Beogram 4002 (5513): Repair of a Broken Sensor Arm Photocell with Three SMD Photodiodes

I recently restored a Beogram 4002 (Type 5513), which had a broken photocell in the sensor arm. One of the electrodes had come off from the photocell:

Here a picture of the removed photocell:

These photocells are not available anymore, and previously a solution involving using a modern phototransistor and a small circuit modification to adapt it onto the Beogram record detection circuitry had been developed.

When I encountered the above, I first wanted to replicate the phototransistor solution, but then I remembered that a photovoltaic cell (aka 'solar cell') is in its simplest form just a pn-diode. Such diodes can actively generate a current when exposed to light if a load is connected. The photocell in the sensor arm is basically such a diode. It generates a light induced AC current (as the platter ribs pass through underneath it) that is fed via a capacitor (to keep DC components from background light out) into the base of a transistor (TR3 in the 5513 circuit) that amplifies this current, which is ultimately used to drive a transistor that disables the arm lowering mechanism.

So the thought occurred to me that one could potentially use a photodiode and simply replace the photocell without further circuit modification. Phototransistors do not generate significant currents by themselves due to their symmetrical npn or pnp junction configurations, and therefore they need added circuitry in this case.

I ordered a few photodiodes for visible light detection that had a suitable form factor for this project. Indeed, most of them showed various voltages when exposed to light. I ended up using a PIN (p-type/insulator/n-type junction) silicon cell made by Osram (SHF 2716). It is a 0805 SMD package. I soldered two 32 gauge wires to it for testing: 

It yielded 0.45V into the 10 MOhm of my multimeter when exposed to my fairly strong bench LED lights:

I installed it using the original insulator tubes:

And I got a decent response at the collector of TR3 when I spun the platter under it and pushed the extracted sensor compartment a couple mm towards the platter. But the signal was still too small and had an inverted look:

I concluded that a single diode did not have enough current capability under the given light situation for driving TR3 properly, so I decided to replace it with three units connected in parallel for my next test. I also reversed the polarity:

I installed the assembly. This time I did not use the insulator tubes, but instead supported the leads by a small piece of suitably thick cardboard. I fixated the leads to it with a dab of epoxy to hold everything in the correct position:

And I measured a nice 5.72V amplitude at the collector at TR3 with the sensor compartment properly inserted into the arm!:

I also tested it with and without record on the platter and the arm did not stop or lower on the empty platter, as it should. All good again in the sensor arm department! This Beogram is ready to be sent back to its owner in California!