Increasing sensitivity (reducing minimum threshold) of CA3080 compressor circuit

[aion]

I am putting the finishing touches on the Lab Series L5 preamp in pedal form. The only thing I still have to tweak is the compressor at the end:



Input is coming from the top-left, output is center-bottom, and the toggle switch controls whether is is engaged or not. (Full schematic here for reference.)

It's working just fine - the only issue is that the minimum signal level to trigger it is too high (i.e. testing with the "Level" control all the way down, and the trimmer set at minimum, so zero resistance to the signal that feeds the level-detection portion) since it was originally designed for an amplifier. So, working as designed - but I need to be able to trigger it at significantly lower signal levels.

Looking at the level-detection portion, it is fed through a non-inverting amplifier with a gain of 11, followed by an inverting amplifier with a gain of 1. Is it just a matter of increasing the gain of A112A, the ratio of R188 to R190? Basically providing a hotter signal just to the compressor without affecting the main signal level? Would this have any side effects or things to watch out for, or is there a better way of doing it?

[R.G.]

Looking at the level-detection portion, it is fed through a non-inverting amplifier with a gain of 11, followed by an inverting amplifier with a gain of 1. Is it just a matter of increasing the gain of A112A, the ratio of R188 to R190? Basically providing a hotter signal just to the compressor without affecting the main signal level?

It's sure what I'd try first. You might also get better low level results by varying the values of R193 and R194 to increase the amount of current fed into Iabc by Q102 for smaller signals.

Would this have any side effects or things to watch out for, or is there a better way of doing it?

You've sidestepped some of the issues with the 3080, in that resistor 183 prevents sudden death syndrome by keeping the Iabc under 1ma, and by using ±15V supplies you're probably safe from signal excursions being too big for the circuit.

A possible improvement would be to convert A112 from two simple rectifier drivers with CR103 and CR104 outside the feedback loop into a full wave precision rectifier. As shown, it takes a signal greater than one silicon diode drop to start pulling the gain back. I'd have to do more thinking to see if this is actually the culprit, but it's certainly a possible one.

[PRR]

Why not just put some gain in front of R176??

That's not a good guitar-input; I would assume Lab had a gain stage there.

As noted, you want about 5V into R176. This seems large but is not atypical for shunt limiters.

- - - - Ah, Lab threw-in everything and the kitchen sink. And I don't like some of it (inverting input). But for "5V" at R176 it notes 55mV and 250mV at the input jack. So try a gain of 20 to 90 between guitar and R176.

Critiquing your plan--

The A112A inverter should stay UNITY gain.

A112B drives CR103 on one half-cycle, A112A drives CR104 on the other half-cycle. If the unity-inverter is not nearly unity, you have a half-wave rectifier, increased ripple, and lower stability.

The A112B gain-of-11 stage can be jazzed-up arbitrarily high. (Gain over 100 will start to lose the top of the guitar band, though this may not be critical.)

However the Q101 A111 sub-system has a rather tight "optimum level". The '3080 either hisses or overloads, and there isn't a huge range between. The inverted connection does flip the problems and slightly extends range, but I assume Lab picked a good operating point and don't think you should randomly change that.

The design opertating point seems to be 1.2V Peak at Compressor pot wiper.

What DC do you have at Q101 Source? The DC bias to Q101 depends on fairly offset in A111, and I wonder if Lab had to select parts to get there.

[PRR]

The non-precision (0.6V threshold) rectifier is not a problem, may improve limiting action. We don't want to know about too-small inputs, only large inputs.

There is a further 0.6V threshold to turn-on Q102. This plan needs LARGE signals. Assuming ample supply voltage, there's no real down-side to this (it simplifies things a lot). If the output is too big for your needs, pot it down.