I'm using the caps as bypass caps in which almost all of the circuit's output current is sourced through a pair of these caps, one for each power supply rail polarity. In doing so, there is ripple on the caps, and also potentially distortion from a nonlinear cap passing signal current from the rail to ground. So, the test was to bias the caps to 22V, which is the DC bias they would operate with, and then dump a lot of signal current into them and then examine the distortion that gets generated across the cap.
For my circuit, the nice thing is that capacitor distortion levels of -150dBV and the like, combined with even modest PSRR from a post-regulator and the subsequent amplifier, result in ridiculously low distortion in the actual circuit - none of the caps is anything close to problematic.
However, it became interesting to me that these caps do actually perform differently in an objective way, and that some may be cleaner than others. What was also surprising was that the audiophile caps actually measured better, so I thought it would be fun to see if the numbers follow any subjective rankings.
As for break-in, I did have to raise the DC bias slowly, just so I did not annoy the analyzer with a large voltage pulse on its input. However, I can't honestly tell you if things improve or not. Powering the cap up to 22V took a minute or so, and then I let it sit for a few more minutes, and then the tests themselves, which were FFTs with a lot of averages, probably took another 15 minutes to sweep over four signal drive levels. If aluminum electrolytics do improve, which I think will only happen subtly with freshly manufactured parts, then my measurements will be pessimistic, and that's OK for me since they all perform extraordinarily well for my circuit. Maybe they get better... maybe not?
I do know that a solid Tantalum will slowly continue to form over time, and for those applications, like the poster who was using a Ta coupling cap for a 0.1-10Hz low noise amplifier, the noise that a slowly forming Ta cap exhibits is irregular shot noise that hopefully decays to zero, but very slowly. As that poster said, some caps will not decay below some leakage (shot noise) level, so for that application, it is useful to spend a lot of time screening and forming one ideal coupling cap. But, for aluminum electrolytics, the forming mechanism isn't a shot noise phenomenon at all, and I feel that you can stay away from significant nonlinearities so as not to worry about forming at all.
I did, only on my caps those were aluminium. This exercise eventually provided me with quite a bit of teflon so i could make my own pF caps for a EAR834 phono clone
Hi,
What is the behavior in frequencies between 20 hz to 20k hz of a decoupling cap ?
at iso voltage and capacitance measured with a good LCR bridge, is there two lytics, diferent model and-or brabd, with the same acurate self resonant frequency, esr, esl, leads thickness & both coper cored with tin, same lytic formula, case size ?
I have no doubt in a sensible circuit two diferent caps sound no the same while not knowing why !
What I can notice, without care too much :
- there is a consistency within a brand model about sound signature
- esr is far to be the only factor that matter while staying important one.
- there is an inconsitency at iso voltage & capacitance & iso esr between two caps models and-or brand
- I can ear it out of the box without break in & I purchase my caps new at Mouser-Farnell reliable distributors.
WHen I set-up a dac, amp, pre, playing with caps & resistors, I'm aware it's not changing 95% of the sound, but the 5% left can sometimes makes a significant diference enough in the whole electrical network of the hifi. I see it as an EQ to my personal tastes and hifi/room.
I always work when I have time to spent on caps game, with low voltages (no tubes gears) without soldering between two caps, and giving me several days between two caps with a same panel of musical material since 20 years now.
And indeed, I can notice a difference. I don't say this is mandatory, but as it's about trade offs and most of time gears are not perfect individually and even less in the hifi network, it just helps to EQ and cunterbalance the gears between each others to acheive the sound I like. And I often check with musician friends without saying I changed something, sometimes they notice it on a same material without I asked !
For illustration I have worked 3 years to improve a dac and setup it. all is hearable in my system ( a guy said me my hifi while being electrodynamic sounded near a panel acuracy ?! ) and I can testimony it's even possible to hear the type of solder : I had to go back from 4% Ag to normal in the output analog circuit because I could ear it ! A bad thing in the speaker filter certainly, but I could ear it !
So yes the caps can make a noticeable diference, now I find the method of measuring their distorsion is just 1% of the problem ! It says too little if nothing most of time with lytics and it's outside of context as measured raw ! Take an Ongaku amp and a little smps one, same V and W outputt in an ABX test, there is few chance the two amps can be sorted regarding their distorsion outputt patern ! OK, it's not about caps here, just to say distorsion and esr are most of time not enough to explain something. It's a beginning of a methodology not its end. Do you see often in the review an acurate measurement with two same caps measured at 1000 hz windows till 20 hz to 20 k hz for instance ? Never... you can see often one or two frequencies, in lytocs often 1 k hz and 10 k hz...
