touane said:
R23-->recommended 22R. Both in simulation and reality, it all sounds horrible with this value, and I ended up setting it around 68R, which is also the value appearing on the original schematics. Could you mention how you set this one?
I did readjust this one a lot, but after I found out my problem was elsewhere I set it back to around 22R. Since I have dual channel unit, I set those to match channel-to-channel gain exactly.
What I found out is that this trimmer simply sets the gain of the unit. But if you go with too much gain it can run the channel into shutdown. So you don't have unlimited range of gain setup.
Be aware that this compressor is feedback topology. If you raise the gain on R23 it affects the gain in the sidechain and that consequently changes the compression amount.
I think that you can set it at whatever resistance you feel it sounds good as long as sidechain is behaving properly!
touane said:
Oscillator: I imagine you also used the 555. Could you mention which freq. and which duty cycle you ended up with?
I never had problems with oscillator. I built it with 555 and first I checked if it is getting proper supply voltage (in my case few mV higher than -12V). Then I used my cheap chinese Hantek to trim chopper pulses to the specified 250kHz. It was no big deal. Once done, I never had to go back to it.
touane said:
Would love to hear more (if you recall) how you proceeded to set it up.
Problem with my dual unit was, as I mentioned somewhere before, with shorted R44 trimmer on one channel. It was manifesting as low output gain on that channel. I first went and tested main signal path on both channels. For it to be tested, I had to disconnect entire sidechain. It was messy since I didn't socket my transistors in the first place and solder pads on this pcb are very small with narrow holes. Disconnecting sidechain is done by lifting mid leg of the chopper transistor PN3640 and disconnecting interstage transformer's primary. That way you can test your input-to-output signal integrity.
I found out that when using BC550A on TR12, as recommended by Abe, is crap. I used 2N4124 as recommended by SR1200 and it performed well. As well, I did my first mistake not taking care when stuffing board in the first place so I put BC550C on that position! 550C has way too much gain. But then I got 550A and it didn't perform well either. When using 2N4124 it was all good.
Luka