Ducker pedal made from Ross/dyna comp circuit?

[ElsStevo]

As far as I know there is no analog (does not contain an AD-DA conv) pedal that can duck effects. Most rack compressors have a key input where you can trigger the compression of the main effect from another effect. The idea I have is to compress a delay pedals signal to make room for the guitar signal until the guitar signal has stopped and lets the delay ring out.

I want to use a Ross/dyna ciruit as a starting point.

Unit would have IN, OUT,SEND,RETURN to make it practical for guitar pedals.

I expect the Input buffers' output would be split to the SEND output and sent to a copy of the compressors "rectifyer-filter" section which controls the sensitivity feedback for the OTA.

The RETURN from the other pedal (delay) would be sent to the input of a replica of the entire compressor circuit but with the sensitivity line feeding the OTA removed and replaced with the one above.

Then a simple blender circuit to blend the original signal that provided the sensitivity and the ducked RETURN output.

There is most certainly better ways to do this ie: there is no need to copy all of the compressor output stage for the ducked circuit. I have a ross/dyna pcb that I'm intending on using for this so I figured I'd try it out.

Can anyone tell me if this is a good idea or if I'm going about it in completely the wrong way.

Cheers,
Steve

[bwanasonic]

As far as I know there is no analog (does not contain an AD-DA conv) pedal that can duck effects

Hmmm, I'm pretty sure ducking existed well before digital effects. I'm sure some studio analog compressors had the capacity for a sidechain input ( see PS below). Not sure about the logistics of say, an all analog *ducked echo* effect patch, but I would think it could be done.

Kerry M

Ohh, wait I realized you were specifically saying "no pedals" that could do this. Well anyway, the RNC link is still useful:

PS.  http://www.fmraudio.com/FAQ.htm  No AD/DA converters here...

[Processaurus]

Its a killer idea.  You could even hook it up to other stuff, like a drum machine, to get a kind of poor man's Seek Trem, or a singers mic, or a mic in a kick drum, or the TV...

[Mark Hammer]

The Line 6 Echo Park, admittedly a digital pedal, does exactly what you describe, and does a nice job of it at a decent price too.

Since the request is for something analog, though, PAiA has a schematic posted at their site for their ducker project that appeared in a magazine article about 12 years ago.

Another route to consider is that used by both the Boss CE-1 and A/DA Flanger for their noise reduction.  Rather than using companding or excessive lowpass filtering to keep clock noise out of the audio output, both used an envelope follower sidechain to control a FET after the BBD chip.  In their case, the envelope follower was not unlike that of some envelope filters in their "downward drive" position, where the control element (a FET in this case) is biased to its lowest resistance to start with and then turned "off" (high resistance) by audio input.  In both the CE-1 and A/DA unit, the FET is used as the ground leg of a resistive attenuator.  So what this means is that if you didn't play anything, the FET went low resistance, bleeding the BBD output (and any clock noise it carried) to ground.  The moment you started to play, the envelope follower would make the FET go high resistance, eliminating the attenuation effect of the wet signal.

You could easily adapt the strategy for your desired application by effectively reversing the action of the circuit, whether by envelope inversion or use of the opposite-channel type of JFET.  So, as long as you're NOT playing, the wet signal goes unattenuated, and once you start to play, the signal gets momentarily bled to ground.  Since the dry signal does not pass through/by that FET, it goes unaffected.  Naturally, you'd need to tinker with the time constants of the envelope follower so that it brings up and cuts out the wet signal appropriately.

Make sense?

[ElsStevo]

Yeah that may be a tad out of my depth at this point. Thanks for the awesome responses guys.
I wasn't actually trying the make a better wheel or go into production or anything like that. What I'm wondering is that since the dyna/ross comp circuit is pretty well known and pcbs for it can be found easily as well as many mods, is there a simple way to make it into a basic ducker with the majority of the circuit in tact.

Basically its just another toy for us less hard core DIYers.

Alternatively, is there a better base circuit design or PCB that could be used as a staring point?

..That being said, Damn you guys know your stuff!

[Mark Hammer]

Did some further thinking about it, and yes, it can be done, though it needs a few more parts.

Duckers work by turning down gain for one signal in response to a second signal.  Compressors work by turning down the gain of a signal in response to itself.  If you can separate the control signal from the one being modulated, you're in business, using a basic Dynacomp circuit.

To start, load up the "Compreross" document from Tonepad (once again, thank you Francisco) so we have a common reference.

