Beogram 6000 (5512): Dead IR Diode in Carriage Position Sensor

After restoring the PCBs and the platter motor of Beogram 6000 (5512) that I am restoring right now it was not working properly. The platter rotated, so I was able to do the RPM stability measurement, but  the carriage did not go to the 30 cm set down point, but rather traveled across the entire platter without intervention. It also immediately switched to 45 RPM after start. This suggested that the culprit was not the record detection system, but rather an issue with the carriage position sensor. This sensor uses an IR diode as emitter and a photo diode as detector. The plastic 'ruler' that is attached to the carriage has black markings on it, which allow the control system to detect what action is needed depending on the carriage position and speed of travel.

That it switched immediately to 45 RPM after traveling for about 1 cm after pressing start suggested that either the IR diode or the photodiode were dead. A strong light beam from a LED flashlight elicited a response from the photodiode. That left the IR diode. I measured the voltage across it and the result was 0.95V instead of the prescribed 1.1V. This was remarkable, since these diodes usually die by going open circuit, i.e. one usually measures a much higher voltage across it. First I tried adjusting R88, which controls the brightness of the diode, but it was maxed out already...so someone tried to adjust it previously...I took the carriage PCB out, which is an easy task in DC motor Beograms since it is just plugged in (one does have to unsolder the two leads to the carriage motor) and exchanged the IR diode with a low intensity amber LED and put the board back in. This is how it looks in action:

One can pretty much use any standard LED to fix a broken IR diode on this board. The photodiode is responsive to visible light. The issue is that modern IR diodes are designed for high output and high current, and so it is better to use an easy to find low output standard LED. The sensor does not need much light intensity. A crucial point of this fix is to adjust the working point of TR17 properly via the LED brightness trimmer R88. This shows my multimeter hooked up to the base of TR17 where 0.7V are specced in the circuit diagram. Just adjust R88 until 0.7V appear while the sensor is under illumination (i.e. none of the black markings can be between the LED and the sensor for this adjustment).

After this procedure the deck operated normally.