muse make excellent output coupling for the single-ended folks out there. Rubycon Blackgates are nice for stuff like feedback decoupling, etc.
However, if you honestly cannot tell the difference between a Teapo electrolytic and a Nichicon Muse electrolytic, then I submit that you either aren't as careful a listener as you may think, or that your equipment is somehow dithering so much that such detail is inaudible.
to refresh a thread
my DA, now working to spec (thanks Demian and Boonton) isn't quite to the level that Cyril described in his article (parts per million) , but I've found some interesting stuff by comparing distortion levels in families of caps:
ASC Metalized Polypropylene are really low distortion, and easily available through Allied Electronics (www.alliedelec.com). Black Cats are a little inferior to ASC, but very good considering that they are (by now) ancient. Orange Drops are a little inferior to Black Cats. These were all in the 400 VDC range. Best of all were some MilSpec TRW's, however, and I have no idea where you can get them except at a surplus house in Cleveland. Some older MilSpec caps from other manufacturers were all over the lot in terms of THD.
My horde of polystyrene caps are nowhere near as good as I had hoped. Siemens Epcos "Silver Caps" in the same voltage rating were very good.
Some garden variety polyster film caps from a well known Japanese mfr, supplied via a reputable US distributor were pretty bad, measuring in the high hundredths of a percent THD.
Does it make a difference in the listening?
my DA, now working to spec (thanks Demian and Boonton) isn't quite to the level that Cyril described in his article (parts per million) , but I've found some interesting stuff by comparing distortion levels in families of caps:
ASC Metalized Polypropylene are really low distortion, and easily available through Allied Electronics (www.alliedelec.com). Black Cats are a little inferior to ASC, but very good considering that they are (by now) ancient. Orange Drops are a little inferior to Black Cats. These were all in the 400 VDC range. Best of all were some MilSpec TRW's, however, and I have no idea where you can get them except at a surplus house in Cleveland. Some older MilSpec caps from other manufacturers were all over the lot in terms of THD.
My horde of polystyrene caps are nowhere near as good as I had hoped. Siemens Epcos "Silver Caps" in the same voltage rating were very good.
Some garden variety polyster film caps from a well known Japanese mfr, supplied via a reputable US distributor were pretty bad, measuring in the high hundredths of a percent THD.
Does it make a difference in the listening?
Hi,
Not to mention the fact that quite often the PS caps are in the signal path as well so they'd better be good too...
From experience, replacing standard quality electrolytics with Black Gate or MKP quality caps makes a huge difference to the final result especially when the rest of the components is already top notch.
Whether replacing a single ordinary filmcap with an exceptional one would make for an audible difference would probably depend on where you'd place it but all in all, yes it surely does make for an audible difference in my book.
DA is just one of the important aspects, self-inductance, foil quality and how tightly the cap is wound, lead out wire quality and how well it mates with the metal(ized) inside the cap, it all makes a difference. Audibly so and measurably so.
Cheers,
Not to mention the fact that quite often the PS caps are in the signal path as well so they'd better be good too...
From experience, replacing standard quality electrolytics with Black Gate or MKP quality caps makes a huge difference to the final result especially when the rest of the components is already top notch.
Whether replacing a single ordinary filmcap with an exceptional one would make for an audible difference would probably depend on where you'd place it but all in all, yes it surely does make for an audible difference in my book.
DA is just one of the important aspects, self-inductance, foil quality and how tightly the cap is wound, lead out wire quality and how well it mates with the metal(ized) inside the cap, it all makes a difference. Audibly so and measurably so.
Cheers,
Hi Frank,
Most of us can't measure things like inductance of a capacitor.
In any case capacitors meant for high speed circuits ( pulse operation ) should have low inductance.
So would this mean that such capacitors might be better for audio. I noticed that the dv/dt capability of normal film is 1/10th to 1/100 th of those meant for switching circuit use.
So that should mean lower inductance of the switching ones.
But do these sound any better . Any idea ?
Cheers,
Ashok.
Most of us can't measure things like inductance of a capacitor.
In any case capacitors meant for high speed circuits ( pulse operation ) should have low inductance.
So would this mean that such capacitors might be better for audio. I noticed that the dv/dt capability of normal film is 1/10th to 1/100 th of those meant for switching circuit use.
So that should mean lower inductance of the switching ones.
But do these sound any better . Any idea ?
Cheers,
Ashok.
Hi,
All in all, yes they do sound better.
However these AC pulse caps aren't always available in the technology that would be ideal for the application.
Often these caps use (or at least used to) paper as a dielectricum, were only available in small values and high voltages etc.
I write used to, as I haven't checked this in a long time.
I can think of a few audio equipment manufacturers that deliberately forewent better dielectrical properties of coupling caps and instead chose to use AC pulse caps because they sounded better in that application.
Usually they were put as a driver to output cap in tube amps where DA is of lesser importance, IIRC.
One aspect that clearly shows their superiority is better stereo imaging for instance.
Nowadays most special audio caps ( AKA "boutique caps) implement Mrs Jung and Marsh's theories with great success but I think the ideal audiocap is still evolving.
Even if it can be built today it's still not economical to go to market with such a range of caps today IMO.
Fortunately the gap is closing fast, both film and electrolytic caps have improved tremendously over the past fifteen years.
Probably regardless of what the audio market said anyway...
Cheers,
All in all, yes they do sound better.
However these AC pulse caps aren't always available in the technology that would be ideal for the application.
Often these caps use (or at least used to) paper as a dielectricum, were only available in small values and high voltages etc.
I write used to, as I haven't checked this in a long time.
I can think of a few audio equipment manufacturers that deliberately forewent better dielectrical properties of coupling caps and instead chose to use AC pulse caps because they sounded better in that application.
Usually they were put as a driver to output cap in tube amps where DA is of lesser importance, IIRC.
One aspect that clearly shows their superiority is better stereo imaging for instance.
Nowadays most special audio caps ( AKA "boutique caps) implement Mrs Jung and Marsh's theories with great success but I think the ideal audiocap is still evolving.
Even if it can be built today it's still not economical to go to market with such a range of caps today IMO.
Fortunately the gap is closing fast, both film and electrolytic caps have improved tremendously over the past fifteen years.
Probably regardless of what the audio market said anyway...
Cheers,
one very low distortion, comparatively inexpensive cap:
I got these from Mouser -- XICON 4.7uF/100V non-inductive metallized polyster -- no measurable distortion (to the extent that I can measure below 0.001% -- Batemen's device adds another degree of precision, to 0.0001%).
For this unit, D measures 0.03.
http://www.mouser.com/catalog/620/524.pdf
i wouldn't recommend them for "active filters" unless you have an acurate capacitance bridge -- the values range from 4.5uF to 4.9uF --
I got these from Mouser -- XICON 4.7uF/100V non-inductive metallized polyster -- no measurable distortion (to the extent that I can measure below 0.001% -- Batemen's device adds another degree of precision, to 0.0001%).
For this unit, D measures 0.03.
http://www.mouser.com/catalog/620/524.pdf
i wouldn't recommend them for "active filters" unless you have an acurate capacitance bridge -- the values range from 4.5uF to 4.9uF --
Polyester is problematic. It doesn't measure non-linear distortion so much, BUT it measures significant LINEAR distortion (DA).
For the record, I used to think that polyester was OK. Then, Noel Lee, then future father of Monster cable was building one of my electronic xover designs, about 25 years ago, and he found that REMOVING the polyester caps made the sound better. I, at first, didn't believe him, BUT further work with Dick Marsh showed me that it was DA that was the problem. I immediately went over to polystyrene and polypropylene caps.
For the record, I used to think that polyester was OK. Then, Noel Lee, then future father of Monster cable was building one of my electronic xover designs, about 25 years ago, and he found that REMOVING the polyester caps made the sound better. I, at first, didn't believe him, BUT further work with Dick Marsh showed me that it was DA that was the problem. I immediately went over to polystyrene and polypropylene caps.
john curl said:
I would like someone to explain the chemistry and its relation to sound to me. Polypropylene is purely aliphatic, there is nothing "polar" in it whereas polyester, aka polyethylene terephthalate isn't.
I have quite a collection of TRW polystyrene caps, their D is negligible (measures 0.0001 on my Genrad Digibridge) but distortion is higher than PET or PP.
Mica caps have very, very low D -- the highest I measured in a mixed lot of 20 was 0.0003 -- these range from very modern to WW-II vintage (yes, I have been a ham radio operator off and on) -- the distortion is usually 0.05% to 0.06%
Got a bunch of ceramics -- from WW-II to surface mount -- the old ones should probably be donated to a museum. I have to test these yet.
I am interested in the "why" of listening results.
for high frequency compensation obviously the cap construction has to have low inductance/high self resonant freq
this leaves out many rolled film-foil types, "stacked metallized film" caps are available in several film dielectrics, PPS is presumably better than mylar but so is NPO/COG ceramic
it is unlikely that the feedback cap would add distortion comparable to the c-b reverse biased diode capacitance – as long as you avoid hi-K ceramics
this leaves out many rolled film-foil types, "stacked metallized film" caps are available in several film dielectrics, PPS is presumably better than mylar but so is NPO/COG ceramic
it is unlikely that the feedback cap would add distortion comparable to the c-b reverse biased diode capacitance – as long as you avoid hi-K ceramics
jackinnj
I am refering to the compensation capacitor(s) needed in power amp designs.
But every op amp also needs such capacitors internally or extermally.
If the pole rolloff starts at few hundred Hz I would think mica is better than ceramic. I have lots of ceramics but n mica ones so I am asking if there are any sonic benefits to mica caps.
Tom
I am refering to the compensation capacitor(s) needed in power amp designs.
But every op amp also needs such capacitors internally or extermally.
If the pole rolloff starts at few hundred Hz I would think mica is better than ceramic. I have lots of ceramics but n mica ones so I am asking if there are any sonic benefits to mica caps.
Tom
I will probably get disagreement on this, but NPO ceramics are supposed to be pretty good. As disc caps, they are usually only available in values < about 1000pF. They are denoted by an NPO on the capacitor body or a black band at the top of the cap.
For values of a few pF, a "gimmick" can be used, consisting of a couple of pieces of teflon insulated wire twisted tightly together and cut to length for proper capacitance. When I do this, I usually put some heat shrink around the wires to make sure they stay tightly twisted.
For values of a few pF, a "gimmick" can be used, consisting of a couple of pieces of teflon insulated wire twisted tightly together and cut to length for proper capacitance. When I do this, I usually put some heat shrink around the wires to make sure they stay tightly twisted.
the gimmick is what we used to trim the frequency of a crystal oscillator --
this is cut and paste from the QRP website:
"Nxxx" on a capacitor specifies the nominal change in capacitance with temperature; N750 means -750PPM/deg, N1500 means -1500PPM/deg, etc. The value is not exact and there will be some variation between capacitors carrying the same designation. These types may also carry an "EIA" designation; the more common ones are:
P2G ® N150
T2H ® N470
U2J ® N750
P3K ® N1500
Nxxx disk ceramics may also be designated by a colored dot on the top:
Black ® NP0
Red ® N075
Orange ® N150
Blue ® N470
Violet ® N750
this is cut and paste from the QRP website:
"Nxxx" on a capacitor specifies the nominal change in capacitance with temperature; N750 means -750PPM/deg, N1500 means -1500PPM/deg, etc. The value is not exact and there will be some variation between capacitors carrying the same designation. These types may also carry an "EIA" designation; the more common ones are:
P2G ® N150
T2H ® N470
U2J ® N750
P3K ® N1500
Nxxx disk ceramics may also be designated by a colored dot on the top:
Black ® NP0
Red ® N075
Orange ® N150
Blue ® N470
Violet ® N750