Any schematics for those PS regulators?
Hearing sound diferences, if your system is transparent enough, should be fine. The question is whether you like or dislike those differences. Or which do you prefer.
Your welcome Karl. Yes Jap verion more evenly balance. The once that you buy from digikey or mouser is more splashy on the treble & bass sounds fatter & does not go that deep. Oh on some applications the sound is more forward.
Strange that pic shows Brown with Gold print Johnego. Could be earlier version when transition was made from Silmic 1 to Silmic ll
Hi Carlmart
Sorry no schema, the regs came ready build. Your right its a matter of whether we like the sound. It’s just kind of head scratching for me that you can hear the difference even when regs are only powering the FGA chip.
Strange that pic shows Brown with Gold print Johnego. Could be earlier version when transition was made from Silmic 1 to Silmic ll
Hi Carlmart
Sorry no schema, the regs came ready build. Your right its a matter of whether we like the sound. It’s just kind of head scratching for me that you can hear the difference even when regs are only powering the FGA chip.
Karl,
There is no "best" capacitor. Much depends on the circuit they are in and frequency ranges they operate at. There isn't a magic capacitor that you can replace others with an end up with perfect solution.
From what I see come across my bench, most would do to get a proper soldering station along with a lot of practice before descending on some poor piece of equipment and hacking it. Of course, its easy to damage capacitors with too much heat, or by not forming the leads properly when the replacements are installed. Update your skills and the equipment and capacitors might survive.
The number of capacitor replacement hacks out there is mind bending. The prices charged for "recapping" a receiver is way out there and may incorrect decisions are made with respect to the capacitor types used. Anyone can mechanically replace parts, but it takes knowledge to use the correct ones. That knowledge is not "extensive listening tests" with your group.
You guys are going to do this anyway no matter what is said. Please at least develop good soldering and desoldering skills before doing any more. It wouldn't hurt to learn about capacitors and which types go where.
If you are working on amplifiers, you need at a minimum:
1) A distortion analyser (THD meter and low distortion oscillator)
2) an oscilloscope
3) a good meter, Fluke or Keysight (HP / Agilent)
4) a variac, variable AC output transformer
5) a good soldering station and desoldering tools
If you don't have those basic instruments you have zero right to touch an amplifier!
-Chris
There is no "best" capacitor. Much depends on the circuit they are in and frequency ranges they operate at. There isn't a magic capacitor that you can replace others with an end up with perfect solution.
From what I see come across my bench, most would do to get a proper soldering station along with a lot of practice before descending on some poor piece of equipment and hacking it. Of course, its easy to damage capacitors with too much heat, or by not forming the leads properly when the replacements are installed. Update your skills and the equipment and capacitors might survive.
The number of capacitor replacement hacks out there is mind bending. The prices charged for "recapping" a receiver is way out there and may incorrect decisions are made with respect to the capacitor types used. Anyone can mechanically replace parts, but it takes knowledge to use the correct ones. That knowledge is not "extensive listening tests" with your group.
You guys are going to do this anyway no matter what is said. Please at least develop good soldering and desoldering skills before doing any more. It wouldn't hurt to learn about capacitors and which types go where.
If you are working on amplifiers, you need at a minimum:
1) A distortion analyser (THD meter and low distortion oscillator)
2) an oscilloscope
3) a good meter, Fluke or Keysight (HP / Agilent)
4) a variac, variable AC output transformer
5) a good soldering station and desoldering tools
If you don't have those basic instruments you have zero right to touch an amplifier!
-Chris
Hi all,
This is a very interesting thread that I check up on from time to time ... and now notice your latest posts.
And I notice your comments about the Silmic II which I have recently listened to (again) in my search for an optimum Ø10 to Ø12.5 mm diameter capacitor. And IMHO (and my humble ears) the Silmic II is very "clean" without any noticable "ringing" or smearing of details (in short). Also the spatial rendering seems to be quite open and with quite good transparency.
That said, I (also) find that the tonal balance is a bit off - to my ears they are somewhat bright.
However, some time ago I happened to give them a very long burn-in period (3 weeks) and this to my ears significantly improved their "seamlessness", the flow of the music and tonal coherence. The difference (to my ears) was quite clear. It also somehow amended the tonal "imbalance" albeit not entirely I would say.
@sumotan: About the Japanese Elna Silmic IIs I found these available on ebay:
2PCS Japan ELNA 1000UF/25V SILMIC II Series High-END HIFI Audio Capacitor CAP | eBay
Not exactly cheap compared with the ones digikey/mouser sell but maybe these are the colors/types that you have bought yourself?
... Any chance you can say a bit more about the differences you find between these Japanese Silmic IIs and the Mouser/Digikey versions?
