Beogram 4004: Restoration of Platter Motor, PCBs and RPM Panel

I recently received DC platter motor, main and output PCBs, and the Keypad of a Beogram 4004 from a customer in New Jersey for restoration:

Shipping went well, but please, do not put any tape on the keypad or any labeling!! These surfaces can be damaged when pulling off the tape. Luckily, in this case it went well, so no worries!

As usual, I worked first on the motor since the oil-infusion of the bearings can take up to 3 days. This shows the motor without its mounting plate:

I disassembled it to get the bearings out. The bearings are the two small donuts on the black pad upfront:

I submerged them in oil and pulled a vacuum. Immediately strong bubbling started:

The bubbling represents air drawn from the porous bearing material. Once the air is out, oil diffuses into the material. These bearings lubricate by slowly diffusing their oil content onto the shaft while the motor runs. At some point the oil is depleted, and that essentially limits the lifetime of the motor. Oil infusion under vacuum resets the clock! I wish old guys could do the same!...;-)

While the oil infusion was going on I worked on the remaining components. This shows the main PCB in original condition:

A detailed view of the RPM section, comprised of RPM relay and RPM trimmers:

I replaced all electrolytic capacitors, power transistors, RPM relay and trimmers and the sensor arm transistor:

Then I focused on the output board. In the 4004 it not only contains the output relay and delay circuit, but also the remote control interface for the Beomaster 2400, whose remote control is able to start and stop the 4004. This shows the board in original condition:

A detailed photo of the output circuit:

I replaced the electrolytic capacitors and the output relay:

This shows the output section magnified. I also installed a switch (red) that allows connecting signal and system grounds. This is helpful if there is a hum when connecting the Beogram to an amplifier:

Then I focused on the RPM panel above the keypad. This shows the panel extracted and flipped over:

Removal of the bulb covers reveals the incandescent bulbs that illuminate the trimmer scales:

In front of the bulbs the Beolover LED assemblies sit ready for installation. This shows the bulbs removed:

The LED assemblies solder directly to the bulb terminals:

This shows one of them magnified:

They do not obstruct the bulb covers, which can be re-installed after the switch:

Now it was time to install the components into one of my Beograms for testing. After bolting in the main PCB I replaced the two power transistors that are mounted on the solder side. Since their alignment is defined by the bolts that hold the PCB in place, it is best to replace them when the board is installed. This shows 1IC1 that controls the 21V rail:

Ii usually is a TIP120 originally. I replace them with a stronger TIP102  hoping to increase longevity:

These modern TIP devices have a tendency to introduce high frequency oscillations superimposed to the 21V rail, and it is advised to add a 100nF capacitor (the yellowish component in the picture) at the emitter to quench this nuisance. I also replaced 1IC4 with a TIP107. The final step in the main PCB restoration was the adjustment of the bias of TR3, which amplifies the sensor signal. I usually replace the bias resistor with a trimmer. This allows compensating for the particular gain of the transistor. Once the bias is adjusted

to get 4V at the collector, the trimmer can be moved to the component side of the board.

Luckily, my customer sent the keypad along. After I removed the RPM panel, cracked panel holders were revealed. Both of them had the same stress fracture:

Luckily, there are faithful plexiglass reproductions available from the beoparts-shop in Denmark:

All that needs to be done before installation is to move the metal spring clips over from the original parts:

This shows the new holders installed:

Beolovely! Now it was time to extract the freshly infused bearings from the motor oil:

I re-assembled the platter motor and also installed it in my Beogram. Then it was time for RPM stability measurements with the BeoloverRPM device:

First I measured in the 'slow' mode of the BeoloverRPM, which yields an RPM measurement every 10 seconds. This is a great way to measure the long-term stability of the RPM. This is the curve I measured after 24 hrs:

Not the best result I ever saw. Such jagged curves are indicative of a top motor bearing that still needs to 'settle'. I am not entirely sure, but I think that the reason for such variations is a slight misalignment of the re-infused bearing relative to the original position. In my experience this will smoothen out after one or two hundred hours of play. Basically the shaft needs to polish the bearing anew in an orientation defined by the pull of the platter belt. At any rate, these current RPM variations that are pretty small, about 0.2%, which is less than what most people can discern.

There is an easy and quick way to a perfect RPM curve: Replace the original DC platter motor with the Beolover SyncDrive. The SyncDrive is based on a synchronous brushless motor, which completely bypasses the analog control system on the main PCB and instead uses a digital on board controller. The SyncDrive gives DC motor Beograms a wow and flutter performance equal to earlier AC platter motors with a long-term stability of later Beogram 8000/8002 models.

The next measurement used the 'fast' mode of the BeoloverRPM, where it logs an RPM value every time a platter rib passes under the sensor. In other words, every platter rotation yields 24 measurement points:

This is a graph that covers about 60 platter rotations (~2 minutes):

What we see is a repeating zig-zag pattern that is caused by minute variations of the rib spacing around the platter superimposed by a wavy waveform. It repeats exactly every platter rotation. This repeating pattern can be considered a measurement artifact that is not related to RPM changes. The longer wavelength wavy pattern however represents 'wow and flutter', i.e. RPM changes caused by the platter motor (and sometimes other factors). The main reason for this pattern is the feedback based RPM regulation of the DC platter motors. Basically, the control electronics continuously measures the RPM and when it is too low, the motor voltage is increased until the motor runs too fast, and then the system decreases the voltage. This usually causes sine wave shaped variations as seen here. All DC platter motors have this phenomenon. This is a common feature observed in all feedback based control mechanisms. The art of designing such control systems is to minimize these variations. But they cannot be fully eliminated.

At any rate this set of Beogram components is ready for duty again.

Beogram 4002 (5513): Complete Functional Restoration and Installation of a Beolover Commander Remote Control

This post describes the full functional restoration of a Beogram 4002 (Type 5513) that I recently received from a customer in Nebraska. My initial assessment of the unit is posted here.

This shows the Beogram with aluminum panels removed:

It was good to see that this unit was in original condition, which is the best starting point for a restoration.

As usual, I started with the platter motor since it can take several days until the platter bearings are re-infused with oil. This shows the motor after extraction:

I disassembled it to get the bearings out. They are the two small donuts on the black pad:

The pulley of this motor showed ugly belt tracks, suggesting replacement with a nice new Beolover replica pulley.

I immersed the bearings into motor oil and pulled a vacuum. Immediately, strong bubbling started:

The bubbling is indicative of the oil infusion process: As the vacuum draws the air from the empty porous bearing material, oil diffuses into the evacuated space.

While this process went on, I focused on the other restoration tasks.

First came the restoration of the mechanical systems, the arm lowering and carriage transport mechanisms. This shows the setup in its original condition:

I removed all the parts:

And cleaned them in an ultrasonic cleaner:

Then I re-assembled everything. I always install a new damper gasket. This ensures that the arm lowering speed is reproducible. The original gaskets are often hardened, which causes intermittent free fall arm drops, which is a hair-raising experience when there is a $800 cartridge mounted on the arm!...;-).

Unfortunately, the spindle nut holder showed a crack. This is a frequent issue with these Beogram models.

Luckily there is a Beolover replacement part:

I reinstalled all the components:

Then it was time to replace the tracking sensor bulb with a Beolover LED assembly. This shows the original black bulb housing in place:

I removed it, which revealed the aperture that is used for the tracking feedback:

As the arm is pulled inwards on a record, the aperture sends more light from the bulb onto a photocell located below the aperture. This is used to determine the appropriate carriage movement to follow the arm.

It is important that the aperture be parallel and close to the bottom part, without touching it.

This shows the Beolover tracking sensor light source in comparison with the original bulb assembly:

The SSMD LED is located at the same spot where the filament of the original bulb is lighting up. This shows the new LED light source installed:

The small white 'box' on top is a trimmer that allows to adjust the intensity of the LED. This can comen in handy for fine-tuning the tracking feedback. Here you can see it together with the new carriage spindle holder installed:

As usual, the original pulley had a hairline crack caused by stress from the set screw:

That is why I always replace this pulley with a Beolover precision machined aluminum replica:

Beolovely! The final step for restoring the carriage was cleaning and re-lubricating the pivot point of the damper-to-arm linkage. It is mounted on the sensor arm assembly. You can see it stick out from the small v-cut in the part that is bolted to the counter weight assembly:

For removal of the linkage one has to remove the sensor arm assembly. Before I worked on the linkage, I repaired the blue wire that goes into the sensor arm. Someone had cut it and then joined it back together in an less-than-professional way:

I removed the electrical tape, and replaced it with a proper piece of blue shrink tubing:

Then I worked on the linkage. Here you can see it in place on the removed arm assembly:

And after removal:

As usual, the small copper plate that ensures good lateral movement of the tonearm in up position was only loosely attached. A light tug with my tweezers pulled it from the degraded double sided tape:

I cleaned everything up and epoxied it back into place:

This concluded my work on the carriage.

