It seems like everyone has their own religion about capacitors. What I find amusing is some simply negate the advantages conferred by their high grade caps by poor implementation.
In decoupling power rails to active devices, I read posts about.....
1) Must use Oscon caps.
2) Must use Panasonic caps.
3) Better to add small resistance in series with cap, to act as filter.
4) Oscons do not like to be bypassed with small value caps.
OK but then they....
1) Use sockets to try out various caps, with enthusiastic reports about the sound. I thought the advantage of those caps were primarily low ESR (what they were designed for).
2) Add series resistance to low ESR caps at the output of regulators. The idea was meant for the main raw supply capacitors and also as local decoupling for individual loads (i.e. opamps). Not at the output of regulators. Some linear regs that don't like low ESR are better off with regular caps.
3) Choose not to bypass electrolytics.
Y5V caps suck bad. I use X6S for bulk decoupling and X7R for high speed bypass. SMD's are the way to go (Howard Johnson's books will tell you why). Even how the short trace and via going to each cap has an impact. Before anyone goes and dismiss this as only applicable for high freq work, there are plenty of high end manufacturers who use high freq EE techniques in their designs. Ceramics are susceptible to temp variations, so do size them correctly. For audio work, I use film PPS SMD or silver mica for the pF range. Been pretty happy with the results.
Thank you both John C and Charlie H for sharing their experiences.
In decoupling power rails to active devices, I read posts about.....
1) Must use Oscon caps.
2) Must use Panasonic caps.
3) Better to add small resistance in series with cap, to act as filter.
4) Oscons do not like to be bypassed with small value caps.
OK but then they....
1) Use sockets to try out various caps, with enthusiastic reports about the sound. I thought the advantage of those caps were primarily low ESR (what they were designed for).
2) Add series resistance to low ESR caps at the output of regulators. The idea was meant for the main raw supply capacitors and also as local decoupling for individual loads (i.e. opamps). Not at the output of regulators. Some linear regs that don't like low ESR are better off with regular caps.
3) Choose not to bypass electrolytics.
Y5V caps suck bad. I use X6S for bulk decoupling and X7R for high speed bypass. SMD's are the way to go (Howard Johnson's books will tell you why). Even how the short trace and via going to each cap has an impact. Before anyone goes and dismiss this as only applicable for high freq work, there are plenty of high end manufacturers who use high freq EE techniques in their designs. Ceramics are susceptible to temp variations, so do size them correctly. For audio work, I use film PPS SMD or silver mica for the pF range. Been pretty happy with the results.
Thank you both John C and Charlie H for sharing their experiences.
Hi John,
Funny story John, I have a somewhat similar one...
Having finished my degree I had originally expected to be able to continue my studies and go into a field that was quite exciting at the time, called in former eastern germany "Cybernetics" or perhaps work in the military research place I did my practicum.
Alas the government had decided that I was politically unreliable, so not only could I not continue my studies despite being in the class's top 3, I could not actually find a decent job in the entire industry in the country with my degree. Everyone was desperate for electronic engineers, yet as soon as they got my personal file they said "Sorry, we do not have any positions available.".
I was almost at the point where my failure to hold a job would have made me liable to a 6 month prison sentence for anti-social behaviour, when a small private company (there where a few, but very few in East Germany, all limited to less than 10 Employees) gave a job despite the black mark in my file.
As it so happened, they made a fair bit of the audio equipment for the east german Radio/TV/Recording Industry, east germany being a small country a few small companies could do most of it. The main products where modular mixer systems, quite similar to the stuff TAB and Neumann as well as other did in west germany.
When I started this was all still completely discrete, much of the circuitry quite reminiscent of the early transistor Neve's and TAB's etc.. The problem was, the customers wanted better EQ facilities and the Boss wanted less labour intensive production, so I was to re-designed this stuff from ground up, first starting with the master modules.
