Thanks for all the input. After seeing the possible negative effects I think I will not use these caps in the older Audio equipment I had intended but will probably use Elna SimilacII or nichicon muse & mabye bypass with a polyprop.
These solid type caps do seem like they would be beneficial in newer circuits designed for lower esr caps. IE products with switching power supplies. I know products with switch supplies aren't generally seen as audiphile components but if someone was modifying a lower quality product such as a low cost DVD player for better audio & video quality this might be a good for the bypass caps in the main supply. What you guys think?
I've gone back to not use capacitors at all in the audio path, DC-coupling and active components for ground. It sounds a lot better, and I don't want to get a capacitor neurosis. Those things are too large and too expensive for my taste.
I agree -My equipment which doesn't have them directly in the audio path does sound better. I'm thinking manufacturers use them at the input of an amp or output of a preamp for protection in case someone connects another component with a problem to the device.Would you know a way to modify the circuit in items that do use them in the audio path or would that be a project you would have to design from the ground up?
I've just recapped a Behringer SRC2496 with Nichicon PLF caps. I also did the standard output stage bypass (adding a Sowter 3603 per channel) at the same time, so I don't know which change brought about the significant improvement in SQ.
A while ago I replaced the BG standard PSU caps in my DAC-AH with NCC solid polymer, which made a significant difference. The BG's sounded warm and fat where the SP's sounded fast and clean (more extension and resolution). I stuck with the polymers ;-)
After changing the caps in the DAC-AH and SRC2496, there seemed to be a 4-6 hour period when the sound was almost edgy and lacking in warmth. After this period the warmth returns but the resolution and HF extension remains. I'm well aware that this could be partly psychological, but I have another system in a different room for comparison so I'm sure it's not just me...
I'm surprised that there have not been more reports of experiments with these caps. Given the vast number of discussions on BGs, silmics, muse etc. I'd have thought people would be all over these caps (given the many technical advantages).
A while ago I replaced the BG standard PSU caps in my DAC-AH with NCC solid polymer, which made a significant difference. The BG's sounded warm and fat where the SP's sounded fast and clean (more extension and resolution). I stuck with the polymers ;-)
After changing the caps in the DAC-AH and SRC2496, there seemed to be a 4-6 hour period when the sound was almost edgy and lacking in warmth. After this period the warmth returns but the resolution and HF extension remains. I'm well aware that this could be partly psychological, but I have another system in a different room for comparison so I'm sure it's not just me...
I'm surprised that there have not been more reports of experiments with these caps. Given the vast number of discussions on BGs, silmics, muse etc. I'd have thought people would be all over these caps (given the many technical advantages).
Heres an update to my capacitor projects. In the Nak amp I replaced the caps in the audio path with Elna similacII's. They sound good after burning in for 50 hours. They were a bit tinny sounding to begin with. My next cap project didn't go as well. I replaced all the lytics in a Krell CD player with Nichicon KW in Main power supply & Nichicon KZ ath the regulator section & audio sections. It sounded great when I finished - a definate improvement over originals but it only lasted about a month. Now the bass is muddy & treble is raspy. I suspect one of caps at the front end has went bad. Has any you guys had caps go bad so fast? I suspect they must have been setting on the shelf for awhile since not many people buy audio grade caps anymore. I was just thinking mabye I should go with non audio grade in the main supply 6800uf at 105 degree just to make sure I get something fresh this time. Any suggestions on a decent large value cap.
I was just looking at the spec sheet of some newer nichicon 16 volt 47uf polymer caps. The ripple current is about 2200ma which is about the same as a regular 6800uf 16 volt cap. Can I replace a 6800uf bypass cap with this 47uf cap at the front end of the supply since the ripple current is about the same or will the lack of capacitance cause problems in a regulated supply. I heard you could do something simular with a switching supply but the example was not as extreme with these value differences.
That 6800uF cap sounds like a smoother/reservior.
The answer to that is no pretty much - you would have much more noise and harmonics on the supply line. This could affect the circuitry working properly etc downstream.
The extra capacitance smooths the waveform (ripple) quite seperate to the ripple current ratings. It forms part of a filter. The larger the capacitance the lower the time constant and more attenuation of noise, ripple etc. The Polymer cap you mention is I suspect spec'd with respect to ripple current at 100 KHZ - there for Low ESR apps / SMPS etc. not LF mains frequencies.
Best to show a schematic if you want some tips on what to improve.