The next step was rebuilding the circuit boards. I usually start with the main PCB. It has two power transistors mounted on the solder side. It is best to replace them before the board is removed for replacing the other components. This shows 1IC1, which is originally a TIP120 Darlington:

I replaced it with a TIP102, a stronger version, but same functionality. For some reason, modern TIP120 or 102 need a 100nF capacitor added to their emitters. Otherwise they can add a high-frequency oscillaton on the power rail, which often confuses the record detection circuit. This shows the new TIP102 and the capacitor in place:

After replacing 1IC4 with a TIP107, I removed the board:

This is a close up shot of the RPM control circuit with the 33/45 relay and the two RPM trimmers:

I replaced all electrolytic capacitors, power transistors, added a bias trimmer to 1TR3 and replaced the RPM section:

This shows a close-up of the restored RPM control circuit:

I usually put in 25 turn precision trimmers, which make it much easier to adjust the RPM. The relay was replaced with a Beolover Siemens style replacement relay.

Then I moved on to the output PCB. Since this is a 4002, there is not much on the board except the output relay and the delay circuit for it. This shows the board in original condition:

I replaced the output relay with the same Beolover replacement as was used for the main PCB:

I also installed a switch (red) that allows connecting system and signal grounds. This can be very useful if there is a hum when connecting to an amplifier.

The next step was to replace the remaining three light bulbs with LED assemblies. Two of the bulbs are in the RPM adjustment panel that is above the keypad. These bulbs provide the illumination of the dials. I removed the panel. This shows it from the back revealing the two bulb compartments:

I removed the covers:

The Beolover LED assemblies are shown in front of the panel, ready for installation.

I removed the bulbs and soldered the LED assemblies in:

They attach directly to the solder points used for the bulbs:

After installation, the bulb covers can be put back into place to maintain the original look:

The final bulb to be replaced is in the front of the sensor arm. This shows the small bulb/sensor compartment pulled out. The original bulb is still in place and the Beolover replacement LED assembly together with its alignment aid are seen to the right:

This shows the LED board installed:

At this point most components were already removed from the enclosure, so it was the perfect moment for cleaning up the mess of the degraded transport lock bushings. This shows one of them:

A 'beautiful' example of degraded bushings!

I removed the original main-capacitor assembly

and also the pinth (for straightening out the metal parts attached to it), followed by removing the floating chassis: 

I vacuumed all the bushing fragments out:

And then it was time to put everything back together. First, I installed a Beolover main-capacitor replacement assembly:

It replaces the entire dual-capacitance setup including the bolted on rectifier for the 30V rail on the output board. This pic interprets the original set-up:

This is the positive front end of the dual-capacitance can. The orange lead is connected to the 1000uF portion and the white lead goes to the 4000uF capacitor that is used to stabilize the main power rail on the main PCB:

This is a shot of the bolted on rectifier. the green leads are the AC inputs from the transformer, and orange and black are +30V and ground:

The installation of the Beolover replacement assembly is easy. Simply remove the original setup including rectifier

then solder the black and white and green wires to the board (the green ones are AC, i.e. it does not matter to which of the two AC input terminals they are soldered):

The next step is to place the alignment piece into the compartment around the screw hole of the original capacitor clamp

and place the Beolover board on it:

Then bolt it in with the original screw. It is a good idea to put a bit of grease on the screw since it needs to be tapped in a bit deeper that for the original setup. The grease helps the self-tapping process:

The final step is to solder the orange lead to the board. It is a good idea to wrap it around one of the capacitors to keep things organized:

Then it was time to install the new transport bushings. This shows the Beolover replacement bushings:

They are each assembled from two identical pieces which makes installation very simple. Another advantageous design property of them is that they are pretty thin. This leaves more leeway for adjusting the floating chassis. 

This is one of the orifices ready for receiving a bushing:

Simply stick in one half of the bushing from below

followed by the other half from the top:

This shows the floating chassis placed into its position with the transport lock screw protruding through the bushing:

After putting the floating chassis back in I placed the keypad back into the enclosure. As so often, the keypad had cracked RPM panel holders:

Both sides had the same crack:

Luckily there are faithful plexiglass reproductions available from the Beoparts-shop in Denmark. This shows two of then together with the broken originals:

All that needs to be done is to transfer the spring clips over to the new parts:

And then they can be bolted in:

Beolovely! Another issue with this keypad were the grease encrusted switch actuators on the keypad PCB:

Beautiful!...;-):

I removed them and cleaned them in an ultrasonic cleaner:

Then reinstalled them:

In my opinion these actuators do not need any lubrication. They are made from smooth plastic parts that have fairly large clearances.

