the choice of the output coupling capacitor

[Dominique]

Hi,

Lately I was thinking about output coupling capacitors.
(more precisely the one of my mic preamp, but this relates to all output caps going into some "line-in")
Until now I used polar electrolytics, because I wasn't aware how low "line" input impedances could go.

But the lowest I could find are not lower than 10k Ohms, even with the portable devices - well - at least among the few data I could find.
E.g. my iriver IFP380T has about 12.5kOhm, my soundcard M-Audio (forgot the type) 10k, most other stuff had at least 20k, but rather 47k.

To calculate the coupling cap capacitance while having some low frequency margin I chose 8Hz as the corner frequency (-3dB) with that lowest input impedance I could find - 10k Ohms. The resulting lowest cpacitance needed is 1,99uF.

In my spice program, the output offset voltage after the coupling cap was lower with an increasing capacitance - is there any relation like that?
I imagine that an electrolytic with it's rather high leakage current produces a higher offset voltage than a polyester cap, doesn't it?

Are there other things to consider when choosing the output coupling cap that I may have overlooked?

If not - with the calculated value, I'd not longer stay with electrolytics, but go for a 2.2uF to 3.3uF polyester cap.
While further up the quality ladder, polypropylene would be too big and expensive for my needs.

Cheers
Dominique

[richie00boy]

All good so far, but you need to double your cap value. The thing you are feeding will have a cap at the input and, say for arguments sake it's the same value as yours, then 2 caps in series halves the value.

This also holds true for the input capacitor we talked about in your other thread.

In my own designs I use 10-100uF for o/p caps and whatever the smallest cap I can get away with for corner below 10Hz for the i/p cap. I try to use non-electro caps such as polyester box where possible.

[Dominique]

Originally posted by richie00boy
All good so far, but you need to double your cap value. The thing you are feeding will have a cap at the input and, say for arguments sake it's the same value as yours, then 2 caps in series halves the value.

Oh yes, of course! Thanks for pointing me at it!

I also already wondered about leaving the output cap away, because almost every input should have a cap, and with my mic preamp, size, sound quality and money ) is an issue.
So I thought one cap is better than two in series.

Or maybe using a smaller cap (like 2.2uF which is should be ok for most inputs) and leaving the possibility to bypass it with a jumper...

What do you think?
Cheers
Dominique

[Sch3mat1c]

You also have to consider output impedance. Ok, I suppose you are doing SSatanic rituals, mmmh? Still applies.

My preamp over << there has some .1 or .082uF on the output, something like that... sounds fine into the >100k tube amps it runs. But it also sounds fine into the 32 ohm headphones I have (until the preamp runs out of headroom, that is ). This is because the amp has follower outputs of maybe 500 ohm impedance. Bass is cut off a bit, but the headphones are boomy anyway.

The disadvantage of a high impedance output is reduced gain feeding low impedances. I only get a few mW into my headphones, and gain has to be nearly cranked.

Oh, and why is 8Hz your chosen F3 anyway? I doubt you can hear (not feel) 20Hz!

Tim

[ashok]

I've been experimenting with coupling capacitors between a tubed preamp and solid state power amp.
here is what I found .. the best is on top.
1. paper in oil cap ( Jensen) ( 0.33u--400V)
2. Poly propylene ( CTR ) ( look like Solen caps) ( 1u--250V)
3. Wima film MKP10 ( 0.47u--250V)
4. Philips polyester. ( 1u-- 250V)

( caps had about 1.2 volts dc across them )

2 and 3 are are almost similar.
1 is really different. You can't make out the differences on casual listening. If you listen carefully you will find that (1) clears up the overall sound. Struck strings --- you can hear each and every string being struck ( in a chord) . With the other caps they sound more and more like one sound. (4) was audibly the worst.

I have some other types of caps but don't have the time to test them out.
Moral ....... test every type of cap you have before deciding what to use. Simple stringed instruments might be one thing to listen to for differences. Guitar strumming can sound very different with different caps. The source has to be good also !
Cheers.

[alecwek]

hey i read spme where ...
something like this:
....anything with coupling caps or transformer outputs can not be concidered audiophile quality....
they were taking about amplifiers....but still anyone know why this is?
...sorry to but in to ur thread.....

[ashok]

The best coupling device is no cap or transformer . Even a series resistor is suspect. ( did someone say 'wire also' ?.............)

So if you must use one of these you will have to check out their performance. Often it also needs a good bank balance !

It is a good idea to try and level match parts of your system and achieve zero dc at the outputs. Not always an easy task.
Cheers.

[Dominique]

thank you for your input!

I'll go for 4.7uF polyester caps at the moment.

Cheers
Dominique