polymer electrolytic capacitor distortion

[twest820]

I'm pulling a blank when searching for distortion measurements of polymer 'lytics. Anyone happened to have taken a look at them? It's a back burner project for me, but I've a bias stabilization application where it seems interesting to try something like Nichicon FPCAP or Cx/Lx series parts in addition to standard choices like Elna RFS or Nichicon KZ. Figured I'd check to see if I'd be repeating existing work before ordering parts.

 

[jcx]

I think Bateman's work pertty conclusively points to the oxide dielectric properties as being the determining factor in measureable electrolytic distortion

Al much better than Ta, and likely Nb

"bipolar" - full thckiness oxide on both foils better than polar caps - I don't think anyone makes Bipolar polymer electrolytics

higher V rating better, again polymer electros are lower V rated

so unless volume, high frequency properties are important I doubt current polymer electrolyte caps are much use for audio

 

[SoNic_real_one]

IMO, if the circuit is designed properly, there is no need of bipolar capacitors. Just keep the + terminal on positive voltage during the whole AC swing.

 

[Lazybutt]

They might be good as psu bypass, due to their low ESR and capacity value at high frequency. The Lampizator swears by them in DACs.

Problem for audio is their limited voltage ratings availability

 

[twest820]

I doubt current polymer electrolyte caps are much use for audio

App note 176 unfortunately doesn't contain distortion measurements but, in my experience, Wolfson is a reliable source. So it's interesting they seem to recommend polymer 'lytics over C0G for AC coupling. The leakage makes them unsuited for the integration tank of a DC servo and the 35+V parts do get a bit pricey. But 50 cents or a couple dollars isn't necessarily a significant in DIY quantities and one can choose a 1uF PPS cap or similar for a servo. 2.5V parts are suffient for bias stabilization (though I'd spec 6.3+V as there's negligible downside to doing so) and if you're working with preamps, DACs, or smaller solid state amplifiers the readily available 10 or 16V polymer 'lytics are likely sufficient.

They might be good as psu bypass, due to their low ESR and capacity value at high frequency.

That and a few other things; have a look at Sanyo's collateral for the Oscons. Nichicon's FPCAP docs are typically terse (despite having acquired FPCAP from Fujitsu) but basic model parameters for some of the caps are on page 53 of the catalog.