Then I installed the restored PCB. It still needed adjustment of the bias of TR3, the transistor that amplifies the sensor signal. I usually install a 5MOhm trimmer to replace R26. This allows precisely adjusting the base bias to yield 4V DC at the collector as is specced in the service manual. The original R26 fixed value setup does not allow this. Hence, in the original setup the collector voltage varies depending on the particular gain of the installed transistor. I adjusted the bias to get the 4V

and then I moved the trimmer to the component side.

In the meantime, the oil infusion of the platter motor bearings had completed and it was time to reassemble the motor. I replaced the ugly original pulley with a brand new Beolover replica pulley:

Beautiful!:

The replacement pulley is crowned like the original ensuring that the belt stays away from the guards:

Then it was time to do a 24 hrs RPM stability measurement with the recently re-designed BeoloverRPM device:

This measurement (done in the 'slow' mode of the BeoloverRPM) yielded this 24 hrs RPM stability curve:

This result is pretty much as good as it gets with Beogram DC motors. They all suffer a bit from temperature drift and other issues. All this is the result of a 1970s style analog feedback control. An upgrade with the recently released Beolover SyncDrive DC motor replacement would improve the situation. SyncDrive is based on a modern digital control system and uses a brushless synchronous motor. It brings the wow and flutter down to AC motor Beogram level with a longterm RPM stability similar to later 8000 and 8002 models.

Additional measurements in the 'fast' mode of the BeoloverRPM device gives a more detailed insight into the wow and flutter of the DC motor platter drive:

These 'fast' measurements also show that the main reason for the measurement noise seen in the 24 hrs graphs comes from minute variations of the spacing of the platter ribs. This measurement shows that a repeating pattern is measured. This graph covers about 10 platter rotations (~20 sec) 

and it is clear that there is a repeating pattern. It repeats exactly every 24 measurements, corresponding to the 24 ribs that are on the Beogram platters. The BeoloverRPM basically measures the time it takes until the next rib passes under the sensor.

If we 'zoom out' a bit another repeating pattern is seen. The graph below covers about 60 platter rotations (120 sec). The superimposed wavy pattern corresponds to the wow and flutter introduced by the analog feedback circuit that keeps the motor running at a constant speed:

These curves look fairly 'dramatic' but one needs to keep in mind that the RPM scales of these graphs are very small, and that these RPM variations are in the <0.01% range, which is pretty much impossible to discern, even for the most advanced audiophiles.

After this excursion into RPM stability measurements it was time to focus on the remaining adjustments.

The most time consuming aspect of adjusting a Beogram 400x is getting the platter parallel to the arms and their movement path, and adjust the platter height properly. Then the platter needs to be adjusted to be flush with the surrounding aluminum panels. Once all this is done the arm lowering limit can finally be adjusted:

When the arm is lowered the tip needs to clear the lower sections of the ribs by about 1 mm. This ensures that the tip is not damaged should the record detection circuit ever fail and accidentally lower the arm without a record present.

After these adjustments I calibrated the tracking force. I usually remove the flimsy retaining washer that holds the screw loosely in place at the end of the counter weight assembly

with an M3 nut:

This allows locking the counterweight calibration into place able to survive the rigors of shipping.

This shows the result of my weight calibration efforts:

I usually try to get the scale on the dial to show the actual tracking weight around 1.2g, which is the weight most B&O cartridges want. The dial is notoriously imprecise and I recommend to re-calibrate the weight once in a while with a digital gauge to make sure.

The final adjustment was the tracking feedback:

Then I plugged in a Beolover Commander remote module:

The Commander allows full control of the deck with an apple remote control. This is the perfect way to prevent further keypad deterioration. The coating on these keypads is pretty fragile and it easily degrades when frequently touched. The acids and fats of the skin seem to react with the coating material causing the often seen smudges on these keypads. The keypad of this Beogram is still nearly pristine and it only has minor damage that can only be seen at certain angles. A perfect candidate for the installation of the Commander! The Commander also adds auto-repeat functionality similar to what the later Beogram 8000 and 8002 models offer.

Before it finally was time for a test spin of this deck I still had to replace the as usual corroded original DIN 5 plug:

This shows the new all-metal plug with gold plated terminals!

Beogolden!...;-).

And finally it was time for a first test spin! I selected an interesting album by MFSB, (Mother Father Sister Brother), which was a group of studio musicians at Philadelphia's Sigma Sound Studios. "Love is the Message" was recorded in 1973 on Philadelphia International Records (KZ 32707).

Of course this lovely album was ultrasonically cleaned on a CleanerVinyl ProXL setup to restore its original sound. A perfect combination with this restored Beogram 4002!

I will now play this deck a bit more to make sure there are no intermittent issues, and then it will be time to return it to its owner in Nebraska!