Beogram 4000/4002/4004: Test of New Custom Manufactured Beolover Carriage Motor

Sometimes carriage motors in Beogram 4000, 4002 and 4004 can be pretty noisy. Most likely their bearings are also in need of an oil infusion under vacuum, like the DC platter motors. Unfortunately, both the original Faulhaber and Maxon motors cannot be opened easily. So the remedy until now was a drop of oil on the shaft and hoping that this would do the trick for some time. After all these motors run much less than the platter motors.

But I always thought this was a less-than-Beolovely solution. Therefore, during the last couple years I tried out a variety of small motors with similar footprint but had to find out that all standard motors were way too noisy and vibration generating. Only after I sent one of the original motors to the supplier of the motors that I use in the Beolover SyncDrive for evaluation it became clear that the carriage motors used in the Beogram 400x are something special:

They are so called "coreless DC motors". Such motors have a rotor constructed as a hollow, self-supporting coil, often in a basket or honeycomb pattern, which eliminates the need for an iron core. This results in reduced mass, reduced eddy currents, lower inertia, and smoother torque/lower vibrations. The result is a motor that runs much more quietly and draws less current compared to normal motors with iron cores for the same performance (but that also has a significantly higher price - you get what you pay for!...;-). The quietness is crucial for driving the carriage since the slightest vibrations on the floating chassis can translate into sound audible in the speakers since they directly couple into the pickup. During my trials of standard motors the carriage drive was quite audible while listening to music at moderate volumes.

My supplier was able to design a modern coreless replacement that has a torque and voltage that is close to the original motors. I had a few samples manufactured and recently received them.

Here is a picture of the new Beolover carriage motor fitted with a custom pulley::

These motors are now available at the Beolover store.

The motor is easily mounted in the original motor enclosure using two EPDM rubber O-rings that help dampen any vibrations the motor may emit and also align it with the enclosure:

Since the new motor is a bit shorter than the original Maxon and Faulhaber motors, the leads can be fed though the cutout in the bottom of the enclosure:

This makes for a nice and clean installation:

It is a good idea to replace the original usually wobbly plastic pulley on the carriage spindle with a new precision machined Beolover aluminum pulley when exchanging the motor.

I performed some measurements that compare an original Maxon motor in one of my Beogram 4002s with the new motors. These oscilloscope traces show the voltage at the motor terminals while playing a record. The first curve corresponds to the original motor:

And the second one to the new Beolover motor:

The oscilloscope settings were the same for both measurements. We see that the new motor responds at a lower voltage, which corresponds to a smaller deflection of the tonearm necessary to get the motor to move. This is most likely a result of the lower friction inside the motor due to its fresh bearings that provide optimal lubrication. The consequence is that after implantation of the new motor the tracking feedback may need to be adjusted slightly to get the specified sensitivity. If a Beolover Tracking Sensor LED Light Source has been already installed in the tracking sensor, this can be achieved by simply reducing the light intensity of the LED slightly by adjusting the built in trimmer.

An important factor in motor replacements is the current draw under identical operational conditions. The H-bridge on the main board already struggles a bit with the current consumption of the original carriage motors, which is evident in the occasional need to replace the power transistors, especially those providing the current for the travel direction towards the home position (>>). So I definitely did not want replacements that would draw more current than the originals.

These are my measurement results for the two motors:

The only difference between the two columns is that the original motor was replaced with a new one. All other parameters like the <, > speed trimmers were kept the same.

It is clear that the new motors draw considerably less current in the same setting. Their performance however is very similar to the original motors as can be seen from the time it takes to complete a full platter sweep. This was done with an empty platter and measuring the time it takes between pressing Start and the moment the deck shuts off after the carriage came home again. The original motor did it in 18.0 seconds while the new motor was slightly faster at 17.5 sec. It should be noted that these absolute numbers will be somewhat different in every Beogram due to friction variations caused by adjustments and lubricants etc...But it is clear that the new motors will result in much less heat load in the H-bridge.

Another happy aspect of the new motors is that during these tests it became apparent that the new motors run noticeably quieter than the original ones. Which may have to do with their more modern design, but also with their fresh bearings and better lubrication.

In summary, it appears that the Beolover carriage motors are an easily installed cure for noisy original motors.

Beogram 4002 (5524): Full Functional Restoration, Installation of Upgrades and a First Spin with Harry Thumann

This post describes the restoration work performed on a Beogram 4002 (Type 5524 with voltage selector) that I received a while ago from a customer in Singapore. There is an earlier post that describes the initial condition of the unit. This project was a bit special in that there was an interesting fault in the electronics (see below and here). It also had a broken off transformer mounting tabs, which I was able to fix with a new Beolover part. This deck also received the Beo4 enabled Beolover Commander remote control system as well as the Beolover internal RIAA pre-amplifier board.

My customer already had installed the Beolover SyncDrive DC motor replacement in an attempt to cure the usual RPM inconsistencies of these decks. Most of the original DC platter motors suffer from dry Oilite bearings these days, which causes the RPM to fluctuate intermittently initially and more permanently later in the cycle. The SyncDrive is a plug and play cure for this issue. You can see it installed on the left side next to the circuit board in this picture:

The original platter motor was included in the shipment, too. I immediately noticed that the pulley was installed upside down, indicative of a precious repair attempt:

I decided to try restoring the motor. After removing the outer housing I saw that the tabs of the upper bearing mount were bent up:

I opened the motor up and found a washer normally located below the commutator installed on the top end of the shaft:

I put it back to here it belonged:

Then I disassembled the motor to extract the two bearings. They are on the black pad up front:

I immersed them in ISO 50 synthetic oil and pulled a vacuum:

Immediately strong bubbling occurred foaming up the surface of the oil. The bubbling is a sign of air being drawn from the empty bearing pores. This makes room for oil to diffuse into the material. 

