nonpolarized caps

You can place THT films over the SMD components, so no space occupied ;-)

This isn't horrible:

and nonpolar Aluminums are too . . . ?

For what kind of service use? Ripple current? Frequency?

Back-to-back 100V polymers, center-biased to -50V?

It's too far from Christmas to go and ask Santa Claus.

RL

Output filter in a 8-channel bipolar switching power supply. The PCB layout will be crazy dense.

I asked for experience and advice, not snark.

Am 10.03.25 um 16:30 schrieb john larkin:

I have used these WIMA capacitors (10uF version) in the input of an ultra low noise FET amplifier. Absolutely no complaints.

Gerhard

Yeah, film is ideal if I can make them fit. As noted, the board will be crazy dense. My production and test people might make funny faces if I suspend giant axial power resistors and caps above surface-mount parts. How about a baby board on hinges?

The design is actually an 8-channel programmable dummy load that simulates both resistance and inductance. But it occurred to me that, having eight isolated bipolar-output class-D amps, we could futz with the output filtering and get an 8-channel power supply from the same PCB layout.

I've wondered if the bipolar aluminum caps work right after being DC biased one way for a long time.

There is one dirty trick of putting two alum caps in series opposing, with a diode across each to prevent back-bias. That's crazy nonlinear but might be OK as a power supply filter.

Maybe jamming film caps on the board is best.