grace with ge transistors as diodes; gain question

[CoolJohnny]

experimenting with LM386 based on runoffgroove's grace/big daddy circuit. i've eliminated the jfet buffer simply because i don't have any J201s handy right now. with the right in/out caps and about 1K across pins 1 and 8 it performs as a nice jangly booster that adds a lot of volume before breaking up.

since i have some pretty leaky AC187s around i set them up as clipping diodes on the output. nice sound indeed, but no matter how much i increase the resistance on the opamp, it doesn't clean up any more. i'm geussing this has something to do with how much the 'diodes' are clipping verses how much voltage the opamp is putting out (the cut down on the overall volume tremendously).

can someone alot smarter than me expain this, and maybe offer some suggestions on how to clean it up if i still want to use the ge 'diodes' at the end and not lose so much headroom?

[MikeH]

Sounds like the Ge transistors a clipping a lot of your signal (makes sense, Ge diodes do too), so I don't think you'll be able to make up for the volume loss without another gain stage in there.  Also, If you want to limit the amount of clipping, you can add a pot between the diodes and the signal path to reduce the clipping.  A 50K would probably work well, set up as a variable resistor, with the transistors connected to ground on the other side.

[CoolJohnny]

i thought about adding a pot there like you suggest. i don't think i'll end up doing it because i had hoped to make this a one knob 'heavy boost' kind of pedal. i've built plenty of multi-knob distortion clippers and was hoping to go simple this time. i even tried putting the trans/diodes in a loop between the gain pins on the opamp, but all this did was darken the signal a little bit. do you don't think actual ge diodes would clip more or less than a pair of leaky transistors? thanks for the input...