In general I find it to be somewhat frustrating - this trying to find the best "higher capacity" capacitor. Having listened to some different electrolytics by now (Rubycon ZL, ZLG, ZLH, MBZ, RX30; Nichicon FG, HZ, KZ, BP, GQ, GU, HD, KX; Panasonic FM, HFZ, HFQ, FC, Pureism; Elna Starget, Cerafines (different versions), Silmic II, Duorex, RSH; AudioNote Kaisei polarized and bipolar, BC Components (don't remember the type)) I still haven't found the capacitor that I would say has the optimum combination of tonal balance, openness incl. spatial rendering and "height", non-smeared transient reproduction - and, importantly, "musicality".
The best ones of these IMHO have been the Elna Starget, Nichicon KZ & KX (large can, 200 VDC), Panasonic HFZ - and to some extent the Kaisei polarized and the Silmic II with the caveats already mentioned.
I've also listened to some polymer types (Sanyo, Rubycon, Nippon Chemicon (I think), Murata) and to my ears most all of them have an up-tilted tonal balance - and in many cases a smeared transient reproduction. Also, the bass region appears light-weight and with less solidity, something I just by thought have attributed to these capacitors' typically very high leakage current (which I assume is not linear wrt the voltage across the capacitor). That said one or two of them have very good mid-range reproduction.
BUT - and this really is the "thing" about this post ... I still haven't found a capacitor that is just subtly sweet and seamlessly tonally rich (beautifully sounding) while at the same time having a natural transient reproduction, spatial rendering, and detail
...
So I am beginning to wonder if such an (electrolytic or the like) capacitor exists at all? Hrmpff!!
I'm sure such a capacitor exists - it's just where to find it ... ... If you think you know of such a capacitor I'd appreciate hearing about it ...
Cheers
Jesper
This is a very interesting thread that I check up on from time to time ... and now notice your latest posts.
And I notice your comments about the Silmic II which I have recently listened to (again) in my search for an optimum Ø10 to Ø12.5 mm diameter capacitor. And IMHO (and my humble ears) the Silmic II is very "clean" without any noticable "ringing" or smearing of details (in short). Also the spatial rendering seems to be quite open and with quite good transparency.
That said, I (also) find that the tonal balance is a bit off - to my ears they are somewhat bright.
However, some time ago I happened to give them a very long burn-in period (3 weeks) and this to my ears significantly improved their "seamlessness", the flow of the music and tonal coherence. The difference (to my ears) was quite clear. It also somehow amended the tonal "imbalance" albeit not entirely I would say.
@sumotan: About the Japanese Elna Silmic IIs I found these available on ebay:
2PCS Japan ELNA 1000UF/25V SILMIC II Series High-END HIFI Audio Capacitor CAP | eBay
Not exactly cheap compared with the ones digikey/mouser sell but maybe these are the colors/types that you have bought yourself?
... Any chance you can say a bit more about the differences you find between these Japanese Silmic IIs and the Mouser/Digikey versions?
In general I find it to be somewhat frustrating - this trying to find the best "higher capacity" capacitor. Having listened to some different electrolytics by now (Rubycon ZL, ZLG, ZLH, MBZ, RX30; Nichicon FG, HZ, KZ, BP, GQ, GU, HD, KX; Panasonic FM, HFZ, HFQ, FC, Pureism; Elna Starget, Cerafines (different versions), Silmic II, Duorex, RSH; AudioNote Kaisei polarized and bipolar, BC Components (don't remember the type)) I still haven't found the capacitor that I would say has the optimum combination of tonal balance, openness incl. spatial rendering and "height", non-smeared transient reproduction - and, importantly, "musicality".
The best ones of these IMHO have been the Elna Starget, Nichicon KZ & KX (large can, 200 VDC), Panasonic HFZ - and to some extent the Kaisei polarized and the Silmic II with the caveats already mentioned.
I've also listened to some polymer types (Sanyo, Rubycon, Nippon Chemicon (I think), Murata) and to my ears most all of them have an up-tilted tonal balance - and in many cases a smeared transient reproduction. Also, the bass region appears light-weight and with less solidity, something I just by thought have attributed to these capacitors' typically very high leakage current (which I assume is not linear wrt the voltage across the capacitor). That said one or two of them have very good mid-range reproduction.
BUT - and this really is the "thing" about this post ... I still haven't found a capacitor that is just subtly sweet and seamlessly tonally rich (beautifully sounding) while at the same time having a natural transient reproduction, spatial rendering, and detail
...So I am beginning to wonder if such an (electrolytic or the like) capacitor exists at all? Hrmpff!!
I'm sure such a capacitor exists - it's just where to find it ...
