You are right. The reality is that I was testing these not just for grins, but with a specific use in mind. Since I had a strict size constraint with the mint tin amp, I just bought all the largest capacitance values of each type that would fit. This skews the test against the Elnas and the Muse to some degree, since they are so huge that I was forced to try smaller values for those types.
XRK971 and BabyDontHertzMe are working on a desktop version of this amp. When I get one of these new boards, size will be moot, so I will do a new test with identically spec'd caps.
Thanks for the heads up, Tom. These look promising indeed! Wow, these guys say that even their standard series is superior to Cerafine and Silmic...
Audio Note
Reasonably priced as well. PartsConnexion has them in the US
Last edited:
Yeah, after this experience, my next comparison will only include the best. Most likely will be KZ, ES, FR, FC, Silmic, and hopefully one of the new Audio Note types.
I have the impression that the caps you're testing for a signal o/p on a headamp are more suitable for power supply applications, Stella, but I might be way off the mark with this
For me, the 'best' treble sounding electro for the o/p on a headamp would be the old Rifa peh169, renowned as the 'fastest' electro caps ever made
For the 'best' midrange, the prize would have to go to Denis Morecroft's 'Slit Foils' caps, now available from Supertec
For the 'best' bass, it's a bit of a crapshoot between the thick foil Siemens/Epcos B41550 and the Nichicon SuperThru - neither in production for some time but there's modern equivalents, maybe some of the Mundorfs?
Nearly all of the modern cap manufacturers are building smaller sized devices and this will naturally change the way they sound, so my selections are probably well out of date, plus there's plenty of other well-known caps that I have little knowledge of
And, yes, these new caps from AudioNote look very interesing
So, not taking anything away from your good work in testing and defining the differences in the caps you have used, which I find extremely useful and informative indeed, and hoping you have a chance to continue with a wider selection of caps
For me, the 'best' treble sounding electro for the o/p on a headamp would be the old Rifa peh169, renowned as the 'fastest' electro caps ever made
For the 'best' midrange, the prize would have to go to Denis Morecroft's 'Slit Foils' caps, now available from Supertec
For the 'best' bass, it's a bit of a crapshoot between the thick foil Siemens/Epcos B41550 and the Nichicon SuperThru - neither in production for some time but there's modern equivalents, maybe some of the Mundorfs?
Nearly all of the modern cap manufacturers are building smaller sized devices and this will naturally change the way they sound, so my selections are probably well out of date, plus there's plenty of other well-known caps that I have little knowledge of
And, yes, these new caps from AudioNote look very interesing
So, not taking anything away from your good work in testing and defining the differences in the caps you have used, which I find extremely useful and informative indeed, and hoping you have a chance to continue with a wider selection of caps
PEH169, B41550 and Super Through are all available from Mouser
. Super Through refers to the KG series with a voltage of 50+, lower voltage ones are called Gold Tune.
They're all pretty huge though. Definitely going to need that desktop board...
With you here, only both imho have insufficient bass with the Rifa pulling towards brightness.
Admittedly my impressions from both are from PS duties. The Rifa i have in various voltage ratings and all sound disturbingly different. Some are much better balanced than others. Very, very annoying.
I think you went to far , this was suppose to be a Coupling capacitor for Headphones amplifier, the slimmest peh169 is twice thicker than nichicon kz 1000uf/50v or SILMIC II 1000/50v (not small caps either) ....And the cost, 7-8 times more...
I think that Kz/es/fg and SILMIC II are, for the money and size THE top contenders here .
And these caps are available and affordable , that's great! We don't have to wait for a shortage or for them to be discontinued products to realize how good they are...
I think that Kz/es/fg and SILMIC II are, for the money and size THE top contenders here .
And these caps are available and affordable , that's great! We don't have to wait for a shortage or for them to be discontinued products to realize how good they are...
Last edited:
I probably created a misleading title to this thread topic. Technically it should have been "Miniature electrolytic capacitors shootout."
I'm happy to know about these other recommended caps, including the huge snap-in and even screw terminal types, though of course there is no way I could have fit either in a pocket amp. Maybe I will try some of the really big guys in the desktop build. Now I am very curious.
It would be interesting to know, for instance, how Nichicon's monstrous KG, which is also manufactured "for audio," compares to their "high end" miniature KZ. I get the idea that KG are more intended for power supply than signal coupling.
I'm happy to know about these other recommended caps, including the huge snap-in and even screw terminal types, though of course there is no way I could have fit either in a pocket amp. Maybe I will try some of the really big guys in the desktop build. Now I am very curious.
It would be interesting to know, for instance, how Nichicon's monstrous KG, which is also manufactured "for audio," compares to their "high end" miniature KZ. I get the idea that KG are more intended for power supply than signal coupling.
Last edited:
If you are interested to try the "perfectly linear electrolytic capacitor" rule of thumb found in Douglas Self's book Small Signal Audio Design (2nd edition), Figure 2.21,
then I recommend you extract the required "AC current delivered into headphones" from the handy table in this web page. When I quickly scan the table, I observe that AC current delivered into headphones is 50mA or less, for about 95% of the headphones listed.
Armed with the necessary numerical values, we simply plug and chug in Self's rule of thumb. AC voltage = 0.08 , AC current = 0.05, frequency = 62 radians/second. And presto, out pops the answer: C >= ten millifarads. (ten thousand microfarads).
Here's a capacitor that sells for $1.44, qty=1, and meets these requirements: link . Remember that Self's data says ANY electrolytic capacitor, from ANY manufacturer, will be perfectly linear when delivering 50mA to a load* ... as long as that electrolytic capacitor is 10,000 microfarads or greater. Even the $1.44 capacitor shown here.
*at frequencies of 62 rad/sec and higher
then I recommend you extract the required "AC current delivered into headphones" from the handy table in this web page. When I quickly scan the table, I observe that AC current delivered into headphones is 50mA or less, for about 95% of the headphones listed.
Armed with the necessary numerical values, we simply plug and chug in Self's rule of thumb. AC voltage = 0.08 , AC current = 0.05, frequency = 62 radians/second. And presto, out pops the answer: C >= ten millifarads. (ten thousand microfarads).
Here's a capacitor that sells for $1.44, qty=1, and meets these requirements: link . Remember that Self's data says ANY electrolytic capacitor, from ANY manufacturer, will be perfectly linear when delivering 50mA to a load* ... as long as that electrolytic capacitor is 10,000 microfarads or greater. Even the $1.44 capacitor shown here.
*at frequencies of 62 rad/sec and higher
Last edited: