Beogram 8002: Record Detection Malfunction

The Beogram 8002 that I am restoring right now had issues with the record detection function. When pressing start with no record on the platter it would drive the arm to the start groove position and then lower it onto the platter. And that is of course not acceptable considering the high prices of rebuilt cartridges these days...so I had another look under the hood. Here is the relevant section of my annotated circuit diagram:

When no record is present the photocell PH1 generates an alternating current due to the intensity change caused by the black strips on the platter alternating with reflective sections. This current charges C15. The resistors R65 and 64 are there to allow the capacitor to discharge when the photocell is off. R64 allows to change the time constant of the discharging process. By optimizing R64 the AC amplitude that goes into the base of TR14 can be maximized. TR14 ad TR15 amplify the AC signal which finally charges C19 and a voltage builds up at the base of TR16. TR16 turns on and HH is pulled to ground. HH feeds directly into the microcontroller, which therefore can detect the absence of a record and prevent arm lowering and/or drive the carriage towards the center of the platter in search of vinyl.

I hooked up my oscilloscope to R71/R72 and measured the signal with the sensor arm over the rotating platter. What I saw was a signal similar to the curve shown in the circuit diagram. At the same time I measured HH, which turned out to be somewhat inconsistent, fluctuating between 0.15 and 0.5V. This pointed towards a too small pull at the base of TR16 which would therefore not turn completely on.

I tried to adjust R64, but was not able to produce a larger signal at R71/72. After some head scratching I realized that I had maxed out R64 to obtain the largest possible signal. This was strange since one would expect that R64 would be dimensioned in a way that the maximum signal would be obtained somewhere in the middle of its resistive track. I measured the resistance across R65/64 and I saw only ~3k. This was strange since in the circuit diagram 15k and 20k are parallel, which should yield 8.5k...I removed R64 and it turned out that it was a 4.7k trimmer. I replaced it with a 50k 25 turn trimmer and did the adjustment again. And this time I was able to adjust to a clear maximum (8V) of the RMS voltage at R71/72. It is a good idea to use the RMS voltage of the sawtooth since its maximum will correspond to the maximum current pumped into the the base of TR16. This is the signal I measured after optimizing R64:

The corresponding voltage at HH was now a constant 0.13V.

Pressing start now also resulted in the appropriate response of the turntable: It did not drop the arm anymore but instead drove it all the way across the platter and then back to the home position. So this may be fixed! On to the repairing the missing digit in the RPM display.