Hi,

I found a noise gate circuit on page 66 of this pdf, also attached as screenshot. How does Q1 switch? Is that some sort of avalanche configuration?

 

It's obviously drawn wrongly, but I'm not sure whether the transistor has its collector and emitter reversed, or whether the LED is backwards and connected to the wrong supply.
I'd go with the former, as it implies the smaller number of drafting errors.

 

I think the 2N3904 has the E and C reversed.

 

Hi Ian, Ron, thanks.

Does the attached driver look ok to you? Are the values of R1, R2 and R3 ok?

 

Hi Ian, Ron, thanks.

Does the attached driver look ok to you? Are the values of R1, R2 and R3 ok?

No - the inputs to the 4066s are back to front. Q2 output should go the the shunt switch and Q3 output to the series switch.
Depending on your application, I wouldn't recommend the 4066 as the switching will be too abrupt. The LDR (although illegal for a commercial product) would probably be better. Commercial analogue noise gates use a voltage controlled amplifier.
If you want it for radio transmission, it will probably be fine.

The values of R1,R2 and R3 are OK.

 

Hi Ian, Ron, thanks.

Does the attached driver look ok to you? Are the values of R1, R2 and R3 ok?

The lousy old LM358 produces much too much noise for it to have a voltage gain of 1000 times and has low gain at "hiss" noise frequencies.

 

The lousy old LM358 produces much too much noise for it to have a voltage gain of 1000 times and has low gain at "hiss" noise frequencies.

And both of us missed that the bass roll off capacitor has gone missing, leaving the LM358 amplifying its offset voltage by a gain of 1000.

 

Indeed I saw this too later, that I forgot the capacitor. I understand that it reduces gain at low frequencies but I didn't understand why it's included. I think I do now: to not amplify offset voltage. Thanks.

As for the opamp I thought I could use a crusty LM358 since it's not directly in the audio path. Perhaps an NE5532 or TL072 would be less sub-optimal?

Those Vactrols are hard to find. If I would use a VCA then presumably I will also need to generate a voltage ramp to feed to the VCA? The application is a guitar noise gate.

 

You can make a simple good-enough-for-guitar VCA with a transistor. The diodes in the side chain need to be the other way round as the output is off when the transistor is on.
The Lm358 might do for the side-chain, but you need something better for the signal path. Nothing wrong with the TL071, it won‘t eat batteries as fast as a NE5534.
Turn the diodes round and you get a sustain pedal.