Thanks
Last edited:
Thanks Fanuc,
That makes sense (100khz) since these caps are generally found in computers & switching supplies. I've noticed most audio manufacturers still use generic 85 degree caps at the main supply. I guess size matters more at the front end for the reason you mentioned. I did get my hopes up-it would have been nice to be able to design a small high quality supply with these lower value polymer caps if they worked like the big ones.
Back to the original topic of lytic vs polymer for audio. I just read that a new tantalium polymer cap has been developed with low dc leakage which was one of the drawbacks of polymers for audio circuits. I will send a link when I find the article again.
That makes sense (100khz) since these caps are generally found in computers & switching supplies. I've noticed most audio manufacturers still use generic 85 degree caps at the main supply. I guess size matters more at the front end for the reason you mentioned. I did get my hopes up-it would have been nice to be able to design a small high quality supply with these lower value polymer caps if they worked like the big ones.
Back to the original topic of lytic vs polymer for audio. I just read that a new tantalium polymer cap has been developed with low dc leakage which was one of the drawbacks of polymers for audio circuits. I will send a link when I find the article again.
Here's a note I snipped from another thread a while back. It was member Stephensank who wrote it, but I don't know the thread off hand:
"After 30+ years of replacing failed and/or crappy electrolytic caps, I can tell you that Panasonic, Sanyo/Oscon, Elna(incl. silmic & cerafine), Rubycon(incl. blackgate), Illinois Capacitor, Samwha, Mallory/Jamicon, Sprague, etc., have terrible long term reliability compared to Nichicon & Nippon/United ChemiCon. And I have yet to hear better performance than Nichicon, with ChemiCon often being a quite close second. Rubycon & Elna have, by far, the worst reliability of caps made in Japan, so it makes me nauseous when "upgraders" sell people on BlackGate & Cerafine. Panasonic & Sanyo are almost as bad."
what about this kind of capacitors for use in power amplifiers?
http://www.muecap.de/fileadmin/pdfs/dc-folien/MUECAP-LNK-P5.pdf
http://www.muecap.de/fileadmin/pdfs/dc-folien/MUECAP-LNK-P5.pdf
Just stumbled on this thread,
RNS1E100MDS1JT | Nichicon 10μF 25 V dc Polymer Aluminium Electrolytic Capacitor, NS Series 2000h 6.3 (Dia.) x 7mm | Nichicon
These are still small in voltage/cap values, however look at the leakage current. Not as good as electrolytic in the leakage department but looking respectable compared to other polymers, anyone here tried these values?
RNS1E100MDS1JT | Nichicon 10μF 25 V dc Polymer Aluminium Electrolytic Capacitor, NS Series 2000h 6.3 (Dia.) x 7mm | Nichicon
These are still small in voltage/cap values, however look at the leakage current. Not as good as electrolytic in the leakage department but looking respectable compared to other polymers, anyone here tried these values?
creek uses ordinary high leakage polymer capacitors* in the signal path of its amplifiers. i have a evolution 50a, which uses them in the bass-control feedback-loop of the pre-amp and on the power amplifier as input dc-blocking, inside the feedback loop (and for idle-voltage stabilisaton). the pre-production schematics shows electrolytic caps, but the final product uses only polymers in the signal path.
the higher leakage seems not to be a problem because they block only small dc-voltage and i assume that creek had measured the resulted low frequency ac distortion.
so maybe the non-linearity of the electrolytic capacitors esr is a bigger problem than the non-linearity of polymer leakage currents. (but booth depends strongly on temperature too - the electrolytics to the better, the polymers to the worse...)
the strange thing: for a dc-blocking input capacitor esr is not so important, but leakage is. the opposite is the case inside the feedback loop.
*100uF 16V, type unkown, but i measured the usual high leakage currents under dc-voltage
the higher leakage seems not to be a problem because they block only small dc-voltage and i assume that creek had measured the resulted low frequency ac distortion.
so maybe the non-linearity of the electrolytic capacitors esr is a bigger problem than the non-linearity of polymer leakage currents. (but booth depends strongly on temperature too - the electrolytics to the better, the polymers to the worse...)
the strange thing: for a dc-blocking input capacitor esr is not so important, but leakage is. the opposite is the case inside the feedback loop.
*100uF 16V, type unkown, but i measured the usual high leakage currents under dc-voltage
Last edited:
That's good info about the creek. The first oem I've heard of to use polymers like that-maybe some designs are changing. Since staring thread 8 years ago I've ended up replacing caps in the signal path with either Nichicon FW or VP. VP is non polar. I stumbled onto the FWs when noticing Yamaha typically uses about 7 or 8 of these in the signal path of each channel of their receivers or amps. Using that many and it still sounding respectable made me think they have low coloration. Been using ever since with good results.