I redesigned with Op-Amp's, using externally compensated ones in mil-spec cerdip housing, split rails, much fewer coupling caps', the miximimg bus was true zero ohm and it measured so much better (THD, frequency response) than our standard modules, I was totally psyched (fresh from the Uni and I was showing these old hands how it is done!) and the boss was really excited.
Then we plugged the new module into the mixer mainframe in our own little listening/evaluation studio. We played some music from a Tape. We looked at each and our faces showed bewilderment and disgust. The new mastering module sounded awful. I mean REALLY BAD.
Over time we came up with hybrids of the old discrete circuitry and a few carefully selected op-amp's where they where of use, I was able to rationalise some of the circuitry by going from single rail to split rail, loosing some coupling caps in the process, using transistor arrays instead of single transistors and so on and got it all to sound good, but it was a hard learning process for the whole company...
A few years later east germany collapsed and they shut doors shortly after. The studio's threw out the old gear by the ton, including excellent discrete and tubed gear, Schulze Coaxial monitors and so much more. Seeing the Skip's at one of them and what was in it I nearly cried in public.
Things have moved on much from those days. But I never forget my first original failure.
Ciao T
Funny story John, I have a somewhat similar one...
Having finished my degree I had originally expected to be able to continue my studies and go into a field that was quite exciting at the time, called in former eastern germany "Cybernetics" or perhaps work in the military research place I did my practicum.
Alas the government had decided that I was politically unreliable, so not only could I not continue my studies despite being in the class's top 3, I could not actually find a decent job in the entire industry in the country with my degree. Everyone was desperate for electronic engineers, yet as soon as they got my personal file they said "Sorry, we do not have any positions available.".
I was almost at the point where my failure to hold a job would have made me liable to a 6 month prison sentence for anti-social behaviour, when a small private company (there where a few, but very few in East Germany, all limited to less than 10 Employees) gave a job despite the black mark in my file.
As it so happened, they made a fair bit of the audio equipment for the east german Radio/TV/Recording Industry, east germany being a small country a few small companies could do most of it. The main products where modular mixer systems, quite similar to the stuff TAB and Neumann as well as other did in west germany.
When I started this was all still completely discrete, much of the circuitry quite reminiscent of the early transistor Neve's and TAB's etc.. The problem was, the customers wanted better EQ facilities and the Boss wanted less labour intensive production, so I was to re-designed this stuff from ground up, first starting with the master modules.
I redesigned with Op-Amp's, using externally compensated ones in mil-spec cerdip housing, split rails, much fewer coupling caps', the miximimg bus was true zero ohm and it measured so much better (THD, frequency response) than our standard modules, I was totally psyched (fresh from the Uni and I was showing these old hands how it is done!) and the boss was really excited.
Then we plugged the new module into the mixer mainframe in our own little listening/evaluation studio. We played some music from a Tape. We looked at each and our faces showed bewilderment and disgust. The new mastering module sounded awful. I mean REALLY BAD.
Over time we came up with hybrids of the old discrete circuitry and a few carefully selected op-amp's where they where of use, I was able to rationalise some of the circuitry by going from single rail to split rail, loosing some coupling caps in the process, using transistor arrays instead of single transistors and so on and got it all to sound good, but it was a hard learning process for the whole company...
A few years later east germany collapsed and they shut doors shortly after. The studio's threw out the old gear by the ton, including excellent discrete and tubed gear, Schulze Coaxial monitors and so much more. Seeing the Skip's at one of them and what was in it I nearly cried in public.
Things have moved on much from those days. But I never forget my first original failure.
Ciao T
Hi Charles,
-Chris
I was thinking more along the lines of a bad batch or some such.
Generally towards smaller parts. That means high-K ceramic dielectrics (= a bad thing!) For electrolytic caps, you want to buy the ones with the largest case size for a given CV. If that means going higher on voltage, do it. Everything is meant as "within reason" of course!
The change is not subtle at all, as you've noted. I find the same results when rebuilding audio equipment. Of course, there are other changes one can make as well that will elevate the performance.
Fair enough. Certainly you know all you need to know.
