problem with parralel processing

[Wounded Paw]

So I'm working on a new design for a two channel pedal with some differences.  I have an overdrive and a fuzz channel and they get processed separately, as in after the pre-amp the two channels split and then recombine right before the output stage.  The problem is I want to be able to switch the fuzz channel on and off while leaving the overdrive channel unaffected.  When I hit the switch to add in the fuzz channel the volume of the overdrive channel drops a bit.  This happens whether I combine them with a blend knob or separate volume controls and it happens whether I'm using a jfet or LPB style output stage.   I know it's because the overall signal level increases when the fuzz channel is added in but is there a way around this?  Would using a op-amp output stage solve this problem?  Is it just the fact that the signal is getting too hot, if I drastically reduced the output level of the two channels before recombining them would it help? 

[aron]

It's worth a try to reduce the level and see what happens.

[ianmgull]

Did you use a separate buffer in front of each effect? How did you go about splitting/recombining the signal?

[Ben N]

I know it's because the overall signal level increases when the fuzz channel is added in but is there a way around this? 

I'm not sure, but I think you may be wrong on this. You haven't described in detail what precedes the two circuits, so we don't know how high the output impedance is. If it is not, very high (e.g. if there is some sort of passive tone stack there, or maybe even if the preamp is a jfet), you may be losing volume from the effective decrease in input impedance when the two effects are on and their inputs are paralleled (especially if the fuzz is a vintage type). If that is the case, the solution is either a very low impedance buffer, or, ideally,  a dual buffer, such as Ian suggested. Suitable candidates may be found at AMZ. However, this may affect the tone of either or both distortions.

I'm guessing that Aron's suggestion is based on power supply sag--that could be an issue if the whole shebang is powered by a 9v battery, but that is less likely with a wallwart power supply.

Also, I'm not clear whether there is an overall drop when you have both distortions on--if so, you should check whether they are in phase with one another.

HTH