input decoupling caps

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

mkoijn

Active member
Joined
Jan 11, 2008
Messages
37
Hi, I notice that on nearly all transformerless preamp schematics, the input caps are 47uF. As I've often read that elctrolytics are not desirable on the input path, what other possibilities exist for the input caps?
 
If foil caps become huge then they get sensitive to EMR. Not so desirable in a preamp. I had to tinker a lot in my supergreen preamps to get rid of noise though the input caps were only 10uF
 
The good old SSL9k has just 2 x sets of 2u2 poly caps, can we follow this
with other designs and what makes that setup "work" in the 9k ??
Is it the following LM394 setup gain stage ?

MM.
 
[quote author="mkoijn"]Hi, I notice that on nearly all transformerless preamp schematics, the input caps are 47uF. As I've often read that elctrolytics are not desirable on the input path, what other possibilities exist for the input caps?[/quote]

There has been some discussion here about DC coupling the input stage and shifting the cap coupling to post that gain stage where it can be managed with a better film cap.

Otherwise for good noise and LF CMR you are stuck with a 22-47uf typical input cap value that is only practical from electrolytic.

JR
 
Thanks, I may look into the dc input servo ides. Do electrolytics really degrade the signal path so much?
 
mkoijn said:
Thanks, I may look into the dc input servo ides.

Do electrolytics really degrade the signal path so much?
You could spend ages discussing that, and it has been done, see the threads here and everywhere on the net.
Alternative is to read a bit, for instance this series:

C. Bateman, Capacitor Sounds, Electronics World, July 2002, pp. 12-18
C. Bateman, Capacitor Sounds 2, Electronics World, September 2002, pp. 16-22
C. Bateman, Capacitor Sounds 3, Electronics World, October 2002, pp. 12-18
C. Bateman, Capacitor Sounds 4, Electronics World, November 2002, pp. 40-50
C. Bateman, Capacitor Sounds 5, Electronics World, December 2002, pp. 44-51
C. Bateman, Capacitor Sounds 6, Electronics World, January 2003, pp. 44-51

To be found here, among other stuff:
http://www.groupdiy.com/index.php?topic=15803


Note that what comes out of the cap at the right is as as much the result of the very cap itself as well as the music you put into it at the left plate.
Better not overspend time & effort on the cap but keep making music as well.
 
[quote author="clintrubber"]BTW, sorry for nitpicking, but let's call it input coupling caps.
Decoupling is to smooth supplies.
[/quote]

Hey Peter. When do you decide which is which? I usually always call it decoupling as we are blocking DC components and hence they have been decoupled from the signal. I just call power supply filtering "filtering". I've never been that sure.
 
[quote author="rodabod"][quote author="clintrubber"]BTW, sorry for nitpicking, but let's call it input coupling caps.
Decoupling is to smooth supplies.
[/quote]

Hey Peter. When do you decide which is which? I usually always call it decoupling as we are blocking DC components and hence they have been decoupled from the signal. I just call power supply filtering "filtering". I've never been that sure.[/quote]
That's how I always thought it without much further thinking, but I may not even be correct there, so let's check:

http://en.wikipedia.org/wiki/Decoupling_capacitor

http://en.wikipedia.org/wiki/Coupling_capacitor

Bye,

Peter
 
[quote author="jensenmann"]If foil caps become huge then they get sensitive to EMR. Not so desirable in a preamp. I had to tinker a lot in my supergreen preamps to get rid of noise though the input caps were only 10uF[/quote]

This came up at Bruno Putzeys' forum recently: http://recforums.prosoundweb.com/index.php/t/21314/0/

Check the Sony photo.

Justin
 
[quote author="flintan"]Maybe something like this?

Available up to 330µF, but big and expensive of course.. :cool:[/quote]

I have these in my power amps - nice!


Justin
 
The "badness" of electrolytic caps is somewhat undeserved, for high impedance applications and can be made better or worse by how used.

The primary consideration for how much non-ideal behavior you will experience with electrolytic dielectric is a function of (AC) terminal voltage and series current.

The terminal voltage is reduced by using larger values. The series current will be impacted by where the nominal 1.5-2k mic termination is located. If the 2k termination is mostly on the mic side of the coupling caps, they only need to drive the much larger value resistors needed to establish a dc path for following high impedance electronics. So the same exact capacitors can act differently in a similar preamp depending on where the termination is located.

Note: in a transformer input you can't change where the termination occurs, but you also don't need capacitor isolation of phantom DC.

Note2: In passive speaker crossovers you have both high current and high AC terminal voltage, worst of both worlds for electrolytic.

JR
 
Agree with the wikis on decoupling/coupling---always the way I understood the terms, with in fact the greater emphasis for "decoupling" on reducing interaction among subcircuits via the power rails.

Some mention of a series R or L-R between the power rail and the shunt cap would be helpful---but then I guess the wiki would be more properly titled "decoupling network".

As far as the noise pickup issue, besides the electric field interception, I've had situations where the simple fact of the larger loop area of the coupling capacitor and nearby plane or trace of signal return, or the other capacitor in a balanced arrangement, did me in. This was on the output (!) of a crossover system where the caps used were large copper foil/paper/oil types, and the charging current spikes were modulating the stray field of a nearby big toroid.

It seemed, in principle at least, that a compensating pickup coil could be arranged so as to cancel the fields. Better was just to move the transformer away, but this meant a bigger box or two-chassis system.

The discussion in the other forum about magnetic shielding is correct but somewhat incomplete, as it doesn't emphasize the importance of having a complete magnetic circuit for the shield. Just substituting ferromagnetic material in place of non-, where the latter was effective for electric fields, may well make the mag field pickup problem worse.

Sometimes you can contrive a given shape and size of such material to shape the field and get improvement though, as long as the interfering source is fixed in position. This won't help for magnetic radiation from other units in proximity though.
 
Some of the folks I work with whom regard themselves as learned engineers mix the two terms. This tends to lead to a lot of head scratching. I tend to use the terms "DC blocking" or "AC coupling" for coupling caps to give it another level of clarity. I leave DEcoupling as it is. This seems to be working pretty well keeping the worker bees straight.
 
Back
Top