Adding a switch w clipping diodes - did I do it right?

[jimbob]

Not sure if i did this right . I thought I did anyhow.  I dug up my old dod-250 clone and decided to add different clippings to it. I found a spdt and used that. On each of the outer lugs of the spdt I connected a pair of diodes, ( 1 pair of 1n914 and the other 1n34a) The middle lug I connected in the hole right before the .001 output cap. I also connectedall the remaining ends of the 4 diodes together and placed that in the hole provided where the label states d2. This switch shows ON OFF ON the side of it. Anyways, when i tested it one end of the switch gives me a louder sound w less distortion the middle also has distortion and is even louder I think and the other end of the switch is a lot less sound but more distortion.

I thought the middle lug was supposed to be off. Is the middle the effects without the clipping at all?
Does what I did make any sense? Sounds good to me.

[Mark Hammer]

That sounds exactly right, Sir James.  The switch can be used to either interrupt the ground connection of the diodes, or else you can leave them all connected to ground and use the switch to make/break their connection to the output cap.  Either ends up being functionally identical. 

The biasing on the Distortion+ is somewhat different than the biasing on a DOD250, despite a lot of other similarities between the two.  On the Distortion+, the Vref is set by a pair of 1M resistors and fed via *another* 1M resistor to the noninverting input.  That's not a helluva lot of biasing current at the input to the 741.  In the case of the 250, the biasing is via a Vref divider made of two 22k resistors and fed to the noninverting input via a 470k resistor.  Big big difference.  Don't know if that is what is responsible but the Diostortion+ will easily go into clipping at even modest gains without any clipping diodes in the circuit.  Not having tried a 250, I don't know if it behaves the same way.

The data sheet for the NJM4558 indicates a typical input bias current of 25na.  I don't have a 741 data sheet handy but the nearest equivalent - an LM1458 - indicates a typical input  bias current of 200na.  My guess is that something is up with those three 1M resistors in the biasing circuit that makes the Dist+ misbehave at modest gains without any diodes, because the difference in specs doesn't seem to explain  the difference in component vales.

Long story short, you should get loud with modest clipping using 1N914s, less volume and more clipping with 1N34s, and loudest but some clipping without diodes at all.

[jimbob]

Very COOL! Thanks Mark.