AGC or Gated effect for reverb circuit

[goldstache]

Hey everyone!  I have a pretty nice spring reverb going on the breadboard and am wanting to set up a LED/LDR to control the wet signals mix.  It's pretty standard fare.  Dry preamp, buffered springs driver via fet into LM386, reverb recovery stage and EQ, virtual earth mixer for wet/dry, output. 

Ideally I want to manipulate the length of my wet sound mix, controlled by the input amplitude into front end. I've tried lots of different setups, and nothing is giving me much control.  I have some vactrols, ldr's of various values, etc..  But I can't quite dial it in. 

So I guess I'm looking for a nice consistent LED driver (detector) and.........

A reasonable way to implement an LDR (recovery stage feedback, series resistance, variable divider, etc.) to have some control of the wet mix (verb only)

Kind of like a gate for the wet return only. 

I don't have a schematic written yet, and I know it would help, but if you have any thoughts with regards to topology or implementation please let it fly. 

Thanks.   

[Mark Hammer]

In the early days of flangers and choruses, one of the strategies employed to keep delay-chip and clock noise out of the final audio path, was to use a FET-based gate on the delay-path.  Look up the A/DA Flanger and Boss CE-1 chorus and you will see a FET to ground at the end of the delay path, just before the wet and dry are mixed.  When playing, the FET is turned "off" so that the FET forms a high-resistance path to ground and minimal delay signal is bled off.  Stop playing, and the FET is turned "on" to drop the resistance very low and bleed off as much of that signal (which would presumably be only hiss and clock noise at that point) to ground.

That seems to be a suitable solution for what you want.

[goldstache]

Mark, thanks for your help.  I will certainly check it out. I tried to go the compander route via 570, as noise was an issue on the verb return, but couldn't get it to work right.  I'll let you know how it goes.

[goldstache]

In this schematic:
http://experimentalistsanonymous.com/diy/Schematics/Chorus/BOSS%20CE-1.jpg

Is the fet in question Q12?
looks like IC-3 (preceding stage may be of importance?) Not sure what's happening there.

Guess I better read up on Fet's as gate/switches.   

[Mark Hammer]

Yes, Q12 is the one.  Normally, the 2k2 and 100k resistor pair form a voltage divider that attenuates the delay signal just a tad.  Q12's drain-source path is in parallel with the 100k, but is such a high off resistance that the combined parallel value of Q12 and 100k is basically 100k.  When Q12 is turned on, its drain-source resistance drops low (presumably lower than 2k2) and in parallel with 100k bleeds off most of the delay path signal.

The circuit built around IC-3 is what senses the input signal level and determines whether Q12 is on or off.

[Kipper4]

I'll wait for the schematic and go from there.

might be something here

http://www.diystompboxes.com/smfforum/index.php?topic=117442.msg1090568#msg1090568

http://www.diystompboxes.com/smfforum/index.php?topic=117301.msg1090402#msg1090402

Rich

[goldstache]

Mark, thanks for confirming and the description. 

So, Input stage in low mode is a gain of around 47.  Then onward to the IC3 where it gets amplified a bunch, reversed biased for a negative supply that varies in accordance with the amplitude of the input to that stage.  I don't totally understand whats going on with the network after IC3.  Especially the other leg of R66 (22k), it seems to go to one of the power supply drops.  Does that arrow mean the -14 supply or the +14 supply.  I'm gonna go with -14, due to the fact that theres a reversed biased diode there and the schemo says -3V. 

I'm gonna mock it up.  Do you think I can get away with standard parts tlo72/1n4148/PF5102?  I haven't tried to cross anything yet.  IC3, thanks to handy voltage chart, appears to be going square, and I think it says 25v peak to peak.  That's not hitting the rails, but its getting there.  Should I be looking for an op amp that can swing to rails? 

Thanks so much for the input, if I can get this part happening or need more ideas, I'll draft a schematic to make it easy on everyone. 

[Mark Hammer]

Maybe take a look at how they do it in the A/DA flanger.  In that application, the FET drain-source path is in series with the delay signal., rather than going to ground.  The envelope detector is simpler since the FET is turned on when there is signal, instead of off in the case of the CE-1.

[duck_arse]

your IC3 is squaring the signal and feeding a rectifier/smoother to produce a negative going envelope signal to control the fet gate.

you need to match the power supply arrows for type and colour, it seems positive supply connections point up, and the opposite, down.

[goldstache]

Kipper, Duck, thanks for the help.  That's what I was thinking regarding IC3.  And the power designations makes sense, thanks.

Kipper, I ended up going down the LM13700 based VCA type thingy rabbit hole.  I used a basic app note with a pot to vary VCC into a buffer, then onto 13700 control pin.  Definitely gates the noise in non playing conditions when CV is low.  Not really what I'm after, but I'm just following any whim in hopes to learn. 

Wanting to control the VCA with my input signal.  So I tried........

Input >  Buffer  >  op amp stage (rectified for CV)  >  control pin of 13700 =  NO DICE!  Sounded like a fart. 

When checking my CV stage, I'm getting DC voltages in the right range.  But it's blowing up the 13700.  Trying to wrap my head around that. 

I've tried many other variations of tracking the input and converting to DC, with resistor in line for current drive, to get the control pin to jive.  But I get all sorts of audio interaction and namely distortion. 
Why when I just send a variable buffered DC source to control pin, it works fine??

The A/DA, and CE-1 are a bit over my head with regards to the concept. Though, I would love to understand this further, as it's nice to be able to gate signals, amplitude based, or even in general really.  I find it funny that I've spent more time getting Fets to turn on, that you'd think I'd be a doctor at getting them to turn off

Thanks for all suggestions and help understanding.  If anyone has any input please don't hesitate to share.

[Kipper4]

You're welcome anything to help.

"When checking my CV stage, I'm getting DC voltages in the right range.  But it's blowing up the 13700.  Trying to wrap my head around that.  "

Maybe too much current.
Check the data sheet.

Some reading.

http://www.geofex.com/Article_Folders/VCA%20Applications.pdf

http://www.nutsvolts.com/uploads/magazine_downloads/11/April%202003%20Ray%20Marston%20-%20Understanding%20and%20Using%20OTA%20Op-Amps.pdf

this too

http://damienclarke.me/effects-pedals/posts/building-a-vactrol-vca

Sorry about that.....My eyes...