-Chris
Good input, T. What was your 'problem' with the state, if you can say?
Personally, I lived in Berkeley in the 1960's, and everybody was considered ' a communist sympathizer' and I almost had similar problems.
I was only in East Berlin, once, in 1967. Still saw bullet holes in buildings, and darn near got in trouble, not getting out of the way of some top communist officials. However, they apparently decided that I was harmless, and got into their limos.
Personally, I lived in Berkeley in the 1960's, and everybody was considered ' a communist sympathizer' and I almost had similar problems.
I was only in East Berlin, once, in 1967. Still saw bullet holes in buildings, and darn near got in trouble, not getting out of the way of some top communist officials. However, they apparently decided that I was harmless, and got into their limos.
Sorry but I cant resist, someone mentioned PCB materials earlier, let me guess, this can influence sound
Do the different grades of FR4 and fibre weave patterns also have an effect then!
Again MLCCs have raised thier head, sorry but for high speed decoupling of digital X7R and COG are tops, due to small size and low inductance design, to have any other view is denial of the requirements of decoupling digital circuitry.
Add Henry Ott and Eric Bogatin to the list of reading.
Do the different grades of FR4 and fibre weave patterns also have an effect then!
Again MLCCs have raised thier head, sorry but for high speed decoupling of digital X7R and COG are tops, due to small size and low inductance design, to have any other view is denial of the requirements of decoupling digital circuitry.
Add Henry Ott and Eric Bogatin to the list of reading.
John,
Well, I was somewhat involved in one of the loose, church based groups that mainly lobbied for nuclear disarmament on the principle that both sides need to make concessions, not just the evil empire in the west and that even 1/10 of nukes the eastern block admitted to having was enough for credible defence and deterrence. Plus, there was some talk that east germany needed internal reforms. There where loads of these in the country.
It was all quite tame, non of us there really where major activists, once when we wanted to attend a planned human chain from the russian to american embassy in east berlin for peace and nuclear disarmament, we all got arrested by the "state security police". After the collapse of the east german regime it was possible for citizens to get a look at their file, formerly held by the secret police, I found it had been the Church's pastor who had informed on us to the secret police. All quite tawdry, sordid and not much to talk about.
It's not like I was an activist of any sorts or a serious dissident. Merely a dude thinking that perhaps we/I could contribute a bit more to society than just "Kuschen und Strammstehen" (Kowing and standing to attention).
Who knows, without these troubles I may have never ended in Audio, but might have instead worked with Manfred von Ardenne. Not sure which would have been more fun.
Well, ancient history from the last millenium. Water down the bridge and all....
Ciao T
Well, I was somewhat involved in one of the loose, church based groups that mainly lobbied for nuclear disarmament on the principle that both sides need to make concessions, not just the evil empire in the west and that even 1/10 of nukes the eastern block admitted to having was enough for credible defence and deterrence. Plus, there was some talk that east germany needed internal reforms. There where loads of these in the country.
It was all quite tame, non of us there really where major activists, once when we wanted to attend a planned human chain from the russian to american embassy in east berlin for peace and nuclear disarmament, we all got arrested by the "state security police". After the collapse of the east german regime it was possible for citizens to get a look at their file, formerly held by the secret police, I found it had been the Church's pastor who had informed on us to the secret police. All quite tawdry, sordid and not much to talk about.
It's not like I was an activist of any sorts or a serious dissident. Merely a dude thinking that perhaps we/I could contribute a bit more to society than just "Kuschen und Strammstehen" (Kowing and standing to attention).
Who knows, without these troubles I may have never ended in Audio, but might have instead worked with Manfred von Ardenne. Not sure which would have been more fun.
Well, ancient history from the last millenium. Water down the bridge and all....
Ciao T
Hi,
Yes. Compared once, all else really being equal, but arguably using quite high impedance circuits, PTFE Board was much better than FR4.
