In audio.
Although ESR does set a limit how far "BIG CAPS!" design can go towards low ripple. (There's always a better way.)
In rapid pulse-work (not bombs, but large switching supplies) a trace of lost energy a million times a second adds-up to heat and failure.
These are the ones I'm pointing finger at; they are the ancestors of the Vishay BCComponents. I even think they're worse than what you call no-name.
More generally: I said it before, and I say it again: coupling caps are signal "pass" devices, and decoupling caps are signal "stop" devices. Both need TLC, equally.
Last edited:
Just recapped my 1998 solid state amp. Not because I wanted to but the fuse blew on one channel on inrush with only caps or FETs being the culprit. FETs tested fine. The amp originally used Jamicon for the 'lytics and any orange/brown drop has is marked with a circle "L" but can't find who manufactured those. The amp is an oven so the 85degC pieces, 24 years on were starting to have capacitance problems leading to higher ripple and inrush - hence the blown fuse. I suspect this is also the reason for the mute circuit diode getting hot (everything in that circuit tests ok out of circuit except the caps being weak).
The amp now contains mostly 105degC pieces:
The amp used a 0.22uF Jamicon electrolytic for signal coupling duties. Now I normally use WIMA FKP1 630Vdc-400Vac caps and if they're between a driver and parallel power tube section I'll use FKP1 0.22uF+MKP 1uF. However I decided to take some hints from designers here and replace the Jamicon with a Panasonic ECW MPP 0.22uF 400V. Currently needs a burn in just as the FKP1s did.
The amp now contains mostly 105degC pieces:
- CDE 380LX 15000uF and 38LX 3300uF 50V pieces - basically for high current (6A) and 20mOhm ESR vs the original 5A or so rating.
- Panasonic EUU FM 470uF for one of the balancing compensation caps and for the decoupling cap for the MC33079P opamp. They're touted as low ESR but anything is better than cooked caps.
- Wurst and Rubicon - mainly because at £0.20/piece they are known brand but cheap enough for the mute enable circle where they're off the direct signal path but not entirely decoupled.
The amp used a 0.22uF Jamicon electrolytic for signal coupling duties. Now I normally use WIMA FKP1 630Vdc-400Vac caps and if they're between a driver and parallel power tube section I'll use FKP1 0.22uF+MKP 1uF. However I decided to take some hints from designers here and replace the Jamicon with a Panasonic ECW MPP 0.22uF 400V. Currently needs a burn in just as the FKP1s did.
I'm happy with the Panasonic ECW, but I wonder where they're made; they look a lot like the Chinese CBB, and priced about the same too.
Most underrated capacitor probably.
Panasonic FM and FC sound horrible. I use to use them and learned it was all a mistake. Very low ESR doesn't translate to good music necessarily.
I found this article interesting
https://www.avnet.com/wps/portal/ab...et-aluminium-vs-polymer-aluminium-capacitors/
The factory electrolytic parts on an old DAC were leaking all over the place on the PS and so I decided to swap out everything to keep it going.
I initially used the same values for everything and also BiPolar around the DAC chip as per the original manufacturer: I used Nichicon 'muse' ES. They looked great in bright green, and the rest of the DAC PS I used Silmic II because they are a reasonable price and I rate their sonic signature from past experience.
Anyway with great expectation bias and high hopes I turned it on and listened. I had the luxury to compare to an identical original DAC I borrowed from my brother as he is not using it atm, and I can then update that too. It has similarly leaked everywhere as well.
After some hours of running in I have to be honest, I preferred the original leaking tired untouched DAC. Yeah the updated DAC had more detail, slightly better bass power and extension and it was more dynamic BUT some how it sounded a bit more HIFI than music, a bit unnatural and false. I waited and listened/swapped over the next 2-3 weeks and then concluded I actually DID prefer the original worn out old DAC. Oh dear, familiarity bias?
So then I changed 4 caps, the +5v Digital and +5v Analogue on the Left and Right DAC from the Muse ES to Oscon Sep C and increased the capacitance. Wow what a difference, I now prefer the updated DAC by quite a margin.
