Who has built a Q&D2 compressor?

[JJGross]

For the life of me, I can't figure out the offboard wiring for this thing.
It looks like there's only 1 wire going to each pot? That doesn't make any sense to me.
I'm sure I'm missing something incredibly stupid here.

How did you wire yours? Even just a clear picture of the guts in yours would be a world of help.

Cheers,
- JJ

[Mark Hammer]

I've made one, and yes several of the various controls consist merely of a variable resistance to ground. You could either wire this up as two wires to the relevant pot, or else solder a pot lug to the pot chassis and let the pot's contact with the pedal chassis take care of grounding.

You probably want to do a thorough search of the archives over at the DIYStompbox forum regarding this little circuit. Jack Orman took it essentially from the SSM application notes for the chip. While there is certainly nothing wrong in that, the chip was really intended for use in mic control strips. As such it expects very low levels from the input signal, and also has an input impedance that would be acceptable for a 600 ohm voice mic but not a 10k guitar output. So, it needs some buffering on the front end, and a volume pot on the output to tame things.

The SSM2166 and the basic design are terrific. make no mistake about that. Bt it needs some additional adaptation to the guitar pedal environment beyond what Jack has done to make it truly great. Much of that is outlined in DIY posts on the forum.

[JJGross]

Thanks Mark, I appreciate the info. From what you say here, I'm thinking maybe something like a LPB2 built into it wouldn't be a bad addition for a stomp. Then a resistor (or trimpot) replacing the single pot in the LPB2 section to give it just enough grunt to drive it, but not enough to overdrive it. Wouldn't an output pot might make it easier to saturate it, kind of like a master volume without the proper support circuitry?

Cheers

[Mark Hammer]

Actually, I think the best strategy is to simply use a single FET unity buffer ahead of the input. There's really so much gain inherent in the 2166 itself that one really needs to find ways to tame it rather than add to it.

Jack's schematic shows a pair of 10k resistors as part of the initial op-amp input stage. It may well be wiser to simply short out pins 5 and 6 and omit the 10k/1uf network to ground. That will turn the built-in op-amp from a 2x gain stage to a unity-gain buffer.

This may or may not offer up the dregree of squish you like/want/need. Keep in mind that the envelope detector is fed by that op-amp stage, so changes in the gain of the stage partially determine the "sensitivity" of the envelope detector and how much the input signal "forces" gain reduction or enhancement. Depending on your needs, you may find that you'd rather have diminished sensitivity but more control over gain further along the circuit, or else more sensitivity up front and less control over gain and stricter control over output level.

It is possible to build reduced versions of the Q&D with fewer controls. Mine is built into a Hammond 1590B and has a compression, level, and noise-gate set control with an toggle to select averaging-cap value. A full-out unit would have a compression-ratio, gain, output level, noise-gate set, and rotation-point control. The noise reduction capabilities of the unit are quite good, so even if you don't feel like using the compression it makes a dandy noise-reduction device when the gate-set control and averaging cap value are set right.

Note that the "gate" is not really a gate but is downward expansion. This does a nicer job than a true gate of preserving the tail of a decaying note before the signal gets attenuated beyond audibility.

Since the averaging cap works a bit like the one in the NE570 compander (one cap determines both attack and decay), you can get some interesting effects with a large value Avg cap and setting of the gate-set control. When the attack time is longer (by virtue of a larger cap value) and the gate-set is set to a slightly higher threshold, you can sometimes get some near bowing effects as the signal is faded in just slow enough to miss the initial attack.

[Artspoke]

Line Level input?

I'm almost done building the Q&D2! The SSM2166P chip is out of production, so I got the SSM2166S (SOIC - surface mount), and a SOIC - DIP adapter from Digikey.

I know this thing is intended for mic-level, but I was wondering if I could mod the circuit in some way to allow for unbalanced line level input. I thought I could simply put a 10k / 100 Ohm voltage divider before the input, but it seems silly to kick the level down, just to pick it back up again with the SSM2166. Is there a more elegant way to utilize this little compressor for line level?

Another question is, will the circuit accept unbalanced line level as is?! In the opening paragraph of its description [http://www.muzique.com/ssm2166.htm], Jack Orman states, "It offers control over the gain, noise gate, and rotation set point, in addition to the compression ratio control offered on the original. This allows more versatility including the ability to use the compressor on line level signals in the studio". Perhaps the gain adjust (R6 off of pin 2) already accomplishes this. ?

Has anyone built any line-level specific compressors based out of one IC, like a THAT Electronics 2252, etc.? I like the small acreage these IC based compressor circuits use - I'm hoping to find aand build a design I like, and then build one into every channel of my two large format recording/mixing consoles. I have a Tascam console that offers an insert on every channel in the form of two RCA jacks, which seems like a fantastic place to insert a mini compressor!

If I get this thing to work, I'll post pictures of my build. I designed the PCB layout myself, and added a 10 segment LED meter.

P.S., I found a picture of a Q&D1 build, which might help you wire your pots:
http://www.radmer.dk/diy/q&d.htm

Jack Orman shows the letters CW next to the four pots in his schematic. The terminal that is the "CW" terminal is the one which, when the knob is turned as far clockwise as it can go, measures 0 Ohms with respect to the wiper terminal. So, turn your pot as far as it goes to the right, put one end of your ohm meter on the wiper terminal (typically the middle terminal), and the other end on one of the other two terminals - one will read the rating of the pot, and the other will read zero. The one that reads zero is the one that goes on the "CW" side of each of the four pots in the schematic.

Hope that helps.