I just repaired my TD-145 that wouldn't auto stop. After reading all the comments about mechanical adjustments I have this to say:
If no one has monkeyed with the mechanical adjustment of the coil and magnet, DO NOT TOUCH IT! The circuit is very sensitive to the HFE of the transistors. After measuring voltages and analyzing the circuit I figured out how to troubleshoot it and repair it.
1. The TD-145 manual is garbage. It talks about measuring the AC voltage at point B but it doesn't tell you what postiion the arm is supposed to be in! If you try to set it for 2V when the arm is in the rest, it won't even get close! The TD126 uses almost the identical circuit. You can even use the schematic to troubleshoot the TD145, it's identical except for the control lines on the right hand side of the schematic, but the oscillator, amp, and detector are the same. There is a cardboard gig to position the arm for adjustment. No pictures of it I could find. But the TD126 manual tells you how to adjust the position of the magnet vs the coil with the cardboard gig in place. If no one has messed with it, all you have to do is move the arm so the leading edge of the pole piece of the magnet is 1/3 across the sensor coil. With the arm there, adjust for 2V at point B. If you don't have a meter that is flat out to 80KHz, use a scope and adjust for 2.6V peak.
2. If all that is ok and it still won't work, look at point C with a scope. Point C should be very sensitive. On mine, after I fixed it, Point C was 5V with the arm at the adjustment position, and if you moved it inwards a bit, Point C went to 0V, I think. If it reads higher, you need to remove T104 and check the HFE. It should be greater at least 100. Mine read 20. Replaced it with a 2N3904 that read 130. All fixed.
3. Note that this circuit responds to relative quick movements of the arm, like the end groove or if it's bumped.