Anyway my experience is expectation bias, familiarity bias, investment bias does not really justify the perceived changes. Measurements may explain the changes of course, but I still think the final arbitrator should be your ears 🙂
https://www.avnet.com/wps/portal/ab...et-aluminium-vs-polymer-aluminium-capacitors/
The factory electrolytic parts on an old DAC were leaking all over the place on the PS and so I decided to swap out everything to keep it going.
I initially used the same values for everything and also BiPolar around the DAC chip as per the original manufacturer: I used Nichicon 'muse' ES. They looked great in bright green, and the rest of the DAC PS I used Silmic II because they are a reasonable price and I rate their sonic signature from past experience.
Anyway with great expectation bias and high hopes I turned it on and listened. I had the luxury to compare to an identical original DAC I borrowed from my brother as he is not using it atm, and I can then update that too. It has similarly leaked everywhere as well.
After some hours of running in I have to be honest, I preferred the original leaking tired untouched DAC. Yeah the updated DAC had more detail, slightly better bass power and extension and it was more dynamic BUT some how it sounded a bit more HIFI than music, a bit unnatural and false. I waited and listened/swapped over the next 2-3 weeks and then concluded I actually DID prefer the original worn out old DAC. Oh dear, familiarity bias?
So then I changed 4 caps, the +5v Digital and +5v Analogue on the Left and Right DAC from the Muse ES to Oscon Sep C and increased the capacitance. Wow what a difference, I now prefer the updated DAC by quite a margin.
Anyway my experience is expectation bias, familiarity bias, investment bias does not really justify the perceived changes. Measurements may explain the changes of course, but I still think the final arbitrator should be your ears 🙂
I will give them a try, let them burn in and when I have time I'll hook up the scope and check the signal on both sides. I typically don't like 0.22uF it depends on what the load on the other side is and what is driving it. I get what you mean - also there is the point that sagging rails (I use that term lightly) means some distortion and if that's 2nd harmonic and that's what you're after then it makesense.
I didn't like the WIMA MKP initially either but I did find it took longer than the FKP to get itself sorted. Next cap I want to try is the FKP4. Slower than the ridiculously fast FKP1 but still way faster than you need for audio.. If it has the same clarity then I'll probably switch.
I have a 55ohm HPA that uses a bank of MUSE ES 50V with FKP in parallel for each channel. I'm impressed by the MUSE ES and the FKP just adds a little more clarity whilst the MUSE seems to work nicely with headphones for pushing low mid frequencies. When I say bank of MUSE, there's four muse + one FKP per phase, with each channel having the two differential phases so each channel has eight 330uF 50V muse (2640.44uF per channel) 😀 Enough for my headphones..
The Panasonic is something I've been meaning to try for a while. However the amp itself isn't great, it does need some work sorting out the top end.
Last edited:
I find the stacked black Ecw radial better than the redish ecw shown above.
So NickNuck, you use the bipolar ES as power bank or load decoupling as well ?
The Panasonic FC are excellent but they should not be used alone in a circuitry. FMs are less good to me because their mid highs that often are fatiguing when used at power bank or smoothing...but rare cases.
Esr here is not a good information as FC and FM have very simikar esr but sounds tottaly different.
There are not better capacitor as some like to repeat it all the time, you need to mix different brands to voice well your circuitry... Designer stop the job at the circuitry design but not that difficult itterating, fatiguing because long job at mixing passive parts try to enhance their design.,,that is apity because there is a lot to do here.
I find a poor argument to choose a capacitor for its standalone paper datasheet and said quqlity in a repair job...this is not the only job for an amp or a dac to last 30 years...It is to sound the best and capacitor are a parg of it..,, I mix all the brands, sometimes same model but at a different voltage or capacitance to adjust the best sounding result in my whole hifi chain just because I can and have not the same goal than a tech that repair an amp or a brand that sell a standalone device.
I can use a Samwha with a black gate and a Nichicon Z with a Pan Fm and elswhere a Pan Fr in the sqme pcb with a decoupling near the load with qcrylic capacitors and all the devices at home sounds very well said all the hifi people that visit me.
