Parallel Caps and Power Amp Hum

[jean-charles maillet]

>I've also heard that it not a great idea to
>put smaller caps in parallel with big smoothers as it can create a
>parallel LCR with the smoother's inductance. Not sure about this one
>either.

In theory ideal caps and inductors have linear impedance frequency
response, though in practice capacitors turn inductive at high frequencies
so that their frequency response looks like a V instead of a monotonic
straight line (goin' down with frequency ...).  I guess there could arise a
problem when the paralle cap values are chosen too wide apart, like say a
100uF and 1nF cap stuck together but this would depend on bonding wire
inductance levels too which can be tightly control depending on fab
quality.  In low voltage IC applications it is recommended practice to
stick a 10uF and 0.1uF cap side by side to shunt high frequency energy
which are inductive to the larger cap and also prevents parallel LC
oscillations though this is always worse case as is usually the case with
app notes (i.e. you could do without the smaller one until you find out one
day, say at a live gig, that you need it ... but I'm digressing).

>I have a big problem with 100Hz noise on my studio 150+150 amp. Maplin
>MOSFET kit. I've tried every possible combination of earthing there is,
>and the noise is still there. It only has about 200mV of 100Hz, it's a
>55volt supply, but it still audible on my speakers. Some things are
>left to try us!

>Tony Allgood.

look for a ground loop between the power supply filter and the front of the
amp.  That is, make sure that the ground return of the filter caps and the
audio circuit (especially the front) are separate.  If the chassis is used
as a common return there's your problem, in which case the grounds need to
be star-ed or
*'ed. Alternately you may want to stick a low resistance (22 Ohm with 10uF
cap in parallel) between chassis ground and front of amp and shouldn't
alter response of the amp significantly.


hope this tickles your monitors ...

jc