Electrolytic coupling capacitors ?

[heater]

I have seen a number of amplifier schematics around that use electrolytic coupling capacitors on the input. One example is this gainclone by plastichead here http://www.plastichead.net/audiowork...schematics.htm

Now I always understood that the thing about electrolytics was that they are polarized and should always be biased the right way around (unpolarized electrolytics aside).

But, as a decoupling cap in such a schematic, the amplifier end is sitting at ground potential while the input end is alternating plus and minus a volt or two with the signal. So at least fifty persent of the time it is biased backwards.

Worse still the whole point of this cap is to block DC offsets on the input, which may be positive or negative, so the situation looks even worse to me.

So without any discussion of the "sound" of electrolytics (yet ) how is this acceptable? Why doesn't the cap fail prematurely? What am I missing here? How did plastichead and others determine which way round the cap should go?

Thanks all.

[cetoole]

I would guess that most people who use 'lytics as coupling caps use something like the Blackgate NX-HQ, which is actually nonpolar. While these are very good electrolytics, I personally wouldnt want them in my signal path, preferring to be DC coupled if at all possible, or using decent film caps, ideally polypropylene film-and-foil.

[heater]

cetoole - I agree no caps would be best.

Apart from sonic issues, buying blackgates or large polyprops is going to double the cost of my amp judging by prices and availability in Helsinki. I'm looking to build four channels.

Now many will say I'm crazy but I would really like to put Rod Eliots active balanced receivers on these amps with nice XLR inputs. http://sound.westhost.com/project87.htm

I need to be able to reduce those caps to 1uF or less to get the price down. Or live without them which seems unwise. Or be happy with my electrolytics, which seems crazy to me just now.

By the way here is another example of electrolytic coupling caps, this time on input and output: http://users.swing.be/edwinpaij/controle_tonalite.htm

Anyone out there know how this is possible ?

[analog_sa]

The general thinking is that a few hundred millivolts reverse bias doesn't matter.

[neutron7]

if you use 2 electrolytics in series, back to back than you have made an non polarized capacitor. (at half the value) then you can put a nice capacitor 0.2 or so in parrallel with both of them.

[heater]

analo_sa - A few hundred millvolts? Like may be 2 volts of my CD player output? Do you have any references for this general thinking?

neutron7 - Interesting idea. Don't we end up with a non-polarized that sounds twice as bad as a single elecrolytic?

At the moment I'm not sure I could ever hear the difference between different caps, or no cap. Guess I'll just have to try them.

[lineup]

A no-name filmcap will beat any 'audiophile' electrolytic as coupling cap.
Even a polyester will do it.
I know some will scream and disagree. But question is if they really go for high fidelity?


One 2.2uF polypropylene cap will do good in most audio inputs.

When input impedance (resistor from input to ground) is 10k or lower
you might want to have one 4.7uF. (or parallell two 2.2uF)

But most inputs have 22k or higher input impedance.


One of the best articles on capacitors is:
PICKING CAPACITORS
Walter G. Jung and Richard Marsh
Audio Magazine, February and March, 1980

http://www.capacitors.com/picking_ca...rs/pickcap.htm

[sagarverma]

Originally posted by heater


At the moment I'm not sure I could ever hear the difference between different caps, or no cap. Guess I'll just have to try them.

i have 'heard' many bad electros.but they were really very very very very very..... bad,in construction,even the filler used was so bad 'looking'(got 4700uf/25 volts for Rs 2 each,[$1=Rs 45].

if the cap that u use is 'really' very bad(like Rs 2) then however good u may make the circuit,howsoever great n fancy cables u use etc etc,the system will sound foul.

use good caps(i use philips or panasonic),they r good,but anything fancy like $100 caps etc etc etc etc r a strict no no )for me atleast).

elctros r sturdy slugs.but avoid them for signal path(m/f or polyester r good).for large capacitances,electros r used,for small signals and faster response,look to m/f etc.

wether dc is + or -,cap blocks dc,its in its 'genes',(Xc =1/cf,f=freq,c=capacitance,Xc=capactive reactance,for dc,f=0 so Xc=infinite)
i have used pol electros in signal path without any probs,but generally they r not recommended

moreover using a 10 ohm res at i/p avoids spurious noise from entering the system(albiet at the cost of decrease in i/p level)

A no-name filmcap will beat any 'audiophile' electrolytic as coupling cap.

yup,what is audiophile cap?how does it 'sound'?it must be above $12345678900987654321? whatever


below is pic of my 12db lr xover with cap(pol electro philips at o/p[on extreme right], i/p caps r 'i-don't-know material,but value is 1uf and r non-pol[on extreme left,sitting small and isolated,orange 'dots']),,
the performance is great.

[lineup]

use good caps(i use philips or panasonic), they r good,
but anything fancy like $100 caps etc etc etc etc r a strict no no )for me atleast).

elctros r sturdy slugs.but avoid them for signal path(m/f or polyester r good).
for large capacitances,electros r used,
for small signals and faster response,look to m/f etc.

In my catalog
2.2uF/100V electrolytic cap cost 1/15 of same value polypropylene

The reason for low cost is mainly, as I can think:
1. lower electrical quality
2. much lower cost to make (filmcaps are not very easy to make)

No wonder some try to sell audiophile electrolytics.
Chances to profit, make money, is good
if you can make people believe they get value for money.

[peranders]

Heater remember that there are two things:

DC voltage

AC voltage.

The DC voltage is very little, normally.

The AC voltage ACROSS the cap is also very little. Most of it lands across the load resistor at the intput.