In a 2400, the insert output will oscillate because the feedback is taken from the other side (output) of the output capacitor. This is a disaster waiting to happen if you have certain transformer-balanced inputs that are not terminated precisely. It can also cause oscillation if you have a long cable patched in. In my case, a Gates M6629 Solid Stateman limiter would do a low-frequency oscillation. Also, just about anything patched in would almost cause an oscillation in the insert output, and changing the op-amp to anything else - even a TLE2072 - caused a horrible mess. So I studied it and analyzed it in SPICE and figured out an easy fix...
I replaced the output capacitor with a 220uF electrolytic, moved the feedback point to before the capacitor, and put a 22pF capacitor from the op-amp output to the inverting input, but just in that one spot. I also looked through the rest of the design for the same problem and found that the schematic did have the same problem on the group cards, but the group cards were wired in the correct manner so I didn't have to make any changes. The schematic didn't follow the PC board, though.
I upgraded the power supply and used a lot of LM833 and NE5532's in there, along with TLE2071 and TLE2072's. I used thost new dual op-amps from National for the mix bus amps and that really cleared up the sound. Replacing all the capacitors got the bass consistent across the whole board, and replacing all of the ceramics cleaned up the HF response when the EQ was in use. I added extra rail decoupling only because the 5532 and LM833 require better decoupling. The only change I percieved with the 5532 or LM833's was a significant drop in hiss. The biggest op-amp improvement was in the mix bus amplifiers.
-Dale