creatiph:
thanks for the great posts[quote author=Terry King link=topic=52354.msg373916#msg373916 date=1297832492]The diode would need to be in parallel with the cap, "pointing down". In series would not allow current to flow OUT for Q2 to swing negative.
yes, but then the capacitor would discharge directly through the diode (?), that would be worse 
The best solution is indeed use a simple polarized cap. It's not so bad since I have a dozen of them.
[/quote]
No, the diode "pointing down" will not discharge the cap if the cap is charged in the normal situation in which it's "top" end is NEGATIVE. It will only conduct if something tries to drive the voltage at the top of the Cap POSITIVE. So in normal operation the diode has no effect. I think it's the easiest solution: diodes are very easy to find; non-polarized caps are more difficult.
But you're in the best position to prove or disprove all this!
Please run both the non-polarized cap AND the polarized cap + diode, and tell us your results.
Mellis, you want a side bet ? ?