Beogram 8002: Repair of Intermittent Tracking Issue

Oh well, I guess I declared victory a bit too early on the Beogram 8002 that I recently restored for a customer in California. After celebrating the restored Beogram with an awesome Manfredo Fest record

I continued playing it for a few days. I always like to do this before I send a Beogram back to its owner trying to reduce the likelihood of intermittent issues coming up later.

During this 'play period' of my restoration process I started noticing that there were sometimes irregular distortions during playing records. Sometimes the needle even skipped occasionally on records that I knew are in pretty decent condition and never skipped before. So I had a closer look at what was going on. After watching the carriage mechanism for a while I finally realized that this unit exhibited 'reluctant tracking' similar to what one notices occasionally in the earlier Beogram 400x series of turntables. Their carriage motors develop too much friction internally and that causes tracking 'hiccups' where the carriage only moves in larger jumps after a few turns of the platter. If the carriage mechanism is properly restored and has low friction, this usually indicates that the motor itself needs a higher voltage to overcome the initial "sticktion", to go from standstill into motion. On the 400x this does not cause very noticeable distortions or skipping, presumably due to the larger mass of the arm assembly. 

But in the 8002 with its ultralight arm and tiny cartridge, such jumps apparently can cause serious vibrations in the needle which can be quite audible. In severe cases this even causes the needle to jump into an adjacent groove.

This experience suggests to me that the 800x turntables now also may start developing carriage motor issues as they enter 'middle age'. 

This meant I had to setup this Beogram 8002 again in service position to have a closer look:

I really started liking my new 'Lazy Susan' bench setup. It allows working on fragile setups from all sides without the need to bend wires all the time etc...Amazing that it took me more than 15 years to come up with this approach!...;-). On the scope you can see already a trace measured after I had fixed it. I only took this picture after I was done already.

Let's see what I ended up doing:

First I studied the circuit diagram. I never had the need to work on the carriage driver circuit of a Beogram 8002 so far. This snippet from the diagram in the service manual shows how it all works:

One of the two opamps integrated in IC2 is responsible for driving the push-pull stage (TR1/2) that comprise the H-bridge half that controls the current through the carriage motor OM1 in forward direction that is used for carriage advancement during play of a record. In this condition the current flows from the +15V rail through TR1, R43, OM1 and then to ground via TR4 on the other side of the motor. For this to happen the TR1 and TR4 need to be conductive. TR4 is turned fully on by the other opamp in IC2 (not seen in the picture above), by pulling the base of TR4 to ground.

The control of the current happens via TR1 whose base is controlled by pin 7 of IC2. This opamp is configured as 'non-inverting': Its "-" input is connected to the comparison voltage defined by the R20/R18 voltage divider between +15V and ground which is fed into the "-" input via R22. This voltage (10k/230k*15V=0.65V) is compared with the voltage on the "+" input, which defines the output voltage of the opamp at pin 7: Without a feedback line between output and "-" input this would result in a digital behavior where pin 7 would shoot up close to the 15V rail voltage whenever the "+" voltage is larger than the "-" voltage, and go near 0V whenever "+" would be smaller than "-". This comes from the fact that opamps are usually setup to have an 'infinite' gain. To get a more controlled behavior where the output actually mirrors the "+" input smoothly as a linear amplifier, a feedback line is implemented between the output (pin 7) and the "-" input. This is marked green in the above picture. This feedback connection via R21 forms an additional voltage divider with R18. This allows setting the gain of the amplifier via R21: A smaller R21 reduces the voltage swing on the output of the amplifier caused by a change of the input voltage present on pin 5, while a larger R21 increases the proportional voltage swing at the output.

The rest of the tracking feedback process is straight forward: The movement of the arm as the needle is pulled towards the center of the record is translated into a positive proportional voltage on pin 5. This happens via an aperture bolted to the arm that increases the amount of light on the photoresistor 5R1 coming from 5IL2. More light means less resistance in a photoresistor. 5R1 forms a voltage divider with R27, which scales the approx. 4.4V that apply to the other end of the photoresistor during play of a record into the "+" input of the opamp (via D4). So whenever the arm gets pulled toward the center of the record, the voltage at the "+" input increases a bit. If it is above the voltage at the "-" input, the output of the opamp goes up, this turns on TR1 a bit more, which increases the current in the motor. After surpassing a certain threshold (defined by friction) the motor starts moving and the carriage starts following the needle. Once the carriage moved enough to reduce the light on the photoresistor sufficiently the voltage on the "+" input drops, causing the motor to stop. The result is that the carriage moves in small bursts after the needle until the end of the record has been reached.

In the case of this Beogram 8002, the bursts were much larger than normal, i.e. the carriage only started moving after the needle had been pulled a few platter rotations inward. When the movement occurred it gave the carriage a fairly strong jolt, which then caused the distortion and skipping issues. This can be seen on this oscilloscope trace, which was measured at the output of the opamp (pin 7) during play of a record:

Each of the peaks corresponds to a platter rotation. The peak is essentially caused by the eccentricity of the record, causing the arm arm to dance left and right a bit during the inward motion. This causes the aperture to swing left and right which causes the pin 7 voltage to vary accordingly. The trace basically shows how the arm gets pulled further and further inward, gradually increasing the motor voltage. This goes on until the voltage of a peak is finally high enough to make the motor move against the friction. Then the carriage adjusts its position and the process starts anew.

When everything is in good shape there should be a small carriage movement about once per rotation. So the question at this point was how to restore this ideal situation.

The first step was of course to have a look at the motor. I had a weak hope it would be the same type that was used in the earlier 400x designs. It would have allowed me replacing it with a new Beolover Carriage Motor for Beogram 4000, 4002, and 4004 and be done with it. But to no avail, the Beogram 800x series has a different setup:

The reason is that the system voltage of the Beogram 8002 is different and so a different motor design was used. 

This left trying to find a workaround until there will be a replacement for 800x carriage motors!

I immediately thought of ways to increase the voltage that is proportionally applied to the motor when the aperture moves a certain distance. This suggested an adjustment of the gain of the opamp that drives TR1. So I replaced the feedback resistor R21 with a 300k trimmer:

Then I increased the resistance gradually while watching the oscilloscope during record play. When I was happy with the carriage movement I measured the resistance and found that I had reached a value of 166 Ohm on the trimmer.

The oscilloscope trace at this point looked like this:

No more steps after a few turns of the record. Visually, the carriage spindle moved a little bit after each rotation of the platter as it should be. I replaced the trimmer with a fixed value resistor and then it was time to play a record and listen to it. I am happy to report that there was no more skipping and I also did not notice distortions anymore. So I hope this issue is fixed for now and this Beogram is ready for service again. I will listen to it for a few more days, and it this turns out to be true it will be finally time to send it back to its owner.