Output Caps Polarity

[Samuel Groner]

Hi

On most schematic I see 'lytic output caps having the negative polarity to the outside world; for best phantom power survivability I'd rather have done it the other way round; any reason against this? Perhaps the outer foil is connected to + and for best shielding we want it look to a low impedance?

I'm sure this has been covered but my search didn't turn up the right thread... Thanks for guiding me further!

Samuel

 

[SSLtech]

I always "show the world the positive plate" on outputs for exactly that reason. Other than that, there's no reason to choose either way, since random DC offsets and polarities will always happen 50% the way you chose and 50% the other way.

However...

Phantom is predictable, and while it should never happen, it usually will!

Keith

 

[bcarso]

In consumer equipment with single supplies and half-supplies derived from them, the tradition is to have a final 'lytic at an output with a 100k ot thereabouts to common. The preceding point is positive so the 'lytic has - to the 100k. Then most inputs (again cheap consumer stuff) have again either nothing, or another 100k or so from tip to sleeve/common and again a positive voltage at the input on the other side of the coupling cap, so once again the - side faces the outside world.

But of course none of this equipment ever is expected to handle a big positive offset, let alone a phantom power voltage.

When I design with coupling caps internally, when they are really needed I look at the polarity of the interfaces even when they are nominally d.c. zero. The op amp input offset voltages have an unpredictable component, but many times the bias currents flowing in the external resistances will cause a predominant shift in output and input voltages, so there will be a preferred polarity for a 'lytic coupling cap.

 

[Samuel Groner]

Thanks for the answers. It looks like there is really no reason for not doing so.

Samuel

 

[TedF]

This thread reminds me of the many pieces of equipment that came back to the service department 'not working' ; some from very respectable studios, and on investigation we found that the output stages had been blown apart by either phantom power or even , in some cases, mains!

The BBC used to put microphone inputs on jackfields, and I never got used to that idea... it always seemed like asking for trouble, even if you were well disciplined... When most output stages became transformerless it was obviously suicidal.... accidentally plug a mic input (with phantom) into a conventional op-amp output stage, and it will go 'pop' whichever way round the cap is!

 

[Samuel Groner]

Well, if you add a 6.8k shunt resistor, use 35 V caps and protection diodes to the supply rails, the amps should survive things, no? At least my A&H board did once...

Samuel

 

[TedF]

Yes, true.... but a service engineer (working on power amplifiers) once said to me; "those output transistors are there to protect the fuse!".

To get back to the topic; where there's no voltage, I go for +ve towards the outside world.... but I don't have a logical reason.