... If you think you know of such a capacitor I'd appreciate hearing about it ... Cheers
Jesper
Hi Gentlevoice,
Like what Anatech says there’s no best caps, what is needed though is to know weakness & strength of each cap. Never populate a board with all the same type or brands of caps you’ll end up with a not too good sounding project. Some caps are good for ps, some for local decoupling on analog & some for digital. How to find the right combi, Im afraid one just has to experiment through trial & listening. Btw even caps of the same value & brands but different voltage ratings will have difference in sound. Like KZ the 50v version sounds much better then 25v types.
Like what Anatech says there’s no best caps, what is needed though is to know weakness & strength of each cap. Never populate a board with all the same type or brands of caps you’ll end up with a not too good sounding project. Some caps are good for ps, some for local decoupling on analog & some for digital. How to find the right combi, Im afraid one just has to experiment through trial & listening. Btw even caps of the same value & brands but different voltage ratings will have difference in sound. Like KZ the 50v version sounds much better then 25v types.
Last edited by a moderator:
Yes Carlmart but then again not all brands of caps behave in this manner. Some lower voltage sounds the same. Just to open up another can of worms, same brand, same voltage & value but one cap short & stubby, the other slim & tall. Slim tall always wins in performance. This is my personal experience will mucking around with caps.
Btw Carlmart, lower esr does not dictate a win win situation. Me do my experiment via listening test so no scientific explanation, again it still depends on where the cap is being applied.
Btw Carlmart, lower esr does not dictate a win win situation. Me do my experiment via listening test so no scientific explanation, again it still depends on where the cap is being applied.
They are current production. More popular locally than the black-gold version.
Attachments
Last edited by a moderator:
As a follow-up, today I received some Nichicon UES1C221MPM1TD (220µF 16V non-polar) and put one on the same test jig that I used for my earlier tests, except without applying any external DC bias, since this is a non-polar cap, to be used without any DC bias.
I found that they had unmeasurable distortion in my test scenario: drive the cap with a 40Ω source 2kHz sine wave through a 100Ω resistor into the cap, which is then connected to ground. The test set analyzer receives a differential voltage across the 100Ω resistor as well as across the cap under test. The cap under test signal is then put into a distortion analyzer that notches the 2kHz fundamental, and 64 complex averages of a 512K point 192kHz FFT are taken from that post notch residual.
The largest applied test signal to the cap under test was +15dBV, and as measured from the 100Ω series resistor’s Kelvin connections, this resulted in 46.6mA of drive current through the capacitor, quite a huge drive level. Still, I could not detect any distortion component down to the noise floor of my test set, which is around -165dBV for this measurement. So, in terms of distortion, it seems clear to me that these caps are extremely clean.
There are many reasons why people prefer one component to another, but I have to say that this Nichicon UES series part behaves extremely well in terms of distortion, even at extremely high drive currents, far better than almost all other quality polar caps I’ve measured. Personally, I was very pleased to see these good test numbers from this cap, since many people seem to like them, and I know that other caps actually do measure worse under the same test.
I also want to point out that most coupling cap applications do not force the cap to handle >40mA of signal current. Often, the current through a coupling cap, which is proportional to its distortion, is a few orders of magnitude lower, more on the scale of tens of microamps, and not tens of milliamps. So, IMHO, these Nichicon UES series caps seem to be blameless re. distortion, even when they are asked to do extremely heavy lifting, and can pass huge signal currents with essentially unmeasurable distortion.
I found that they had unmeasurable distortion in my test scenario: drive the cap with a 40Ω source 2kHz sine wave through a 100Ω resistor into the cap, which is then connected to ground. The test set analyzer receives a differential voltage across the 100Ω resistor as well as across the cap under test. The cap under test signal is then put into a distortion analyzer that notches the 2kHz fundamental, and 64 complex averages of a 512K point 192kHz FFT are taken from that post notch residual.
The largest applied test signal to the cap under test was +15dBV, and as measured from the 100Ω series resistor’s Kelvin connections, this resulted in 46.6mA of drive current through the capacitor, quite a huge drive level. Still, I could not detect any distortion component down to the noise floor of my test set, which is around -165dBV for this measurement. So, in terms of distortion, it seems clear to me that these caps are extremely clean.
There are many reasons why people prefer one component to another, but I have to say that this Nichicon UES series part behaves extremely well in terms of distortion, even at extremely high drive currents, far better than almost all other quality polar caps I’ve measured. Personally, I was very pleased to see these good test numbers from this cap, since many people seem to like them, and I know that other caps actually do measure worse under the same test.
I also want to point out that most coupling cap applications do not force the cap to handle >40mA of signal current. Often, the current through a coupling cap, which is proportional to its distortion, is a few orders of magnitude lower, more on the scale of tens of microamps, and not tens of milliamps. So, IMHO, these Nichicon UES series caps seem to be blameless re. distortion, even when they are asked to do extremely heavy lifting, and can pass huge signal currents with essentially unmeasurable distortion.
