Channel switching via reverse-bias diode?

[schnarf]

I had this idea for channel switching. Say you've got your clean and distorted signal paths running in parallel. For each channel, run it through a reverse bias diode such that the diode's breakdown voltage is greater than the maximum signal voltage. After the diode, put a large enough capacitor to filter out a DC bias. Before the diode, attach a relay or switch to connect it to a DC source of enough voltage to ensure that the diode will always be on when the DC bias is connected (if the diode's breakdown voltage is x, and the voltage of the signal is y volts, then make the DC bias x+y volts). Would this be a decent way to do channel switching without terrible popping? The DC bias relay could also be connected to another capacitor to allow the DC bias voltage to rise somewhat slowly when turned on rather than to turn on immediately.

[gez]

If I've understood you correctly, then no, it wouldn't work that well.  Suddenly introducing a DC bias to 'open' the diode switches is going to create a shift in the DC level, which can easily translate as a pop.

Secondly, diodes only allow signal one way.  Granted, if it's biased it will allow a signal through, but it's still closing off slightly during half of the signal's cycle, so you're going to end up with distortion of the signal.  This could even be quite severe if the signal is large.  Then there's crossover distortion - even with bias will you have met each diode's threshold?

Channel switching shouldn't be that difficult to achieve using conventional methods.

[Nasse]

slacktrem

[R.G.]

However, you get a candy bar for a consolation prize in consideration of your original thinking. Diode switching is used all the time in RF circuits where the signals are only millivolts or fractions of a millivolt.

It is not used for audio because of the reasons listed.

However however there are some audio circuits that use a variant of this. It's a diode bridge. The bridge is arranged so the audio is applied to one of the  AC points, and taken out of the other AC point. The + and - points are fed a current symmetrically through large resistors. When current flows, the diodes turn on and the resistance between AC in and out is small. When current stops, the diodes all turn off and no signal flows. The + and - antiphase drive on the resistors balance in the audio path and no pop occurs. You still really should keep the audio under 50mV, but at least it can be made to work. The tremolo on Thomas Vox amplifiers works this way, and works very well. The only problem is the noise caused by running at such a low signal level.  But it complex and touchy, requires frequent readjustment to be perfect.