Sorry a bit long, not intended to open and re open the eternal cap can worm. But yes it's hearable, no it's not 95% of the whole result but just a little tool in the tool box if having time. According to me Silmic I & II and the oldest ones have all a different sounding effect at iso V & F specs. I don't like Silmic II but if I want to acheive & copensate in a particular situation. If N Pass talked about the Tonorek caps (after all he never talked about the silmic II in particular but just of the silked caps of the brand), they can give sometimes on SS this smooth signature of bad designed tubes amps while being indeed free of hearable distorsions. But a Nichicon but some specific ones (ES, KZ) acheive also something smooth yu may or may not like according the gear you tweak.
My point is people, magazine, some designers focus a lot on caps % distorsion because they can measure it, so it's an objective data for communication while most of time explaining nothing in real life context : I wonder if someone for instance has ever listen to a preamp without its inteconects, the amp & the speakers...oh the source as well ! I bet on the fact when I listen to music I eventually don't watch a scope screen
So why bother, just do what you want as far you're happy with it. The today paradigm is about distorsion and esr often when talking about caps and sound : it's not false and ok in the context of the paradigm, not false...just not enough for some, I don't bother anymore about the church I'm into here And often your own receips doesn't work in the hifi network of a friend (though some tweak may work 100% positive and very great enough) hifi network ! Sowhy discuss it yet
A bit long, just my point, YMMV as usual.
What is the behavior in frequencies between 20 hz to 20k hz of a decoupling cap ?
at iso voltage and capacitance measured with a good LCR bridge, is there two lytics, diferent model and-or brabd, with the same acurate self resonant frequency, esr, esl, leads thickness & both coper cored with tin, same lytic formula, case size ?
I have no doubt in a sensible circuit two diferent caps sound no the same while not knowing why !
What I can notice, without care too much :
- there is a consistency within a brand model about sound signature
- esr is far to be the only factor that matter while staying important one.
- there is an inconsitency at iso voltage & capacitance & iso esr between two caps models and-or brand
- I can ear it out of the box without break in & I purchase my caps new at Mouser-Farnell reliable distributors.
WHen I set-up a dac, amp, pre, playing with caps & resistors, I'm aware it's not changing 95% of the sound, but the 5% left can sometimes makes a significant diference enough in the whole electrical network of the hifi. I see it as an EQ to my personal tastes and hifi/room.
I always work when I have time to spent on caps game, with low voltages (no tubes gears) without soldering between two caps, and giving me several days between two caps with a same panel of musical material since 20 years now.
And indeed, I can notice a difference. I don't say this is mandatory, but as it's about trade offs and most of time gears are not perfect individually and even less in the hifi network, it just helps to EQ and cunterbalance the gears between each others to acheive the sound I like. And I often check with musician friends without saying I changed something, sometimes they notice it on a same material without I asked !
For illustration I have worked 3 years to improve a dac and setup it. all is hearable in my system ( a guy said me my hifi while being electrodynamic sounded near a panel acuracy ?!
) and I can testimony it's even possible to hear the type of solder : I had to go back from 4% Ag to normal in the output analog circuit because I could ear it ! A bad thing in the speaker filter certainly, but I could ear it ! So yes the caps can make a noticeable diference, now I find the method of measuring their distorsion is just 1% of the problem ! It says too little if nothing most of time with lytics and it's outside of context as measured raw ! Take an Ongaku amp and a little smps one, same V and W outputt in an ABX test, there is few chance the two amps can be sorted regarding their distorsion outputt patern ! OK, it's not about caps here, just to say distorsion and esr are most of time not enough to explain something. It's a beginning of a methodology not its end. Do you see often in the review an acurate measurement with two same caps measured at 1000 hz windows till 20 hz to 20 k hz for instance ? Never... you can see often one or two frequencies, in lytocs often 1 k hz and 10 k hz...
Sorry a bit long, not intended to open and re open the eternal cap can worm. But yes it's hearable, no it's not 95% of the whole result but just a little tool in the tool box if having time. According to me Silmic I & II and the oldest ones have all a different sounding effect at iso V & F specs. I don't like Silmic II but if I want to acheive & copensate in a particular situation. If N Pass talked about the Tonorek caps (after all he never talked about the silmic II in particular but just of the silked caps of the brand), they can give sometimes on SS this smooth signature of bad designed tubes amps while being indeed free of hearable distorsions. But a Nichicon but some specific ones (ES, KZ) acheive also something smooth yu may or may not like according the gear you tweak.
