Beogram 4002 (5501): Repair of a Strange Fault Caused by a Shorted Diode D6

After replacing the reservoir capacitors in the Beogram 4002 (5501) that I am restoring right now, a strange control system malfunction remained. Here is a description:

After pressing START, the carriage would start moving but then after about 20 mm it would stop and the lights would go out. Then after a few seconds, the lights would come back on and the motor would run on 45 RPM. Once on 45 RPM, I was able to drive the carriage with the >> << buttons, but arm lowering would not work anymore. Switching back to 33 RPM killed everything again, but lowering would briefly function before the lights went out. Pretty strange all this!

Luckily the root cause turned out to be an easy fix. Here is the relevant part from my annotated circuit diagram:

After a bit of head scratching and doing some measurements, it occurred to me that a short circuited D6 would kill the 22V rail if TR2 is on, as it is the case when 33 RPM is selected. This was quickly confirmed by a 0V diode tester reading across D6 in both polarities. After replacing D6 with a new 1N4148 signal diode everything was working again properly.

This is how this fault 'worked': When TR2 is on then its collector is on GND level. If D6 is making a short, GND is applied to R2/D1 (dashed red lines). This kills the voltage on the base of TR1 which then kills the voltage on the base of 0TR1. This turns the 22V rail off and the lights go off. The time delay that was inherent in this process is enabled by C1, which preserves the base voltage on 0TR1 for a couple seconds until the capacitor is empty. At this point TR2 turns off. This allows the voltage to recover since now the anode of D6 is no longer grounded. So it does not matter if the diode is shorting. The voltage rail recovers, and that turns TR3 on via L1 and R8, and the deck is now operating on 45 RPM. Ahhh, the mysteries of analog control systems! This is Beolove!