So clearly this ranking like the best is silmic 2 or such is stupid...however they have all a typical sound and voicing in a circuitry at iso use: smoothing, coupling, decoupling, power tank, etc... Oh and I use the violett Elna and wometimes Tonorsx silk as smoothing cap in digital power supply for clock... Go figure why.
So NickNuck, you use the bipolar ES as power bank or load decoupling as well ?
The Panasonic FC are excellent but they should not be used alone in a circuitry. FMs are less good to me because their mid highs that often are fatiguing when used at power bank or smoothing...but rare cases.
Esr here is not a good information as FC and FM have very simikar esr but sounds tottaly different.
There are not better capacitor as some like to repeat it all the time, you need to mix different brands to voice well your circuitry... Designer stop the job at the circuitry design but not that difficult itterating, fatiguing because long job at mixing passive parts try to enhance their design.,,that is apity because there is a lot to do here.
I find a poor argument to choose a capacitor for its standalone paper datasheet and said quqlity in a repair job...this is not the only job for an amp or a dac to last 30 years...It is to sound the best and capacitor are a parg of it..,, I mix all the brands, sometimes same model but at a different voltage or capacitance to adjust the best sounding result in my whole hifi chain just because I can and have not the same goal than a tech that repair an amp or a brand that sell a standalone device.
I can use a Samwha with a black gate and a Nichicon Z with a Pan Fm and elswhere a Pan Fr in the sqme pcb with a decoupling near the load with qcrylic capacitors and all the devices at home sounds very well said all the hifi people that visit me.
So clearly this ranking like the best is silmic 2 or such is stupid...however they have all a typical sound and voicing in a circuitry at iso use: smoothing, coupling, decoupling, power tank, etc... Oh and I use the violett Elna and wometimes Tonorsx silk as smoothing cap in digital power supply for clock... Go figure why.
Pfffff,
Just try a 10 000 uF /50V 105°Jamicon with a Panasonic FC 1200/50V low heigth factor as power bank of an amp and also smoothing if the behavior is shared because no Pi... Then report please. I invit Papa aka Nelson to try it and report as well the subjective sounding tests... Use simple Nich KZ in 100V rating low capacitance elswhere... for simplicity of the test. Destroyer OS you know how design amps then do some other caps tests please.
Anatech, Let us talk with concrete experiments in spite of all this critisisms. I do not care of Mc Intosh...most sounded not fine. Only few sounded excelent. It is not about fixing amps only for refurbishing them for solidity butvalso try to henhance them a littme and adapt in a particular hifi chain at home and all are different so you can voice with caps a little and sometimeq a lot...most often it will sounds better elswhere too... good design is mandatory, ok but good parts choice is a part of the often not tottaly finished devlopment, imho.
Last edited:
I service Leslie amps that used Mallory FP cans, and there was a period from about 1969-1973 when Mallory appears to have made capacitors that greatly exceed expectations for long-life for reasons I'm sure we'll never know. They may have used chemicals that are now banned. But I have a 1971 Leslie amp of my own that has its original Mallory FP can that still tests perfect on all tests, ESR, impedance bridge, and full-voltage leakage. If you go to ~1968 and older, they will usually have at least one section that is going bad.
What I'm saying is that the early 70s Mallory FP cans are outliers. Mallory was doing something right that even they couldn't maintain. And that's something to keep in mind. Not all product lines are consistent over time.
I know FR has lower esr...I use it in dac power supply digital side...often good here. But not always. ESR is a poor indicator unless you have enough to avoid ringing in some situations imho.
And if you ask me I know also the FS and FK and KL...I have tons of caps.. always bougth from official distributor but also well known shop...here it is not enough . For instance I have seen yedterday while I was searching a 1 uF 50V lytics in my wild collection two Silmic 2...same casing but difetent diameter...both 1 uF 50V...but not purchased the same place and maybe the same year...is it for one a fake or just different production year and sligth change. One Mouser the other for a known german shop... idem with a mox resistor at two most known eshop in France: mundorf model but grey not the same and size diameter not the same ? Fake or Mundorf changed their suppliers ? Go figure again...
