psu bank with silmic 2?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
There is a huge difference in the sound of Elna Silmic caps and Nichicon KG power caps so it's a bit hard to directly compare them. Using 10 x 1,000uF Silmic caps in parallel has been tried and discussed a lot in other threads.

Personally, I tried this idea as filter caps in power supplies with Panasonics, Silmics, Nichicons, Epcos, Rifa, etc and unfortunately, none of the parallel smaller caps seemed to work for me (including Silmics) in the F3 and F5 power amps - I did do okay using about 3 parallel caps for supply filters, but more than that didn't seem to achieve any benefits that I could hear and snubbing the resonances was a real PIA.

There does seem to be plenty of other people that have achieved good results using the smaller parallel caps for power supply filters, assuming that is your requirement, and the existing threads are full of designs and experience.
 
Member
Joined 2010
Paid Member
Why ever do you look for low ESR in a power supply for class A? The best kinds of capacitors for filtering roles will be somewhat lossy, not like a zero ESR source.

It once seemed a big deal to have hundreds of tiny low ESR caps on show (look at all the Eastern European pics from about 3-10 years ago.) Maybe the idea has come full circle now as the weak points of such power supplies, like ringing and oscillation rear their heads. You certainly don't see these supplies in top-flight gear any more.

Think this through rather than just assuming what seems like a good idead actually has merit in the application(s).
 
Last edited:
I used to think a flashy main reservoir was the way forward. For example, with film caps as well as electrolytics. Then I read about the potential for oscillations with this topology and the need for an RC snubber when film caps are used.

Further reading suggested that all the main psu reservoir does is charge the smaller local decoupling caps. Which meant all the main reservoir needs to be is big and basic.

If any change in SQ is to be found it is likely to be at the local decoupling caps.
 
I used to think a flashy main reservoir was the way forward. For example, with film caps as well as electrolytics. Then I read about the potential for oscillations with this topology and the need for an RC snubber when film caps are used.

Further reading suggested that all the main psu reservoir does is charge the smaller local decoupling caps. Which meant all the main reservoir needs to be is big and basic.

If any change in SQ is to be found it is likely to be at the local decoupling caps.
It took me a long time to come to this realisation.
 
As you said "class A", the goal is to get a silent PSU at constant high current.
Lytics need to come with proper ripple current rating and you want as much
capacity as you can afford. I'd suggest 3 times as much as for AB. Consider
to design the PSU at least "CRC", better "CLC" with sufficiently rated inductors.


BTW:
http://www.diyaudio.com/forums/parts/151392-best-electrolytic-capacitors.html
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.