Ciao T
Sorry but I cant resist, someone mentioned PCB materials earlier, let me guess, this can influence sound
Yes. Compared once, all else really being equal, but arguably using quite high impedance circuits, PTFE Board was much better than FR4.
Ciao T
Another factor could have been the source impedance. If you sourced your amp with a tube amp with say 1k Zout, that filter now cuts in at 10kHz or so, and that IS audible.
The 150 ohms and 10nF is in any case an unlucky combination with it's sensitivity to source Z.
jan didden
I swear that I don't understand why someone has to explain away removing a CERAMIC cap and hearing a difference. I used to to do it just to MOD, Sony audio equipment. Swap out the CERAMICS for something different, and 'improved audio quality'.
I realize that many of you have never tested caps using the Tektronix method of modifying a 577 curve tracer and externally driving it with a function generator, but if you ever did, you would understand that most ceramics are fatally flawed, with a 'non return to zero' defect easily seen in the larger caps.
Distortion measurement by either IM or harmonic distortion methods is also possible, but you have to know what frequencies and method works best, to see large distortion.
The ear picks it up rather easily, apparently.
I realize that many of you have never tested caps using the Tektronix method of modifying a 577 curve tracer and externally driving it with a function generator, but if you ever did, you would understand that most ceramics are fatally flawed, with a 'non return to zero' defect easily seen in the larger caps.
Distortion measurement by either IM or harmonic distortion methods is also possible, but you have to know what frequencies and method works best, to see large distortion.
The ear picks it up rather easily, apparently.
Last edited:
I find live amplified rock a social experience not a sonic one (at least without earplugs). I have several friends who love close seats and always wear them. My comment was intended as an exaggeration. I would not use live rock music to test amplifier architectures. Why do you think tubes with their overdrive characteristics are prefered?
In overload, common in music production and performance, tubes distort. They may distort differently than some solid state circuits and that distortion may be preferred by some performers. But their linearity short of overload is probably not relevant, since solid state circuits can be just as linear or even more so.
Interesting point about musicians John.
I have a musician friend come over and listen sometimes. His hearing has been damaged some over the years, and yet I played a recording he didn't know at all and he identifies the Bass player. He listens to different things than I do but always has a valuable comment. You have to have some ears to trust besides your own.
I have a musician friend come over and listen sometimes. His hearing has been damaged some over the years, and yet I played a recording he didn't know at all and he identifies the Bass player. He listens to different things than I do but always has a valuable comment. You have to have some ears to trust besides your own.
Classical musicians are an interesting bunch: They usually do not need real hi fi, because they can listen to a performance through a good quality clock radio and get more from the performance than I ever could. However, when it comes to their personal instrument, they are VERY touchy. Often, they will invest 'everything' into it.
Sy,
So, all the classical recording of the "golden age" that where made with tubes (Mercury Living Presence, RCA Living Stereo's, Decca's et al) used overload?
I think not.
In fact, overload is used only in a few specific genres of contemporary music, emphatically not in others.
So it is not helpful to lump everything together. In fact, only for E-guitar and to a lesser degree for E-Bass and some keyboards is appreciable overload used routinely.
It is extremely rarely if ever used on Vocals, Brass and Woodwind's, Drumkits, classical orchestras and so on. Hence I would actually say that overload is rare in music production, with the exception of electronically amplified stringed instruments, where it is part of the "tone" and artistic expression.
Ciao T
So, all the classical recording of the "golden age" that where made with tubes (Mercury Living Presence, RCA Living Stereo's, Decca's et al) used overload?
I think not.
In fact, overload is used only in a few specific genres of contemporary music, emphatically not in others.
So it is not helpful to lump everything together. In fact, only for E-guitar and to a lesser degree for E-Bass and some keyboards is appreciable overload used routinely.
It is extremely rarely if ever used on Vocals, Brass and Woodwind's, Drumkits, classical orchestras and so on. Hence I would actually say that overload is rare in music production, with the exception of electronically amplified stringed instruments, where it is part of the "tone" and artistic expression.
Ciao T
- Status
- Not open for further replies.