Simple FET compressor

Hi all,

I have a basic FET compressor breadboarded and functional. I though I might run it by folks here before committing it to solder. I am happy with how it sounds, but it's cobbled together from reading a few different schematics, and I wouldn't be surprised if there's something wrong, or at least not ideal, here. Notice anything odd? Anything you might do differently?

I've been hanging around here for a while, reading and learning. It feels good to post something.

If the 470k and .01uF cap network are to reduce distortion that needs to be a constant ratio (-6dB below drain voltage IIRC). Your threshold pot that is varying the gain of that buffer will also vary the contribution of the 470k degrading the distortion reduction. Since you are operating the buffer after the JFET shunt at fixed gain you could divide down that output voltage  (that won't change with threshold) for your distortion reduction tweak.

JR

Ah, I see that. Thanks! I'll try attaching that RC network after the output amp with a divider and see what happens.

It has also occurred to me that I could run the sidechain buffer at fixed gain and control the threshold with a 100k pot to ground before the buffer (taking the place of that 100k resistor before the output amp).

rp said:

Ah, I see that. Thanks! I'll try attaching that RC network after the output amp with a divider and see what happens.

It has also occurred to me that I could run the sidechain buffer at fixed gain and control the threshold with a 100k pot to ground before the buffer (taking the place of that 100k resistor before the output amp).

Click to expand...

As long as you manage to get the side-chain gain structure correct, why not? I don't think it would change significantly the operation. You need variable gain here, in order to vary threshold. There are several ways of making gain variable; all are correct. Your choice will be guided by your abilities, your economics, what you have in your drawers...

Back in the '70s I messed around with linearizing the control law for a JFET by using a dual n ch JFET, with both sources grounded. I wrapped an op amp feedback loop around one half to make it a voltage controlled resistance, the other half tracked reasonably well, but with a linear not log law, so less convenient for dynamics processing. 
Around the same time I applied the same technique to center tapped cds cells. to control the one half by putting the other half in a feedback loop.

This may be too much work just to make a JFET or cds cell deliver a repeatable transfer function. (you could always just buy a VCA).  ;D

JR 

Thanks to both of you again for the feedback. I ended up scrapping the threshold knob. Both additional methods worked, but I decided that adjusting the threshold with the input/output pots was fine for my needs, and it makes for one less knob on what will be a tight front panel. I also changed various R values, for someone who sees this and might want to fiddle with it.

I notice there's a thread in the brewery on this topic currently. I'll add to it here and say that FET compression is a pretty fun and useful sound to have at your fingertips. Definitely not the cleanest or most precise. In appreciation of this, I added a resistor in line with the bias trimmer that one can switch in or out to bump up the FET bias, which lends itself to getting that "drums are exploding" sound. From reading, it seems like the 1176 does this as part of the sidechain, which is what you abuse when you go all buttons in.  I don't completely understand that circuit, though (hence making a simpler one).

JR, sounds like an interesting thought experiment with the dual JFET.