problems designing a compressor/feedforward

[loss1234]

i aam trying to bang out a simple Vactrol based drum compressor.

so far so good when using feedback mode (it sounds great at high ratios)

basically here is how i have it setup on the breadboard


signal goes into an opamp (GAIN STAGE/VCA) with a VTLC2 in the feedback loop.
there is a ratio control that determines how MUCH of the CV goes into the LED.

I have on the board an opamp FWR.

it is a design I have used before and I know it works.

In fact, when i test all of this with a simple LED just to make sure it works, it definitely lights up.

BUT it only works when i take the OUTPUT of the GAIN STAGE /VCA) and stick THAT into the input of the FWR.

I believe that is called FEEDBACK MODE.

however, i have read that feedforward works better for drums.

SO, i have a switch setup which allows me to go from either the OUPUT of the GAIN STAGE/VCA or the the INPUT SIGNAL which goes through a buffer.

NO DICE on feedforward mode.

SO my question is this

1. Do i need to add gain to the feed forward signal before it hits the FWR? and if so, by how much?


also...am i correct in thinking i add THRESHOLD by adding variable DC offset to my FWR out BEFORE the Ratio Control?

thanks so much

[loss1234]

OK

seem to have solved my own problem

WOW

one thing i must say

Since most guitar related designs seem to use feedback, I was blown away by what a difference adding a feed forward stage made!

for those interested in compressing not just gtr but also drums or vocals with their stompbox, please try this.

all i ended up doing was taking the INPUT signal and splitting it

side A went into the OPAMP with a vactrol in its feedback loop

SIDE B went into a TL072 with a 100k pot in its feedback loop (and a 10k resistor going into the input)

well...With little gain, the compressor wont work in feed forward mode

BUT as you start adding gain (which is only going into your FWR) wow...the drums start getting that CLASSIC pumping sound where the SNARE turns into a soft POP

I must say I am falling in love with compressor DIY!! BUT not the kind where you need weird parts or  expensive  Transformers...I knew there had to be a way to do it on the cheap!!

BTW...i have yet to add any Threshold or attack/decay

But that seems as easy as adding an A/R gen after the FWR'ed signal/


thanks guys

[Mark Hammer]

Though not strictly the case, feedforward tends to be more true of what some folks mean by "limiting", while feedback is more true of "compression".

With feedforward the signal your rectifier is responding to is just the input, not a modified version of the input.  In contrast, feedback mode drives the rectifier with a signal that has already been altered.  The result is that compression/feedback tends to result in a more constant level, while feedforward tends to simply keep the peaks in check and do much less to the soft parts.  This is the principal reason why someone might prefer compression on guitar but limiting on drums.  The caveat might be the case of cymbals, where compression with long decay can produce some interesting effects.

[loss1234]

thanks

you know it's interesting, I have been a studio guy for years, In fact that is what I have done for a living for many years.
But you find out a lot of things about compression very quickly when you try to actually build one!

Granted I have always known the difference between an Opto compressor, a limiter, high ratios and low, etc BUT when you actually have the damn thing on the breadboard and can tweak every value, it opens up a lot of new possibilites and makes terms like feedback and feedforward actually mean something on a visceral level.

anyway...thanks so much for all the help.

i suppose the next thing to do is try the same idea with a VCA instead of an opto to see how that sounds, and then of course, try to add some KNEE, attack, and release to the CV itself.

[Boogdish]

If you have time for it, please post a schematic, visual learners like myself get a lot more from a drawing than sentences describing a circuit.

[frank_p]

If you have time for it, please post a schematic, visual learners like myself get a lot more from a drawing than sentences describing a circuit.

Sometimes I wish there were a blackboard option on this forum.

[arawn]

I work in the cable repair industry and had been wondering if an automatic gain control could be adapted to compressor duty??

[Mark Hammer]

AGC = compression in a great many instances, though when it gets called AGC it tends to be a bare-bones curcuit with less quality. AGC is what you find on cheap cassette recorders and CB mics.

[arawn]

I see, I wasn't sure and it seemed an idea. Besides I was looking at the datasheet for the lm358 and there is an agc design in there. Just rung a bell for me

[Mark Hammer]

AGC/ALC circuits, in my experience anyway, are explicitly intended to keep volume constant at any cost, and are principally directed at processing voice.  Speech comes in little bursts rather than passages, and the speaker can also regulate and maintain volume if they wish, so the target of such circuits is to keep other sensitive circuits/devices from getting overloaded (e.g., saturating tape). 

As such the little things intended to allow a sustained note to be held longer, or to allow a strummed chord to fade out gracefully, tend to be missing.  That's not a reason why such a circuit could not be adapted to function as a compressor, but as built directly from a published schematic there is a high likelihood you will find too much envelope ripple for your tastes.  Again, that could be easily ..... fixed (had to stop myself from saying "rectified"  ).