This is getting a bit difficult to debug without an oscilloscope.
The "fluctuation" on the anode side of D37 should be a very short pulse (about 1ms) of about 0.8v when a note is played (this is the "trigger" for the envelope). Naturally measuring this with a multimeter may not result in a meaningful display, even if that pulse is ok. Since there actually is something happening on that side I'll guess the trigger is there.
On the cathode you should get the envelope portion: On pressing a note it should jump to about 12v and then fall again. My guess now is that the problem is somewhere around Q39 to Q41. Please inspect them visually (if they are the right types and correctly positioned). You can also check if 12v are present at the middle pins of Q40 and Q41.
Perhaps you may want to check before if R152 and R138 have to correct values.