After reinstalling the drive unit into the Beocord 5000 (4715), I began testing. Initial impressions were positive, with fast forward, rewind, play, and stop functions all working. However, the Auto Stop function would cease working after a few minutes. The tape would reach the end, but the drive mechanism would continue running, potentially harming the motor. The only solution was to turn off the deck for an hour to reset the function.

The Pause function was also inconsistent, often failing to engage. Further testing revealed that rewinding was significantly slower than fast forwarding, with the speed dropping worryingly low at the end of the rewind cycle. This was unexpected for a dual-motor drive mechanism. The speed difference was perplexing, as friction should be similar in both directions. While the free-wheeling mechanism’s pulleys reverse the winding direction, they had been lubricated and shouldn’t cause such a discrepancy. Manual cranking of the motor also suggested minimal difference in torque between directions, ruling out friction as the primary cause.

Extensive investigation and voltage measurements, compared against the circuit diagram, revealed a problem with transistor 3TR27 in the Pause flip-flop. Its base measured -3.8V when paused and -10.8V when running, a significant difference from the expected -10.2V vs. -10.8V based on the transistor’s silicon turn-on voltage. This contrasted with the correct values exhibited by its counterpart, transistor 3TR26. Replacing TR27 with a 2N2222 NPN transistor resolved all the issues. Pause and Auto Stop functions became consistent, and the rewinding speed matched the fast forward speed.

Interestingly, the extracted BC548B transistor tested fine, suggesting a possible bad solder joint as the culprit. However, the possibility of a heat-related issue affecting performance over time could not be ruled out. Regardless, the repair was successful. This experience highlighted the intriguing nature of working with analog networks.

For reference, post-repair winding times for a new TDK SA90 tape were approximately 85 seconds in both directions. Auto stop time from the tape stopping was around 7 seconds.

Before replacing TR26, rewinding took 120 seconds, while fast forwarding took 85 seconds. Auto Stop, when functional, also took ~7 seconds.

During troubleshooting, I investigated the Auto Stop circuit, initially misunderstanding the 1977 circuit diagram’s terminology.

The 0 to -10.9V step at 3TR12’s collector was meant to be a pulse, indicated by the » « symbols (initially misinterpreted as related to Fast Forward and Rewind). Connecting a multimeter didn’t initially reveal the voltage jump. However, an oscilloscope showed the pulse:

This pulse, lasting only 10-20 microseconds, charges 3C5 (100nF) via 3R15 (120R). The short duration aligns with the combination’s time constant of roughly 10 microseconds. However, it’s sufficient to briefly trigger 3TR4’s base, initiating the Stop function via 3TR5.

The pulse at 3TR12’s base was also measured. It appears as follows (scale: -11V [baseline] to -10.4V [peak]):

Connecting an oscilloscope or multimeter to either side of 3R41 (1M) deactivated the Auto Stop function due to the measurement instruments’ internal impedance. This high-impedance network requires careful consideration during measurements to avoid impacting performance.

With the repairs complete, it’s time to experiment with recording and perhaps even create a mix tape.