ebartlet
Well-known member
Just finished the usual mods on two 3630's - upgrade power supplies, op-amps, and converted one unit to feed back compression.
After upgrading the audio and side-chain op-amps for the Rev B unit to ST MC33079's I had in stock, I noticed some parasitic oscillation (1.6Mhz) that occurred at higher input signal levels on the negative peaks of the output. Turns out there is exactly "0" bandwidth limiting on the input OA stage, so a cap accross the NFB resistor was in order.. I decided on 100pF (across 8.2k) for a -3dB point around 200KHz (which will vary depending on the setting of the -10/+4 switch)
On the Rev D, after upgrading op-amps, channel A seemed to be fine, but channel B oscillated without any signal at a level of -20dB!! Swapped op-amps and the problem stayed with channel B, must be something with the PCB layout, installed aforementioned caps on these as well and all is well.
Decided to mod the gates on these as well. I don't use gates very often, but when I do, I prefer to have the gate perform reduction rather than switch on and off with signal level. The choppiness is too obvious, and really not necessary as a moderate attenuation will reduce the offending noise enough, especially in the context of a mix.
It would have been nice to be able to install a pot to adjust the amount of attenuation when the gate is closed, but there is very little extra room on the front panel, so a fixed value would have to do.
I settled on a -22dB gain reduction when the gate is closed, which is achieved by changing resistors R32 and R36 from 820 ohms to 2.2K. Obviously this value can be played with to achieve varying levels of reduction. I've attached a jpg showing the location of R32/R36.
After upgrading the audio and side-chain op-amps for the Rev B unit to ST MC33079's I had in stock, I noticed some parasitic oscillation (1.6Mhz) that occurred at higher input signal levels on the negative peaks of the output. Turns out there is exactly "0" bandwidth limiting on the input OA stage, so a cap accross the NFB resistor was in order.. I decided on 100pF (across 8.2k) for a -3dB point around 200KHz (which will vary depending on the setting of the -10/+4 switch)
On the Rev D, after upgrading op-amps, channel A seemed to be fine, but channel B oscillated without any signal at a level of -20dB!! Swapped op-amps and the problem stayed with channel B, must be something with the PCB layout, installed aforementioned caps on these as well and all is well.
Decided to mod the gates on these as well. I don't use gates very often, but when I do, I prefer to have the gate perform reduction rather than switch on and off with signal level. The choppiness is too obvious, and really not necessary as a moderate attenuation will reduce the offending noise enough, especially in the context of a mix.
It would have been nice to be able to install a pot to adjust the amount of attenuation when the gate is closed, but there is very little extra room on the front panel, so a fixed value would have to do.
I settled on a -22dB gain reduction when the gate is closed, which is achieved by changing resistors R32 and R36 from 820 ohms to 2.2K. Obviously this value can be played with to achieve varying levels of reduction. I've attached a jpg showing the location of R32/R36.
