tantalum caps instead of electrolytics as coupling caps?

[dramadisease]

good idea or bad idea?

 

[microx]

In my experience tantalums are better caps for audio but they are prone to making crackling noises when they age. The modern ones may well be better quality.
Steve

 

[Steve Jones]

You have to be careful of making sure they never go reverse polarity also if you use them for audio coupling, or they eventually throw a fit.

 

[VacuumVoodoo]

Tantalums are ok for decoupling, but as Steve said, not very well suited for signal coupling. One can use 2 caps in series with reversed polarity + - - + for that purpose or use nonpolarised tantalum caps. I refrain from commenting on the sonic aspects of tantalums as coupling caps.

 

[Svart]

A BIG no no. use polypropylene or styrene, heck even teflon if you can afford em and rest assured you can't get anything better for audio coupling. use polyester polypro for decoupling.

 

[gyraf]

It's not all that black-and-white,

..the Neve, Calrec and NTP stuff we all like are full of these things. The important aspect is to come up with a technically correct design, as tantalum caps are much less forgiving than electrolytics (or other caps)..

But please do not use them for psu decoupling - unless you have local fuses on your pcb. Where electrolytics tend to go open-circuit when they fail, tantalums has a bad habit of shorting..

Jakob E.

 

[Svart]

yes Gyraf, agreed.. However, if someone is asking such a question, then I am sure they aren't familiar with advanced design using tants in the first place.. :wink: :guinness:

 

[DrFrankencopter]

You'll want a steady DC bias across a tantalum cap used for couplig, otherwise you'll get some distortion from it. Neve could get away with it because they use single supply circuits (+24V), so everything had to be biased anyways.

Using a DC bias is good for aluminum electrolytics as coupling caps too...it helps to linearize the cap's performance. One design trick that I haven't seen all to often is to take to polar electrolytics, and put them back to back (attach their anodes together) and tie the adnodes to the negative supply rail through a resistor. This biases each electrolytic cap, and the whole thing is non-polar. However, this might start to cost as much and take up as much space as a good film cap.

Cheers,

Kris

 

[dramadisease]

thanks for all the info!

i do understand enough to know that when tants go bad, they short - bad for psu decouping - but i was mostly wondering about sonic aspects of these caps. im going to be replacing some electrolytic signal coupling caps in some mic pres and was wondering if a tant would sound better? im always in such a rush when i write posts! forgive me for that.

Frankencopta` -> im curious of the trick you mentioned with the electrolytics - when you say it will linearize performance, are you saying that it will have less phase shift and a greater frequency vs. amplitude balance? by annodes you mean neg side i assume (note the ass in assume)
i have the space for doing that, but film caps @ 10uf are costly...and bigger...
im basically gearing up to expierment with lots of coupling caps in ss circuitry - i love doing it in tube stuff so i figure i should learn somemore with the transistorized curcuits and caps.
how much of a bias is appropriate? 3v? 5v? .5v?
any tips on resistor value?

thank you all so much for this info!
-bryan