Beogram 4000: Carriage Returns Home Immediately After Setting Down on the Record

Oh well, the Beogram 4000 that I restored in January 2025 came back to my bench exhibiting an issue I had not experienced before:

After pressing ON the carriage went and found the LP setdown point and put the arm down. Immediately after hitting the run in groove, the carriage returned home with the arm down. It did that at a fairly high speed. Once home the unit shut off and the arm lifted.

Since the << and >> buttons worked fine with the arm up, a problem with the H-bridge was unlikely. This meant there must be an issue with the tracking sensor.

The Beogram 4000 features two photoresistors FC1 and FC2 that control the reverse and forward directions of the carriage motor (later 4002/4 models are differently set up, with only one photoresistor for the forward direction). This means the dual photoresistor tracking sensor can control the carriage in both directions, forward and reverse. The carriage is at rest when the forward and backward currents through the H-bridge (i.e. the motor) just cancel out. This condition is reached when the aperture in the tracking sensor illuminates both photoresistors with the same intensity, i.e. when it is positioned just between the photoresistors. If everything is properly adjusted, this condition coincides with the tonearm being parallel to the sensor arm. As the arm gets pulled towards the record center during play the arm gets deflected laterally, which causes the forward current to become larger than the reverse current, and the carriage starts moving forward, following the motion of the arm.

A consequence of this setup is that if one photoresistor circuit is not working properly, the carriage motor will already see a significant current even with the tonearm in parallel alignment, and the carriage will start moving. If the failed photoresistor circuit is the one that controls the forward motion, the carriage will always go reverse, independent of the lateral deflection of the tonearm.

With this in mind, I started examining the forward direction circuit. A good rule of thumb with the Beogram 4000 is that 90% of its issues are typically caused by bad contact terminals in the many switches, or by a wiring problem. The Beogram 4000 is an early 1970s design, when wire-to-board connections were still hardwired (instead of using modern wire-to-board headers and plugs, which are mechanically much more resilient). These soldered wires have a tendency to break off at the solder terminals when the Beogram is handled or transported.

This personal 'statistic' was confirmed in this unit: When I had a closer look at the wiring I immediately found the problem. The wire connecting the forward photoresistor FC2 to the H-bridge was broken off at its terminal on the main PCB (yellow marking in the photo):

I cleaned wire end up and soldered it back to the terminal. And the deck was operating normally again! I will play this Beogram a bit and then it should be time for sending it back to its owner in Australia!

Beogram 4002 (5501): Random Stop Issue Fixed - Defective Photo Resistor in the Spindle RPM Sensor

A Beogram 4002 (5501) that I restored last year developed a strange new problem: When playing a record the OFF (>>) function  would be triggered randomly during play as if someone pressed the >> key on the keypad.

Before I will discuss the fix, let's have a look at the circuit diagram (click on the diagram to get it in full resolution):

TR17 in the 'Electronic Switch' is responsible for triggering the >> function of the turntable. Whenever its base is pulled to GND its collector goes high to ~18V and then the arm is lifted and the carriage is driven home. There are three ways this can occur: Via two mechanical switches, one being the "End Switch" (ES) under the carriage and the other the ">>" key on the keypad. The End Switch is the one that sends the carriage back whenever it is driven all the way to the left (i.e. manually via the << key or automatically if there is no record on the platter and the arms go all the way searching for one).

The third way to trigger >> is via the end groove detection mechanism ("RUN-OFF STOP" on the diagram). This works via TR20 whose base is pulled up when the end groove is detected. That connects its collector to ground, and with that the >> function is activated.

The end groove detection mechanism works via an 'analog rotary encoder' that detects light flashes impinging on a photo resistor (OR2) from a bulb (OIL1) that shines its light though four holes in the carriage pulley as they pass by (creating light fluctuations on OR2). 

These light pulses short OR2 to GND causing TR21 to shut down, which increases the voltage at the collector of TR21 to about 19V. These voltage pulses (see oscilloscope trace schematic on the circuit diagram) charge C33, which, via the voltage divider formed by R88/89, pulls up the base of TR20, triggering the >> function.

It is interesting to to note that the bulb OIL0 only comes on when its switch to GND is closed. This switch is activated by the carriage when the tone arm gets close to the record label. This switch is activated by the same tab on the carriage assembly that activates the 17 cm (singles) set-down point switch ("B" in the service manual). This is the reason that the tab that activates B has a long flat shape. Anyway, this switch aims to prevent triggering the the RUN-OFF STOP mechanism when the carriage briefly moves faster between two tracks of a record when the carriage is still far away from the label.

Why does the mechanism not trigger RUN-OFF STOP when the carriage advances normally while playing the last track when OIL1 is already on? C34, R91 and D30 form a network that discharges C33 when there is no light on the sensor. This means there is a competition between charging during illumination episodes, and discharging when there is no light on the sensor. So if the pulley rotates only slowly and occasionally during playback of a track, discharging 'wins' and the base of TR20 is not pulled up high enough. But when the light flashes happen quickly like when the carriage moves fast pulled along by the end groove, then the charging mechanism wins and TR20 is turned on. The beauty of analog control systems!!

Ok, back to the 'Random Stop Issue': After verifying that the ES and OFF switches were working properly, I traced the signals from the base of TR17 into the RUN-OFF STOP circuit. This shows the oscilloscope traces that I measured:

The traces are assigned as follows:

• Yellow: TR17 collector (when this signal goes high the measurement was triggered - the graph shows ±5 sec around the >> event)
• Blue: TR20 base
• Green: TR21 collector
• Red: TR21 base

These traces show that before the >> event happens there is a random statistical fluctuation of the voltage at the collector of TR21 (instead of being close to GND if there is no end groove event). These voltage flashes obviously correlate with the signal at the base of TR21 (suggesting that TR21 is working properly). This leaves only one conclusion: The sensor OR2 randomly short circuits to GND causing these voltage spikes. These spikes can be enough to charge C33 and when the voltage one C33 gets high enough due to these random events, the >> function is triggered. This is seen in the blue trace, which shows that the voltage exceeds the ~0.6-7V threshold needed to turn on a standard silicon based transistor when TR17 is activated.

All this told me that OR2 was faulty. This meant it needed to be replaced. Since this is a special B&O part, which is not available anymore, I decided to design a 3D printed assembly to replace the entire bulb/sensor assembly on the pulley. This shows the setup after I implanted an LED to replace the light bulb last year:

The orange part contains the LED (Newark 78R6602) and its 2k current limiting resistor (Newark 26R3983). It was designed to stick onto the OR2 sensor housing. The current design replaces the entire encoder setup. These are the structural parts of the assembly:

Assembled they look like this:

And with LED, resistor and photo resistor (type "5516", ~10k resistance when dark, ~500 Ohm when illuminated) installed:

And after installation and in action:

Once the set-up was installed I measured the signal at the collector of TR21 when the LED came on:

This signal looks pretty much like what the manual demands. I tested the Beogram by playing some records, and it seems everything is working again! So I am hopeful that this fix took care of the issue.