And if you ask me I know also the FS and FK and KL...I have tons of caps.. always bougth from official distributor but also well known shop...here it is not enough . For instance I have seen yedterday while I was searching a 1 uF 50V lytics in my wild collection two Silmic 2...same casing but difetent diameter...both 1 uF 50V...but not purchased the same place and maybe the same year...is it for one a fake or just different production year and sligth change. One Mouser the other for a known german shop... idem with a mox resistor at two most known eshop in France: mundorf model but grey not the same and size diameter not the same ? Fake or Mundorf changed their suppliers ? Go figure again...
Last edited:
So I have made a proposition with a Jamicon // panasonic FC with precise ratings and circuit position...I invit people and even brand designer to report their finding if they hear a difference for instance with another caps couple same ratings but different models...for instance an Elna Tonorex and a Nichicon... Real life testing no blah...me I know the result but fed up with the formulas giver that never try. 🙂
Let us experiment.
Let us experiment.
I've had the opportunity to compare a Jamicon and a Nichicon Muse at the output of my Musical Fidelity A3.24 DAC. The original Jamicons failed and leaked DC to the output jacks, and I replaced them with Nichicon Muse ES, 13 years ago. The Nichicons have not failed even though they've now been in the unit longer than the originals.
I have genuine Jamicon one rated 105° ...worse area: smoothing of a big amp...15 years...no problem.
So youvused bipolar muse ES for decoupling and eventually find it better with their 85° limit better than a 105° degree cap ? No coment...
I said the blue one rated 105 degrees. Or perhaps yoy talk of DC blocking?
So youvused bipolar muse ES for decoupling and eventually find it better with their 85° limit better than a 105° degree cap ? No coment...
I said the blue one rated 105 degrees. Or perhaps yoy talk of DC blocking?
Hi PRR,
All,
Switching power supplies require different capacitors that are designed for the job. Imagine that, using a part designed for the job you are asking of it!
Now, coupling capacitors shouldn't have signal voltages across them. That matters, think about it. Bypass capacitors shouldn't have AC signal component across them either, but the effective circuit impedance we are talking about is vastly different in the two cases as well.
People constantly talk about your ears being the final arbiter. Too bad your ears lie to you constantly. The correct measurements point the way and keep you honest. I listen as well just in case there is something else going on. I'll say one thing, the measurements I'm looking at seem to agree with what honest observers are saying about the sound quality.
I am pretty convinced that those who argue against measuring equipment do not have access to the required test gear. All they have is nothing. Their ears, so obviously nothing else matters. Okay, whatever ... One point to consider is that there is zero difference between an audio signal and any other signal. It could be from a video signal, process control, or any other sensor. Often times the fidelity of whatever signal is a lot more critical than what you hear. Targeting for a missile, use your imagination. If this stuff isn't bang on, people can die. Stuff can explode or we don't see something deep in space. So we have engineers highly trained in dealing with various signals. There is nothing special about your stereo except it is a lot less critical. Then, there are your speakers, horribly non-linear with other impairments.
So signal amplification is a science. Very accurate test equipment exists for qualifying various signals, so it is used whenever it really matters. Not one single manufacturer of scientific or industrial signal amplifiers has people listen to the product. Why do you think that is? What we have to use in audio, the best stuff, has industrial applications or it wouldn't exist. There are "best practices" engineers use to get it right, but most audio manufacturers don't follow them, the really high end "design by ear" crowd hasn't got a clue. I use the best I can afford, and it never steers me wrong. Not these days.
So, use your noodle, your common sense, and think on this. The high end guys will hate me for it because they can't hide from the cold light of truth that can be quantified and backed up. There are designers and teams of engineer / audio designers who get it right. More that don't. This industry only survives on a good story these days. In the early 1980's, equipment was sold on performance and build quality. That petered out quickly as those who couldn't design things well opted for features and blah, blah, blah. That worked, so now you have the industry that runs on this and contract manufacturing. Physics has not changed, but somehow the new feature is a must have and your older stereo is garbage. Go figure. Profits are higher and quality is pretty low comparatively speaking.