Capacitor is only a small part of the system. It's only when you have improved the critical active parts you should look into boutique caps. I have seen many hobbyists trying to improve poor circuit with boutique parts. Best cap is then only best for that system. You change your speaker and it may not be the best cap anymore
From my experience 25 years ago, I found that Black Gate is the best for PS. I still own NOS of them but I don't care so much about cap because now they are too expensive. This thread is started by a master who thinks that Elna Silk is cheap and good. So, why can't it be good enough for you too?So I am beginning to wonder if such an (electrolytic or the like) capacitor exists at all? Hrmpff!! I'm sure such a capacitor exists - it's just where to find it ...
I'll be happy using Silmic for the main supply. Any sound deficiencies will be addressed somewhere else. Supply bypass, grounding and wiring, regulators and even diodes can become other objects of tweaking.
Build the Jung Super Regulator with low noise Dale resistors and top quality op-amp (AD797, OPA627, LT1028) as well as the best caps like Black Gate (Std, F, N, NX), and I will be happy not to use it. Best cap will not automatically give best sound.
And who knows, it might be the cheap power supply diodes that gives you the problems you expect the best cap can solve?
I thought you are abroad as you're not familiar with the brown Silmic II
Last edited by a moderator:
I have Silmic II but i don't pay attention to how they sound (don't have the time). Long time ago I considered them (all the Silmic) as average (I preferred red Cerafine). Pureism was avoided after comparing with the red Cerafine.
I also like to try out stuffs, find out how things affect sound, but more with active parts.
Last edited by a moderator:
Hi all,
I have just written - and erased due to happening to click an advertisement that did not open in another window - one hour of reply time on your questions and comments - so now it will have to be somewhat shorter.
@carlmart:
@sumotan: Thanks for your feedback.
@Monte McQuire:
@johnego:
Cheers,
Jesper
I have just written - and erased due to happening to click an advertisement that did not open in another window - one hour of reply time on your questions and comments - so now it will have to be somewhat shorter.
@carlmart:
Appr. 2 Vpp music into 8 ohms into the capacitor connected to ground. Bias 4 VDC.
Actually the very UES types that Monte McQuire has posted distortion measurements for this morning. Also some very large size Panasonics (BP type) that sounded quite good but were too big.
No, I didn't. My amp does not have GNFB.
@sumotan: Thanks for your feedback.
I agree. Would be great though if there was just one cap "type" that had most all the qualities desired and only needed just a little correction. I reckon it would make things easier ... And similar for caps for digital circuit use ...
Good to know ... thanks for the tip
That is also what I have heard.
@Monte McQuire:
Hi Monte ... If possible can I ask you to just briefly clarify what your measurement setup looks like? I've attached an LTSpice screendump of what I reckon it could be but I am a little unsure of how you connect things around the resistor and capacitor ... BTW - have you ever tried to measure the caps with higher frequencies e.g. 50 kHz? Could be interesting to see how they respond to higher frequencies ...
@johnego:
In my experience everything interacts, yet somehow the capacitors' sonic imprint seems to get into the overall sound nonetheless. But, yes, circuit topology, resistors, transistors, layout, IC choice etc. all matters. The fair complex electronic world .
Cheers,
Jesper
Attachments
Last edited by a moderator:
Most capacitors and power supplies are designed mainly for low ripple. That's the first mistake.Today we live in environment full of interference. An amplifier has a capability called PSRR and CMRR. It is possible to have an amp with good PSRR/CMRR but unfortunately it will still very far from perfect (they collapses at HF). Imo, this issue should be addressed at both amplifier level and power supply level.
I'm not so sure about we are having different taste. "Too much highs". What does that mean? Tweeter has it's frequency response. Amplifier has its HF response. Why put the blame on the cap? If you like non-FB amp, may be your amp has high level of 2nd and 3rd order distortion such that the sound is too forward sounding and the cap just don't want to interfere?
Yes, the capacitor matters and it is a complex system. That's why it is almost useless to generalize the sound of caps. 'We' don't know when we need a high ESR and when we need it low.You prefer the sound of 1000uF cap A over 1000uF of cap B. Then you add 330nF MKP to cap B and suddenly cap A cannot compete even with the addition of similar 330nF MKP! So which one is the best cap you think? It could be A, you just need 100n instead of 330n.In my experience everything interacts, yet somehow the capacitors' sonic imprint seems to get into the overall sound nonetheless. But, yes, circuit topology, resistors, transistors, layout, IC choice etc. all matters. The fair complex electronic world ...
Last edited by a moderator: