I am restoring a Harman Kardon A50K (kit version of A500) tube amp. A first tube project and first audio project for me. As I replace aged power supply caps and signal filter caps I am finding the art of cap selection mysterious. Armed with a Mouser catalog and online access to antique audio web sites, I just cant figure out the mistique of cap selection, and think maybe it is part real, part myth (bunk!). People advise that when replacing a .25 uf signal cap with an available .22 uf that there will be a significant change in frequency response. And that JJ caps will perform better for audio than another cap. I don't get it. I would think the thing to look for is max life at high temp, tolerance, capacitance coefficient to heat. And since the life is certified at voltage,one would tend to choose a much over-rated voltage to get more life.
How does one go about selecting caps for audio circuits, and how much is myth?
How does one go about selecting caps for audio circuits, and how much is myth?
It seems to me that the audiopholes in this forum can have two caps with identical published specs and say that the audio quality can be very different. If this is true, then there are characteristics that the specs do not identify. What is it that makes a cap sound better than another when it has the same published spec, assuming the spec is accurate?
I would like to understand this before I come to believe it is really smoke and mirrors, bragging rights, and misconceptions. I tend to believe there really is something more, but why is it that no one knows what it is?
I would like to understand this before I come to believe it is really smoke and mirrors, bragging rights, and misconceptions. I tend to believe there really is something more, but why is it that no one knows what it is?
Datasheet will show cap value, tolerance, reliability rating for a given voltage/temp combination, a temperature coefficient rating to show consistency over temp range, dissipation factor (whit is this?), CV (dont know what this is), impedance/equivalent series resistance, leakage current, ESR (what is this?), DCL, an on and on.
So why in these many responses on this topic doesnt anyone offer any info with real value? There are a number of threads, but nothing of substance to indicate what makes a good audio cap - if it is indeed not something that can be captured by a spec.
Maybe in many cases people perceive differences that are not real.
The previous response: "published where - on the side of the caps" is not helpful. Of course, the manufacturers provide datasheets - online and easy to access.
I have very limited expertise here and am looking for someone to share some experience. I have not problem with electronics and am an engineer, but am frustrated by the fact that so many people pretend to know about things they dont understand and knowing that there are so many really smart, knowledgeable people that could help with this question am frustrated that no one is willing to serioulsy offer some quidance.
I am about to spend over $100 on parts to restore this amp - mostly for a few caps, and I have no idea if these "audio grade" caps are really any better than more common industrial ones, that can go head to head in terms of the detailed psecs notes above. I am learning that the material a cap is made of makes an audio difference, but again, this could be that the material impacts the published spec and the real performance can be understood from the spec. The datasheet has al ot to offer - does this only get us part of the way to understanding the audio quality, and if so, what is it that additional specs, say an audio spec, would say?
So why in these many responses on this topic doesnt anyone offer any info with real value? There are a number of threads, but nothing of substance to indicate what makes a good audio cap - if it is indeed not something that can be captured by a spec.
Maybe in many cases people perceive differences that are not real.
The previous response: "published where - on the side of the caps" is not helpful. Of course, the manufacturers provide datasheets - online and easy to access.
I have very limited expertise here and am looking for someone to share some experience. I have not problem with electronics and am an engineer, but am frustrated by the fact that so many people pretend to know about things they dont understand and knowing that there are so many really smart, knowledgeable people that could help with this question am frustrated that no one is willing to serioulsy offer some quidance.
I am about to spend over $100 on parts to restore this amp - mostly for a few caps, and I have no idea if these "audio grade" caps are really any better than more common industrial ones, that can go head to head in terms of the detailed psecs notes above. I am learning that the material a cap is made of makes an audio difference, but again, this could be that the material impacts the published spec and the real performance can be understood from the spec. The datasheet has al ot to offer - does this only get us part of the way to understanding the audio quality, and if so, what is it that additional specs, say an audio spec, would say?
Before you get too wound up, start by reading Walt Jung and Richard Marsh's capacitor selection articles. They were originally published in Audio magazine, but are available on the web. Try www.capacitors.com or Walt's website. He's a member here (WaltJ).
Yes, there are things that don't show up in the standard measurements. Jung & Marsh went a long ways towards proving what audiophiles already knew. There's more work to be done, but they laid a very strong foundation. That will go a long way towards clearing up some of the so-called 'myths' about capacitors.
The upshot is that caps make a tremendous difference in sound quality. The prices charged for some of the good caps are scandalous, but at least you can go in knowing what you're getting into. Do not let the sound quality question and the price issue get confused in your mind. They are two separate matters.
Once you've done your homework on capacitor selection, chosing values for DC blocking caps is a simple matter of plugging numbers into a simple formula and cranking the handle:
F=1/(2*PI*R*C)
where:
F= frequency
PI=3.14159
R=impedance in Ohms
C=capacitance in Farads (note that microfarads are 1e-6)
Most caps are 10% tolerance, so the difference between .22 and .25uF isn't going to be that big a problem. For DC blocking caps, bigger is usually better (low frequency response improves), although you will sometimes run across a circuit that acts funny if you put in too large a value. Don't fret about that, it's rare. Putting film bypasses on the power supply is a good idea, as electrolytics have lousy high frequency characteristics. Ignore those who scream that bypasses in power supplies are bad.
For that matter, ignore anyone who screams, period.
Grey
Yes, there are things that don't show up in the standard measurements. Jung & Marsh went a long ways towards proving what audiophiles already knew. There's more work to be done, but they laid a very strong foundation. That will go a long way towards clearing up some of the so-called 'myths' about capacitors.
The upshot is that caps make a tremendous difference in sound quality. The prices charged for some of the good caps are scandalous, but at least you can go in knowing what you're getting into. Do not let the sound quality question and the price issue get confused in your mind. They are two separate matters.
Once you've done your homework on capacitor selection, chosing values for DC blocking caps is a simple matter of plugging numbers into a simple formula and cranking the handle:
F=1/(2*PI*R*C)
where:
F= frequency
PI=3.14159
R=impedance in Ohms
C=capacitance in Farads (note that microfarads are 1e-6)
Most caps are 10% tolerance, so the difference between .22 and .25uF isn't going to be that big a problem. For DC blocking caps, bigger is usually better (low frequency response improves), although you will sometimes run across a circuit that acts funny if you put in too large a value. Don't fret about that, it's rare. Putting film bypasses on the power supply is a good idea, as electrolytics have lousy high frequency characteristics. Ignore those who scream that bypasses in power supplies are bad.
For that matter, ignore anyone who screams, period.
Grey
ransom peek said:
What does high ESR sound like?
My responses may not seem helpful to you, but my point is that either you trust your ears or you don't or you fall somewhere in between.
If "audio grade" caps seem like nonsense to you, don't buy them. If you're curious, investigate either by listening comparisons or by measuring. If you're willing to take someone's word for it, then ask people what they like to use and copy what they do.
You aren't going to convince anyone that they are delusional because they can't establish the existance of an "audio grade" cap through specifications, whether published or not. People have tried and nothing comes of it but arguing. I'm plenty guilty myself.
See the linked thread below for examples of useful measurements and pointless argument. Unfortunately, I have only contributed to the latter.
http://www.diyaudio.com/forums/showthread.php?goto=newpost&threadid=48942
Btw, I used to live in Ft Collins before I moved to Australia. Ft Collins is a brilliant place to live.
ransom peek,
Give this a read:
http://www.capacitors.com/picking_capacitors/pickcap.htm
Mostly folks can't tell you how to choose what sounds good to you in your circuit and topology. Same amp, different speakers, different sound.
Then there's always the religion that tubes sound better than transistors and vice versa. Once you start down the slippery slope of making amps/speakers/pre-amps and so forth, you are making that hole in the water that boat owners/sailors throw their money in.
Prosit
Give this a read:
http://www.capacitors.com/picking_capacitors/pickcap.htm
Mostly folks can't tell you how to choose what sounds good to you in your circuit and topology. Same amp, different speakers, different sound.
Then there's always the religion that tubes sound better than transistors and vice versa. Once you start down the slippery slope of making amps/speakers/pre-amps and so forth, you are making that hole in the water that boat owners/sailors throw their money in.
Prosit
These sites have so much information, they can be overwhelming! There is not much in the way of summary info, but I understand the detail to show:
1. Dialectric choice is biggest factor is audio performance
2.The Dialectric material affects critical parameters that affect audio, and if these parameters are known, you can predict audio quality to a great extent
Surely manufacturer can, using identical materials, produce varying degrees of results through each design, and quality of materials and workmanship. So, I will choose caps fro my Harman Kardon A50K (A500) restoration based on dialectric material and recommended manufacturers and recommended devices. Throwing the extra money in the hole (thanks "acenovelty" for clarifying the terminolgy for this technique!) will be well spent. I am sure there is more to it than this, but at least I have moved on from newbie to beginner.
Thanks to all for your patience with me and the excellent information.
Ransom Peek
1. Dialectric choice is biggest factor is audio performance
2.The Dialectric material affects critical parameters that affect audio, and if these parameters are known, you can predict audio quality to a great extent
Surely manufacturer can, using identical materials, produce varying degrees of results through each design, and quality of materials and workmanship. So, I will choose caps fro my Harman Kardon A50K (A500) restoration based on dialectric material and recommended manufacturers and recommended devices. Throwing the extra money in the hole (thanks "acenovelty" for clarifying the terminolgy for this technique!) will be well spent. I am sure there is more to it than this, but at least I have moved on from newbie to beginner.
Thanks to all for your patience with me and the excellent information.
Ransom Peek
Ransom, allow me to give a simplified view of capacitor sonics. The dielectric constant (DC) of a capacitor's dielectric correlates well with the capacitor's AC accuracy (ie, its sound). Dielectric absorption (DA) is a function of DC, and DA seems to be the main or at least the worst sounding non-linearity capacitors introduce into a circuit, all other things being equal. The lower the DC, the lower the DA, the better the sound---again, all other things being equal.
These observations are not controversial and are well-known to cost-no-object industries like military and aerospace.
Capacitors therefore rank, sonically, as a function of their DC/DA: air (DC = 1) sounds better than teflon (2.1 DC) sounds better than polypropylene (2.2 DC) sounds better than polyester/mylar (3+ DC) sounds better than electrolytic (8+ DC) etc. Certain adjustments to a capacitor's parameters (metallized vs foil plate etc) can hone in (ie, best utilize or least diminish) a given capacitor's best functioning. Hence, careful attention to quality control and materials selection and winding processes etc are important.
The main problem with lower DC is the attendant increase in size of a capacitor of a given capacitance. Size does matter for many an important design consideration.
These observations are not controversial and are well-known to cost-no-object industries like military and aerospace.
Capacitors therefore rank, sonically, as a function of their DC/DA: air (DC = 1) sounds better than teflon (2.1 DC) sounds better than polypropylene (2.2 DC) sounds better than polyester/mylar (3+ DC) sounds better than electrolytic (8+ DC) etc. Certain adjustments to a capacitor's parameters (metallized vs foil plate etc) can hone in (ie, best utilize or least diminish) a given capacitor's best functioning. Hence, careful attention to quality control and materials selection and winding processes etc are important.
The main problem with lower DC is the attendant increase in size of a capacitor of a given capacitance. Size does matter for many an important design consideration.
That's really helpful serengettiplains. Now I also read from the previoulsy referenced websites that
"up to values of 10,000 pf polystyrene is the best all around choice, as it has reasonable size and is readily avialable in many sizes with tight tolerances available. Above 10,000 pf and up to 0.1, it can still be used but is harder to obtain. Above 01. uf polypropylene (or metalized polypropylene) is the dielectric of choice, as it has hearly the same relative qualities of DF and DA as polystyrene. Tight tolerances are available (but will be special order) and you can get capacitors up to 10 uf or more."
serengetiplinas - for some reason you dont reference polystyrene...
"up to values of 10,000 pf polystyrene is the best all around choice, as it has reasonable size and is readily avialable in many sizes with tight tolerances available. Above 10,000 pf and up to 0.1, it can still be used but is harder to obtain. Above 01. uf polypropylene (or metalized polypropylene) is the dielectric of choice, as it has hearly the same relative qualities of DF and DA as polystyrene. Tight tolerances are available (but will be special order) and you can get capacitors up to 10 uf or more."
serengetiplinas - for some reason you dont reference polystyrene...
It would be helpful if you were to first understand how a particular cap is used in a circuit. At least determine whether or not it is in the signal path.
One hint toward understanding: think of a cap as a resistor whose resistance value goes down as frequency goes up. The Capacitor's value determins the frequency where the resistance starts to fall. The lower the cap's value the higher starting frequency. Now review what you know about voltage dividers and look for cap-resistor combinations that form a "frequency dependent voltage divider".
In modern SS amps very few cap are actually directly in the signal path. These are mostly DC blockers and would seem to be the ones where audiophile concerns are most relevant, although sometimes the simple expediant of substantially increasing the capacitor value trumps nuances of construction. In some other locations in the circuit the the capacitor is not directly in the signal path (although not totally dissaciated with it) but has such a small capacitance value that it it can only affect signal far above the audio band. In these cases, it is possible that there is very little tolerance with regard to the value selected. Then there are caps used as filter in XOs, phono pre-amps, EQs. These are in the signal path and audio qualiity is a consideration.
Don't give too much regard to statements like "ceramic caps are bad" or "electrolytic caps are bad". Most of the applications for these are one where the issue is moot -- no alternative type exists. There are areas where there is an overlap of types, but even there cost and size may be as important as anything else.
One hint toward understanding: think of a cap as a resistor whose resistance value goes down as frequency goes up. The Capacitor's value determins the frequency where the resistance starts to fall. The lower the cap's value the higher starting frequency. Now review what you know about voltage dividers and look for cap-resistor combinations that form a "frequency dependent voltage divider".
In modern SS amps very few cap are actually directly in the signal path. These are mostly DC blockers and would seem to be the ones where audiophile concerns are most relevant, although sometimes the simple expediant of substantially increasing the capacitor value trumps nuances of construction. In some other locations in the circuit the the capacitor is not directly in the signal path (although not totally dissaciated with it) but has such a small capacitance value that it it can only affect signal far above the audio band. In these cases, it is possible that there is very little tolerance with regard to the value selected. Then there are caps used as filter in XOs, phono pre-amps, EQs. These are in the signal path and audio qualiity is a consideration.
Don't give too much regard to statements like "ceramic caps are bad" or "electrolytic caps are bad". Most of the applications for these are one where the issue is moot -- no alternative type exists. There are areas where there is an overlap of types, but even there cost and size may be as important as anything else.
Polystyrene has some quirky dielectric characteristics I don't understand, and manufacture of polystyrene, I think, is being phased out, making styrene caps more difficult to find. I think in the general ranking, polystyrene sits between teflon and polypropylene. That is both my experience and my quasi-educated opinion having, among other things, spoken with certain capacitor engineers at a reputable military firm.
I also did not mention paper-in-oil and wax. Certain waxes have very low DC/DA, even < teflon, and thus boast excellent sonics. Jupiter caps, made with beeswax, are well regarded among tube aficionados.
Paper likewise can have a low DC, as can certain forms of oil. PIO capacitors, moreover, can benefit from the heat and vibration reducing effects of their liquid dielectric. Capacitor plates *do* move, despite being wound under tension, and the internal damping offered by a liquid dielectric can be beneficial.
One important non-dielectric factor is plate material. Metallized plates sound worst, IME, as they exhibit higher resistance and noise (noise because plate deposition is via a not-too-exact sputtering, creating discontinuities etc). Aluminum plates are better than metallized, but aluminum cannot be affixed to copper, requiring an intermediate substance like zinc. Inserting aluminum-zinc-copper connections into your circuit is not advised. Aluminum also does not seem to sound good as a conductor: note the lack of use of which in any form of audio gear.
Tin plates are better than aluminum because can be directly soldered to the lead.
Copper plates .... I'm skeptical of their value outside their use in PIO caps. Copper oxidizes, and over a given time a copper plate can be expected to degrade significantly, except perhaps in PIO caps for obvious reasons.
Silver plates .... ahh, what can I say? If you have the money ....
I also did not mention paper-in-oil and wax. Certain waxes have very low DC/DA, even < teflon, and thus boast excellent sonics. Jupiter caps, made with beeswax, are well regarded among tube aficionados.
Paper likewise can have a low DC, as can certain forms of oil. PIO capacitors, moreover, can benefit from the heat and vibration reducing effects of their liquid dielectric. Capacitor plates *do* move, despite being wound under tension, and the internal damping offered by a liquid dielectric can be beneficial.
One important non-dielectric factor is plate material. Metallized plates sound worst, IME, as they exhibit higher resistance and noise (noise because plate deposition is via a not-too-exact sputtering, creating discontinuities etc). Aluminum plates are better than metallized, but aluminum cannot be affixed to copper, requiring an intermediate substance like zinc. Inserting aluminum-zinc-copper connections into your circuit is not advised. Aluminum also does not seem to sound good as a conductor: note the lack of use of which in any form of audio gear.
Tin plates are better than aluminum because can be directly soldered to the lead.
Copper plates .... I'm skeptical of their value outside their use in PIO caps. Copper oxidizes, and over a given time a copper plate can be expected to degrade significantly, except perhaps in PIO caps for obvious reasons.
Silver plates .... ahh, what can I say? If you have the money ....
BTW, re Sam9's comments regarding what's "in series" with the signal, you might want to read something like Lynn Olson's presentation on current loops. IMO, everything is in some degree in series with the signal, even that lousy aluminum wire connecting AC to your house. Power supply capacitors, to the degree their AC non-linearities are not cancelled by perfect differential operation, are certainly in series with the signal.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.