Components sold to the high end audio community is just low hanging fruit. Why not? A great marketing opportunity which has not been lost on a number of smart people.
I agree 100%. And yes, there are far better ways than bulk capacitance beyond a reasonable level. Low frequency ripple is much easier to deal with than the HF content from huge caps also.
All,
Switching power supplies require different capacitors that are designed for the job. Imagine that, using a part designed for the job you are asking of it!
Now, coupling capacitors shouldn't have signal voltages across them. That matters, think about it. Bypass capacitors shouldn't have AC signal component across them either, but the effective circuit impedance we are talking about is vastly different in the two cases as well.
People constantly talk about your ears being the final arbiter. Too bad your ears lie to you constantly. The correct measurements point the way and keep you honest. I listen as well just in case there is something else going on. I'll say one thing, the measurements I'm looking at seem to agree with what honest observers are saying about the sound quality.
I am pretty convinced that those who argue against measuring equipment do not have access to the required test gear. All they have is nothing. Their ears, so obviously nothing else matters. Okay, whatever ... One point to consider is that there is zero difference between an audio signal and any other signal. It could be from a video signal, process control, or any other sensor. Often times the fidelity of whatever signal is a lot more critical than what you hear. Targeting for a missile, use your imagination. If this stuff isn't bang on, people can die. Stuff can explode or we don't see something deep in space. So we have engineers highly trained in dealing with various signals. There is nothing special about your stereo except it is a lot less critical. Then, there are your speakers, horribly non-linear with other impairments.
So signal amplification is a science. Very accurate test equipment exists for qualifying various signals, so it is used whenever it really matters. Not one single manufacturer of scientific or industrial signal amplifiers has people listen to the product. Why do you think that is? What we have to use in audio, the best stuff, has industrial applications or it wouldn't exist. There are "best practices" engineers use to get it right, but most audio manufacturers don't follow them, the really high end "design by ear" crowd hasn't got a clue. I use the best I can afford, and it never steers me wrong. Not these days.
So, use your noodle, your common sense, and think on this. The high end guys will hate me for it because they can't hide from the cold light of truth that can be quantified and backed up. There are designers and teams of engineer / audio designers who get it right. More that don't. This industry only survives on a good story these days. In the early 1980's, equipment was sold on performance and build quality. That petered out quickly as those who couldn't design things well opted for features and blah, blah, blah. That worked, so now you have the industry that runs on this and contract manufacturing. Physics has not changed, but somehow the new feature is a must have and your older stereo is garbage. Go figure. Profits are higher and quality is pretty low comparatively speaking.
Components sold to the high end audio community is just low hanging fruit. Why not? A great marketing opportunity which has not been lost on a number of smart people.
I have no doubt all the gears sold are ade by engineers, but some sound bad, so it is not sufisant.
Well, not all engineers are engineers as they claim. Every product is produced to marketing department requirements, then costs reduced by procurement people. Finally, board layout and physical construction is done by others. So what an engineer designs may look completely different than the original design. So the original design may not have been intended to sound very good to begin with. Sad but true.
However, measurements quickly sort out things that don't sound good from things that do. So you can't throw out the idea of measurements not being sufficient simply because some equipment is not presented with full measurements and you don't like the sound of it. I guess any reason to believe there isn't a science behind this to preserve the idea you can somehow hear things that can't be measured is important to some folks. But if you had the equipment and knowledge to use it, you would definitely be singing a different tune.
However, measurements quickly sort out things that don't sound good from things that do. So you can't throw out the idea of measurements not being sufficient simply because some equipment is not presented with full measurements and you don't like the sound of it. I guess any reason to believe there isn't a science behind this to preserve the idea you can somehow hear things that can't be measured is important to some folks. But if you had the equipment and knowledge to use it, you would definitely be singing a different tune.
Panasonic used to claim knowledge in specialized audio construction electrolytics too. They used to have the PZ,PX,PXL Pureism and Master Ta-Keh lines. I had even seen an early to mid 90s Technics RS-B965 top of the line cassette deck with many Pureism caps inside.