Discrete compressor. So simple.

[midwayfair]

I think I totally messed up here.

A little.

Take a look at the previous schematic again. I can't "draw" a block diagram, but I can describe it.

The guitar input goes to Q1, which is a voltage amplifier -- you know, a booster. The output is split to two places: the output of the effect and the input of Q3.

The gate of Q1 is held to ground by Q2, which is a MOSFET being used as a variable resistor.

Q3 is a buffer (a current amplifier) -- its input comes from the output of Q1. The signal is then taken from its source and when the wave swings (one way ... I'm not sure if negative or positive) it lowers the resistance between the drain and source of Q2. Q2 is now a variable resistor.

I'm not actually sure how Q4 and the compression pot works.

I scrolled up a little further and I must have missed the previous version of the schematic -- it's closer to what I was describing in that I was suggesting skipping the FETs for BJTs.

---

Yours cures the problem where the variable resistor loads the guitar, but there are two problems:

1) The BJT is not the device I would pick as a variable resistor. I've tried them in the past and the MOSFET just works better in this configuration. I still think a linearized FET selected for best performance is even better, but it would be a little harder to get going.

2) You've neglected to account for how the variable resistor actually works! It's one leg of a divider. How low can you drive a MOSFET? How high will the resistance be at idle? And what's the OTHER LEG of the voltage divider in this situation? Calculate what kind of decibel swing you can get as the effect currently stands!

3) Q1 in your schematic might as well be a BJT (like I was saying before). Unless you're intentionally using the distortion/low headroom of the J201, of course.

[Kipper4]

Great explanation Jon thanks.
I'll look into the improvements as its on the breadboard.
Before I do though i might just try the bjt ver1 from reply 16.

The way I see it the 50k trimmer is biasing the Q3 fet base. turn it one way and the q3 turns off (Not sure if its just off or in saturation) either way. You get no effect at all.
Turn the trimmer all the way the other way. it just distorts and the comp pot does begger all. As if its just a booster. Q1 only.
im getting some compression with the voltages i posted earlier.

Smoke me a kipper. I'll be back for breakfast.

[duck_arse]

the original Q4, kipper's Q6, is wired as a constant current source, exactly as it appears in the orange squeezer. I'd imagine it would bet set and forget in the same manner as the squeezer.

but don't quote me.