Back to Back Polarized Electro's

[EchoWars]

I thought that if you wanted to make a bipolar out of two back-to-back polarized electrolytics that you connected the negs to each other, and the positives to the outside..

device > + - - + > device

However, working on a few old amps where they have the positives together, and the negs to the outside...

device > - + + - > device

Perhaps it doesn't matter. Is that the case?

[thoriated]

To me, thinking about the workings of back to back or nonpolar 'lytics (same thing) is like thinking about how sausages are made or politicians do their thing. Whatever is supposed to happen sorta does, even though the hidden details are fairly gruesome. So connecting the pluses vs the minuses to make a nonpolar probably works equally unwell. If you want to improve the situation, try adding a low current dc bias (through a highish value series resistor) in excess of the peak value of the signal to be passed to the tied together node of the 'lytics.

[EE_Mark]

It doesn't matter whether the + terminals join or the - terminals join. It makes absolutely no difference whatsoever. I either case one of the caps will be reverse biased and will look lie a low impedance so the voltage will dropn across mthe capacitor that is seeing the correct polarity.

[EE_Mark]

It doesn't matter whether the + terminals join or the - terminals join. It makes absolutely no difference whatsoever. In either case one of the caps will be reverse biased and will look like a low impedance so the voltage will drop across the capacitor that is seeing the correct polarity.

__________________

[Cobra2]

But since negative terminal is often connected to the can, it would be more practical to connect the negative terminals.
(just a good idea if they are close or naked...)

Arne K

[jcx]

Cyril Bateman talked about al electros in his "Capacitor Sound" series in Electornics World mag last year, with a recent amp re-cap distortion measurement follow up just this month

basically he claims the thin spontaneous film on one electrode in a polar cap causes large distortion, for lowest distortion he recomends bipoalr electros which are made with full thickness oxide on both foils, further lowering V stress on the oxide by series connecting bipolar electros is his strong recomendation for lowest distortion in large value cap applications

[EE_Mark]

Originally posted by jcx
Cyril Bateman talked about al electros in his "Capacitor Sound" series in Electornics World mag last year, with a recent amp re-cap distortion measurement follow up just this month

basically he claims the thin spontaneous film on one electrode in a polar cap causes large distortion, for lowest distortion he recomends bipoalr electros which are made with full thickness oxide on both foils, further lowering V stress on the oxide by series connecting bipolar electros is his strong recomendation for lowest distortion in large value cap applications

I think what he's trying to say is I'm right. If you're really worried about it drop $5 and buy some Sanyo SG Series Os-Con caps.

http://www.bdent.com/

[Elso Kwak]

Originally posted by EchoWars
I thought that if you wanted to make a bipolar out of two back-to-back polarized electrolytics that you connected the negs to each other, and the positives to the outside..

device > + - - + > device

However, working on a few old amps where they have the positives together, and the negs to the outside...

device > - + + - > device

Perhaps it doesn't matter. Is that the case?

HiEchoWars,
Better use bipolar electrolytics if you have to use elco's.

[EchoWars]

Naaa....not really doing anything to this amp, it runs fine...just saw the back to back electrolytics with positive sides connected wheras I've always seen the negatives connected.

Thanks for the reply guys...always learning something.

[SY]

EW, we had a long thread about this a few months back which included some experimental resuts (including my clumsy attempts at scope photography). It's probably worth doing a quick search.