You will note that the output of the CA3080 goes to Q2.  Q2 uses equal value emitter and collector resistors (10k) allowing it to function as a phase splitter  (output at E is equal in amplitude but opposite phase to output at C).  The emitter output (arrow side) of Q2 goes to two places/nodes.  One is a complementary envelope follower formed around Q3, including the 1M resistor cap and diode.  I describe this as "complementary" because even though it LOOKS the same as what comes out of the collector of Q2, it is working with an opposite phase signal.  Put the collector outputs of Q3 and Q4 together and you have a full-wave rectifier.  The pooled Q3/Q4 output drives Q5, whose output feeds current to the Iabc line (pin 5) of the CA3080.

The other "place" the Q2 emitter output goes to is to the .05uf cap, 10k resistor, and 50k output level pot.

Let's ask ourselves, do the .01uf caps normally tied to the collector and emitter of Q2 have to be tied to Q2?  The answer is no.  Suppose you stuck another jack in the chassis and fed an external signal to a simple transistor phase-splitter stage just like the one normally there, and THAT phase splitter now had the two .01uf caps connected to it in the same way.  You would now have a CA3080 based VCA stage, and a completely separate sidechain (just like in "real" duckers).  IN the absence of any input to the external jack, the gain would remain high for the audio input.  Once anything was played into the external jack, though, the gain would be instantly reduced by the amount set via the 500k compression pot.  Voila, instant ducking.  There are a few details to attend to, with respect to the transistor stage for the sidechain, but in principal that ought to work.

If you want to build completely from scratch, in addition to the PAiA schematic/project, Rolls has a manual pdf for their DU30 ducker that contains the schematic.

[sean k]

Thats great Mark,
                         so in effect,excuse the pun,the output from pin 6 of the CA3080 goes to the emitter of Q2 and an extra input goes to the base of Q2 given the signal from source had been split into two in phase signals with one going to a delay,etc, and into the input of the compressor,now a ducker,while the other goes to the extra input?
 Would the output at pin 6 need the bias voltage applied to it and/or a cap before joining the emitter of Q2 and would the extra input to the base of Q2 require any amplification an/or inversion?

[Mark Hammer]

Thats great Mark,
                         so in effect,excuse the pun,the output from pin 6 of the CA3080 goes to the emitter of Q2 and an extra input goes to the base of Q2 given the signal from source had been split into two in phase signals with one going to a delay,etc, and into the input of the compressor,now a ducker,while the other goes to the extra input?
 Would the output at pin 6 need the bias voltage applied to it and/or a cap before joining the emitter of Q2 and would the extra input to the base of Q2 require any amplification an/or inversion?

Tell me it's great AFTER it is confirmed to work! :lol:

The CA3080 needs a transistor output buffer between it and the volume pot. as far as I know.  In the case of the Dynacomp circuit, the SAME transistor serves as both phase splitter and buffer, which is why the emitter output goes to BOTH the output circuit and one half of the rectifier circuit.  I'm naive in the ways of transistors, but I assume pin 6 goes to the base of Q2 and not the emitter.  The emitter, in turn is what goes to the output pot.

What I *attempted* to describe in the previous post is that a ducker can be made by installation of a duplicate of Q2, whose base is taking a signal from an external jack.  The way the rectifier circuit is designed, it keeps charging up the 10uf cap, which keeps the gain of the 3080 high when it is fully charged.  The envelope essentially counteracts that, and the "recovery/attack" mod I've mentioned in past (which is not my idea, only what is in evidence on the Boss CS-2, the Nobels compressor, and others) simply adjusts the recovery/recharging of the cap (note that the stock resistance of 150k makes that cap charge up slowly).

If you want to go crazy cuckoo nuts, you can even make this into a de-esser by preceding the substitute Q2 with an input gain stage for the external signal, and a tunable bandpass filter.  This will result in gain reduction for the band you select but no gain alteration for the rest of the spectrum.  De-essers are essentially compressors that only affect harsh sibilants in the lower treble range.

[Torchy]

http://www.paia.com/duckerda.htm



Mods

[Mark Hammer]

It's a greatlittle circuit but you will note that it is essentially designed for line level signal.  The audio input being ducked by the external signal simply passes through a FET-based attenuator circuit with NO gain applied whatsoever.

That doesn't mean that it could NOT be modded, but Steve's original request was really for something that could be directly applied to guitar signals within the context of a pedal-board.