clipping diode arrangement

[jokn]

Hey guys, I need someone smarter than me to help me out a bit: I'm working on a two channel recording preamp, and I'm trying to figure out a way to incorporate adjustable diode clipping (heresy to some, I know ) for the dirty channel.

Does the attached schematic snippet make any sense? The intention is to have a switch to select the clipping diodes in or out of circuit, and when they're in circuit, to have the two signals , clipped and unclipped, mixed via a pot and fed to the next gain stage through a Peavey-esque noise gate (which will add its own piquant flavor) Thanks!

[shooter]

not sure it'll float, having a hard time getting much signal through 1meg + 470k/vr2 voltage divider
1a will kill the signal, guessing that's it's purpose
 the diodes an pot to ground should be just fine to square up your signal, then just tap off the signal on the left of your input cap and give it it's own cap and send it downstream to meet up with the square waves

there's how the SS guys do it in pedals, (D3 &4)

[d95err]

As drawn, it will not work at all, so I assume the connection that shorts the diodes to ground is a mistake and should not be there. Even without that, I don't think it will work well.

The Peavey style noise-gate requires a very high signal level to work, otherwise you get wierd crossover (non-musical) distortion. Since the clipped signal has a relatively low level, the noise-gate diodes will barely let it throught at all.

Also, the diodes will clip the signal extremely hard. The signal level before clipping can be about 100V peak and it will be clipped down to about 1.4V. I would consider using zener diode clipping instead (like e.g. "Jose"/Friedman styld circuits) and use an additional series resistance to soften the clipping up a bit. Something like 5-20V zeners are typically used for this.

For variable clipping, you could simply put a pot wired as a variable resistance between the diodes and ground. Low resistance will be harder clipping. Higher resistance will be softer clipping. At max resistance, the diodes will effectively be out of the circuit. I'd try a 200-250k pot.

[jokn]

Thank you for the input! Considering your comments, I redrew a simplified version.

I've left out the noise gate, as I'll move that downstream closer to the TS where the signal voltage is higher, and I've removed the voltage divider and coupling cap.

If this were placed early in the signal path, with appropriate attenuation upstream (input around 2-4 Vp-p) would this work with the usual diodes (1N4007, 1N4148, red LED, etc.) ?. I've kept the switch as it would be nice to have the clipping diodes switchable on/off (for live use), even though the pot alone would accomplish the same, as d95err pointed out.

[shooter]

I'd be inclined to build a little 2X2 tack in board, then decide if you like it, what you'd change, etc
here's a quickie that might work to blend the clean/clipped.
and ya, you do want it closer to the left side of your schematic, maybe after V1a but before TS 

[pdf64]

I think that a series resistor is needed before the clipper, otherwise it will clip both signal paths (depending of the frequency, C and R values, source impedance etc) 
Maybe the 'C's should be resistors rather than caps?

[jokn]

How 'bout having the pot on the left, and mixing resistors on the right? Just a sec, I'll sketch it up..

[jokn]

Like this, perhaps?

[shooter]

when they're in circuit, to have the two signals , clipped and unclipped,

I think a breadboard and experimenting is your next step, you don't need to even cut up your amp if have a handful of transistors.  V1 typically just amplifies, stays class A.

this is my tinker board, does both analog and digital 

[jokn]

Nice setup, Shooter! I’ve got an amp I can use as a test bed, so I’ll cobble together a version of this, probably on veroboard, and get to experimenting. I’ll report back with the results!

[d95err]

Like this, perhaps?

Don’t think it will work as intended. This solution still adds series resistance between the diodes and ground as the oit is turned, which affects the clipping threshold. I think it would have more or less the same effect as the variable series resistance. I.e. It will be a variable clipping threshold rather than a mix between clipped and unclipped signals.