My point is people, magazine, some designers focus a lot on caps % distorsion because they can measure it, so it's an objective data for communication while most of time explaining nothing in real life context : I wonder if someone for instance has ever listen to a preamp without its inteconects, the amp & the speakers...oh the source as well
! I bet on the fact when I listen to music I eventually don't watch a scope screen So why bother, just do what you want as far you're happy with it. The today paradigm is about distorsion and esr often when talking about caps and sound : it's not false and ok in the context of the paradigm, not false...just not enough for some, I don't bother anymore about the church I'm into here
And often your own receips doesn't work in the hifi network of a friend (though some tweak may work 100% positive and very great enough) hifi network ! Sowhy discuss it yet A bit long, just my point, YMMV as usual.
How many people have built some version of Jung/Didden superegulator and what capacitors brand and types did you use for the 120uF units?
The ones Walt mentioned on his last regulator version were HQ and FC, which are not made anymore. But I'm sure others used other brands and types.
Did you notice any difference in the audio quality? Note I said "difference", not good or bad.
The ones Walt mentioned on his last regulator version were HQ and FC, which are not made anymore. But I'm sure others used other brands and types.
Did you notice any difference in the audio quality? Note I said "difference", not good or bad.
This thread contains claims that some particular electrolytic capacitors by particular manufactures are superior compared with the rest. However, can anyone explain what makes these capacitors 'superior' with reference to an electronic equivalent circuit for such capacitors? It is a well known fact electrolytics may not respond fast enough to very high dI/dt, but this can be corrected by connecting a faster but much smaller capacitor in parallel with the main capacitor.
The crucial question is: What makes these capacitors better electronically?
The crucial question is: What makes these capacitors better electronically?
If you download the datasheet, there's this warning: This Series is no longer available for purchase in the country of Japan.
So what Panasonic model would it be replaceable with? Or any other brand?
Mine has to be 50v or 63v.
I use "Premium grade “nichicon MUSE” acoustic series" because I trust the Nichicon brand. This is for the power supply bypass caps. I cannot answer the question, "What makes these capacitors better electronically?"
https://www.mouser.com/datasheet/2/293/e-ukz-1513448.pdf
It may be a little more expensive than the Panasonic or others. Being electrolytic, this "premium grade, audio series" Nichicon is still very reasonably priced.
For the input signal path cap, I follow John Curl's recommendation. The Elna SILMIC series Silk fiber using audio purpose capacitor.
https://www.mouser.com/datasheet/2/129/rfs_e-606.pdf
Last edited:
Here is Panasonic's site for you:
Aluminum Electrolytic Capacitors - Industrial Devices & Solutions - Panasonic
Aluminum Electrolytic Capacitors - Industrial Devices & Solutions - Panasonic
I am sorry about the mix-up. It was Nelson Pass who started this thread, not John Curl. I typed without checking first.
I can`t understand why people haven`t considered using solid polymer electrolytics for upgrading.They have limitations with regard to availability of certain capacitance and voltage values but they have better specs than standard "wet" electrolytics.
They typicaly have lower ESR,lower noise,higher ripple current rating and longer life and don`t cost that much more than standard electrolytics.
They typicaly have lower ESR,lower noise,higher ripple current rating and longer life and don`t cost that much more than standard electrolytics.
Please, excuse my ignorance. Which manufacturer makes this solid polymer electrolytics? I need something at 35V or above for power supply application.
They are OS-CON caps. Sanyo is my favorite manufacturer. Nichicon also makes them.
And Panasonic too (if these are what you mean)
https://www.mouser.ca/datasheet/2/315/panasonic_07172019_Capacitor_OSCON_(SXE_Series)_04-1622870.pdf
From Mouser.
https://www.mouser.ca/datasheet/2/315/panasonic_07172019_Capacitor_OSCON_(SXE_Series)_04-1622870.pdf
From Mouser.
Thank you. I found them at Mouser.
Panasonic Radial Aluminum Polymer Capacitors Electrolytic Capacitors Aluminum Organic Polymer Capacitors | Mouser
Sanyo Oscons are no longer made,Panasonic make solid polymer types under that name but strangely the original Oscons used a conductive salt electrolyte which is different technology.
Keep in mind that polymers don't have any self healing capabilities and usually got worse ripple ratings then electrolytics. I won't use them as power supply caps. They won't last long. Also the best ploymers got worse leakage specs then an average chinese electrolytic. The best cap is one that fits physically and electrically in your circuit.