Beogram 4000: A Fun Time with a Suddenly Activating Solenoid, an Anemic Carriage Motor, and an AC Motor Driver with a Weak Driver Amplitude

I am making good progress with the Beogram 4000 that I recently started to restore. More about the standard restoration tasks in a following post. This post focuses on two non-standard issues with this Beogram that I encountered after rebuilding its boards and motors.

I started realizing there was something wrong when I adjusted the platter height relative to the arms. During this process I needed to drive the carriage, so I plugged the Beogram in and I realized that the carriage movement was a bit slow compared to other 4000s I had worked on previously. I measured the 24V rail, and it only showed 21V. A bit odd. Usually this is 25-26V due to a warm Schottky diode, or 23.8V in my case since I now always replace the original 1.3W diodes with a cool running modern 5W type, which brings the voltage down to 23.8V at the capacitor terminal. But in any way the presented voltage was way too low.

Then I accidentally touched the carriage in a way that it was lifted up in the back a bit. This briefly activated the arm lowering solenoid. Very strange!!

I was able to reproduce this and it dawned on me that there must be a major issue, somehow putting a significant voltage onto the carriage when it touches the floating chassis (the carriage is normally insulated due to the plastic components that hold it on the guiding rods. The solenoid issue even 'worked' with the Beogram turned off (but plugged in).

I put the multimeter to it, and after poking around for a while, I realized that the enclosure of the Beogram was at ~30V relative to the system ground. Not a good sign, I counted myself lucky that there had not already been some 'magic smoke' emitted from an unlucky component. I also realized at this point that the transformer got unnaturally warm. 

30V 'smelled' like the unwanted voltage originated from the 24V system, which is fed by about that voltage before the Schottky diode (OD2)/transistor (TR1) regulator stabilizes it to ~24V. 

I determined that there was a ~9 Ohm short between the positive end of the 24V 3000uF reservoir cap (0C4) and ground:

More precisely, the short was between the red lead connected to the cap and the chassis. I followed the red lead, which goes to the collector of 0TR1. So first I suspected 0TR1 might have an issue, but it turned out to be o.k..

But at the collector of the chassis-mounted 0TR1, the red wire from the cap meets a thin blue wire, which goes straight to the carriage circuit board, and from there to the right contact of the solenoid, supplying it with un-stabilized 30V for maximum action! The plot thickened. I successively unsoldered/measured/re-soldered all the connections as I 'moved away' from the collector of 0TR1 keeping an eye on the resistance to the chassis. I made it to the left solenoid contact, and then to the terminal where the solenoid connects to the collector of its driver, 0TR4. All the while the short remained, i.e. all that separated me from the chassis was now 0TR4. A closer look at 0TR4 revealed that previous 'creative human interaction' had caused the bolt insulator to go AWOL:

The way this transistor is bolted down, a direct short is made between its heat sink (=collector) and the chassis via the bolt. What was missing is a part like this, which I fortunately had in my stash:

I installed it

and this cured the spontaneous solenoid activation issue. Details, details!!...;-)

I measured the 24V rail again, but the voltage was still only at ~21V and the carriage still moved too slow. So there had to be a secondary issue draining power from the system. Also the transformer still ran pretty hot. The other 'power electronics' system of the Beogram 4000 is the AC platter motor. So I measured the AC platter motor signal just to make sure this was working properly. What I saw was a much too low signal, however:

The RMS voltage was only 2.5V with the motor pot adjusted to the maximum smooth waveform. It should be about 5.5-7V when everything is up to specs. This suggested the secondary issue was in the motor driver. I moved the main PCB out of the way and had a look at the push-pull output stage of the motor oscillator beneath it:

The left transistor is 0TR2 (NPN, push) and the right is 0TR3 (PNP, pull). Since the motor wave was about 50% of normal, I immediately suspected one of them being in trouble, causing the motor to run only on one 'cylinder' instead of two. I touched the transistors and I found the PNP to be pretty hot. I extracted it and measured it with my transistor tester, and it appeared to be in perfect working order. So it had to be the other one! I extracted it, too, and indeed, it had turned into a small resistor:

This basically caused the NPN transistor to carry a big DC component along with the AC fed to the motor, explaining the high heat load.

I replaced 0TR2 with a BD437G power transistor (45V, 4A - a bit stronger than the 3A original TIP32),

(this modern package does not need a bolt insulator since the heat sink is encased by plastic on all its bolt facing surfaces)

and that cooled down things while yielding a normal strong motor signal for 33 RPM

and 45 RPM:

After this the 24V rail also was fully restored to 23.8V, and the carriage now moved much more quickly. 

On to do the final adjustments and giving this baby its first restored spin!

UPDATE: 

My customer suggested: "You write that 0TR3 (PNP, pull) was hot but working.  Running transistors hot obviously shortens their lives and I’m just wondering, now that you know this circuit had a problem, would it be worthwhile to just replace 0TR3 pro-actively to make sure that there isn’t a latent issue that could lead to another failure downstream?"

I thought that made a lot of sense!...;-), and so I put in the complementary PNP (BD437) in, too:

If you do this at home, note that the BD437/438 SOT-32 packages have base and emitter reversed compared to the original TIP31/32 TO-220 types, i.e. the brown and blue/orange wires need to be reversed.