Capacitor Compressor?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

Phrazemaster

Well-known member
Joined
Oct 2, 2006
Messages
2,010
Location
Southern CA
Forgive the probable stupidity of this idea - it's probably been done - but can a compressor be made using capacitors and resistors only? I know that resistors change the time constant of the capacitor, so would that provide a way to smooth out peaks...?

Please no flames - I've been around here awhile but I'm a total noobie when it comes to electronics.

Thx,

Mike
 
No,  a compressor is a nonlinear circuit, and resistors and capacitors are essentially linear components - no combination of them will make enough nonlinearity to perform noticeable compression.

Capacitors "smoothing out" signals is another way to describe a filter,  a circuit that has different gains for signals of different frequencies. However, given one frequency, there's no way to make a filter that treats large signals differently than small signals - that's what a nonlinear circuit could do, not a linear circuit like a filter.
 
There is a class of compressors often used in amateur radio that simply gently clips the signal; its a bit like over driving a guitar amp. You can make a version of this using a couple of back to back diodes and a resistor or two.

Cheers

Ian
 
For a capacitative compressor, you would have to be able to electrically control the value of a capacitor.

The only parts that i know of that can do this are varicap- or tuning diodes that change capacitance slightly as function of an applied voltage.

But the range of these is sub-nanofarad, and quite nonlinear, so probably most suitable at modulated RF frequencies

Maybe something like a RF mic circuit (see discussion in the /mic subforum) could make it possible to AM modulate audio and thus compress? I'm pretty sure that it has not been done before..

Jakob E.
 
gyraf said:
For a capacitative compressor, you would have to be able to electrically control the value of a capacitor.

The only parts that i know of that can do this are varicap- or tuning diodes that change capacitance slightly as function of an applied voltage.

But the range of these is sub-nanofarad, and quite nonlinear, so probably most suitable at modulated RF frequencies

Maybe something like a RF mic circuit (see discussion in the /mic subforum) could make it possible to AM modulate audio and thus compress? I'm pretty sure that it has not been done before..

Jakob E.
Yes, but even that is arguably an active device using voltage control.
=======
Mike, perhaps start by reading about how resistors and capacitors behave. 

You could also search for "Capacitor compressors" , and might find a machine that squashes capacitors. (bad joke, actually you will find motor start capacitor discussions).

As you already observed passive Rs and Cs can create frequency dependent transfer functions, but not level dependent compression, which generally requires active (semi-conductor) circuitry.

JR
 
Thanks everyone.

I do understand cap/resistor basics. I was just thinking how since caps smooth peaks...that’s kind of what a compressor does. But I see it’s quite a bit more complicated than that.

The frequency dependent aspects of caps is what makes it untenable I can see...

Thx
 
Pip said:
Just to add in. Capacitors "smooth the peaks" or filter the ripple by discharging during the troughs not by limiting the peaks
Actually they do, in conjunction with a resistor in a RC low-pass filter.

so in a certain sense they add energy
No they don't.

as a filter they are high pass not low pass.
A filter involves at least two elements, e.g. a cap and a resistor. Connected R-series/C-shunt they constitute a low-pass filter. Connected C-series/R-shunt they constitute a high-pass filter.
Please refrain from trying to explain things you don't master enough.
 
abbey road d enfer said:
Actually they do, in conjunction with a resistor in a RC low-pass filter.
No they don't.
A filter involves at least two elements, e.g. a cap and a resistor. Connected R-series/C-shunt they constitute a low-pass filter. Connected C-series/R-shunt they constitute a high-pass filter.
Please refrain from trying to explain things you don't master enough.

Original Post Removed.

Sorry if I led people astray.

You are right I am wrong. It was a simplification is all and overly so. I will refrain from such contributing in the future.

In an attempt to make it right!

http://www.learningaboutelectronics.com/Articles/High-pass-filter.php

https://www.electronics-tutorials.ws/filter/filter_2.html
 
Here is what I thought of as a control element that you could somehow call 'capacitive' in that the capacitance of the MOS is varied in order to change the signal gain.

It will inherently low pass the control voltage so would only give you slow attack/release times, as you can see it'd have a very limited control range of just over 3 dB and signal would have to very small to avoid distortion.

All in all I'd say this is a poor choice for gain control. Aside from its obvious drawbacks I just find that designing signal paths in a capacitive regime is annoying and doing it intentionally seems masochistic. Sometimes you inherit a capacitive regime (typically from a sensor: mic, piezo etc.) and the first thing you do is typically to buffer/"read" and then complete the signal path in a stiff bridging regime (ie. driving high Z loads from low Z sources).

Thanks for the thread, always good to think about various approaches. I'd love to hear if someone has a trick that would make this approach work well.
 

Attachments

  • MOS_gaincontrol.png
    MOS_gaincontrol.png
    295.5 KB · Views: 21
Back
Top