I was sent an Oberheim SEM / ARP  Odyssey Filter Set module to repair.

Construction

This is a three PCB sandwich. The first layer is the panel PCB with all the controls. The second layer is the ARP Odyssey filter.

The third and rear PCB is the Oberheim SEM filter.

Operation

This started as a simple calibration so I didn't take any scope photos. However, I found the variable waveshape control never worked. It was designed with a LM3914 LED driver using the first four segments to select the variable waveform. These outputs were chopped with a 50 KHz square wave to control a MAX364 CMOS switch. Each switch was for each of the filter outputs. These were summed in an op-amp. I found that op-amp biased to the rails in most settings so little sound came out. A a 5 piece bused 100K network was used for the four input resistors and the feedback resistor. I thought at first the feedback resistor was picking up crosstalk or leakage so moved it out of the resistor pack. That did improve operation, but not nearly enough. I found if I decreased the values to 10K then the waveforms nearly correct but with some offset. I simply added a 47 uF NP capacitor in series to center them about ground. You can see that capacitor sideways along with my 4 resistor "network" in the above photo.

back