Check the capacitor and and measure the 50 microphones, there is continuity in the RS RC CS compressor terminals and none is grounded. Even so the compressor may be locked? If so, what would be a solution?
I tried it in a direct way and what it does is an assumption like wanting to come in and after a moment it sounds like something opens inside. (pouf) and if I leave it repeat the same and get hot.
You took the consumption in the test you put direct voltage with a new capacitor but in any case if the test is with new and direct capacitor it is logical that the compressor is locked and you will have to replace it is a shame is the worst thing that can happen to a split