This process can take a couple days. While this was underway I focused on the remaining tasks of the restoration.

First came the cleaning and re-lubricating of all the mechanical moving parts responsible for driving the carriage and lowering the arm. This shows the setup in its original condition:

I extracted all the parts for ultrasonic cleaning:

This shows the liberated carriage:

I usually place it on a soft pad to protect the fine wiring on its bottom side. Next I removed the solenoid to check on the magnetism of the plunger. This shows the plunger removed:

I usually test it with a small ferrous set screw:

A magnetized plunger will attract the screw and it will roll towards it. In this case the plunger was slightly magnetic. When the magnetism gets too strong it can delay arm lifting during carriage return at the end of a record. It is no happy sight if a $800 cartridge is dragged across the record for some distance until the solenoid spring finally wins against the magnetic attraction between plunger and solenoid mounting bracket. I demagnetized the plunger until the set screw showed no longer any attraction to it.

In the mean time the ultrasonic cleaner had done its thing. This shows the clean parts ready for re-installation:

As usual, I installed a new damper gasket:

It is a great idea to do this since the original gaskets are often hardened which can cause inconsistent arm lowering speeds. In bad cases the arm occasionally just drops without any damping, which can be a hair raising experience!

After putting the parts back in I also installed a new carriage pulley. When I put the belt on I realized that the carriage motor was completely loose in its housing. This usually means that the foam pad that dampens its vibration has degraded. In this case it basically had turned into a fine powder:

I cleaned the mess up and installed a new polyurethane foam pad cut to size and glued in place with double sided tape. Hopefully this will last for the next 40-50 years...;-):

This shows the new precision machined pulley in place:

The final mechanical part needing attention was the linkage between damper and tonearm assembly. Here you can see it sticking out from the V-cut in the small arm bolted to the back of the counter weight:

To get to the pivot point for cleaning and re-lubrication the sensor arm needs to be taken out:

After removing the locking washer the linkage can be removed:

As usual the small copper plate that helps the tonearm moving laterally when up came loose after only a weak tug with my tweezers. I cleaned the degraded double sided tape off

and glued it in place with a dab of epoxy:

After re-installing the sensor arm it needs to be properly aligned to be perpendicular to the rods the carriage travels on while maintaining the proper distance and parallelism between the arms:

My next task was the restoration of the PCBs. I started with the main board. It is best to replace the two power Darlingtons mounted on its solder side while the board is still in place. This makes their alignment straight forward. This shows the original IC1 in place:

I usually replace these with a stronger type, a TIP 102:

The small yellow package is a 100nF capacitor between emitter and ground. It is necessary since these modern TIP Darlingtons tend to enter a high-frequent oscillatory state on their output in this circuit configuration. This can throw off the record detection mechanism. Adding this capacitor cures the issue and stabilizes the power rail properly.

I also replaced IC4 with a TIP107 and then removed the board. This shows it extracted with the components up:

A detail shot of the 'RPM section', composed of a Siemens RPM relay and RPM adjustment trimmers:

I replaced all electrolytic capacitors, power transistors and the RPM controlling components:

This shows the rebuilt RPM section with a Beolover Siemens style relay replacement and modern encapsulated 25 turn trimmers for RPM adjustment:

Next came the output board:

This board contains the output relay and the delay circuit that ensures that the relay remains closed until after the needle touches down in a groove:

I replaced the electrolytic capacitor and the relay with another Beolover Siemens style relay replacement. I also installed a switch (red) that allows connecting signal and system grounds in case a hum is experienced (mainly when connecting the deck to RCA inputs via an adapter)

At this point the enclosure was already almost empty. What was left to remove was the original reservoir capacitor. In this case it was a single capacitance unit:

This shows the empty enclosure

As usual with this vintage 4002 the transport lock bushings had disintegrated ad their fragments were strewn around the enclosure: 

I vacuumed the mess:

Now it was a good moment to address the broken out transformer. Sadly, some later DC motor 4002s were fitted with transformers enclosed in plastic housings. Obviously their routing tabs are not strong enough to withstand the rigors of transport. In this case both tabs had broken off and the transformer could be lifted out easily:

I had the same issue with another 4002 that I restored earlier this year. To fix it I developed a mounting bracket that fit the transformer shape:

It basically fits over the transformer housing and holds it down via two screws that bolt into the original mounting holes in the enclosure bottom. For installation the remnants of the broken off tabs need to be cut off with a carpet knife or similar:

After trimming the tabs the bracket fits precisely. Here some impressions of the nice clean fit:

Use the original wire clamp to attach the power cord to the transformer corner:

Beolovely!

After this was done, I installed a new Main Reservoir Capacitor for Beogram 4002 and 4004:

This capacitor assembly can be used for replacing both single and double capacitance reservoir capacitors found across the DC motor Beogram 4002/4004 line.

I also replaced the cracked output cable clamp with a Beolover Lead Holder and Output Board Bracket for Beogram 4002 and 4004:

This shows original cracked and new clamps next to each other:

And with output board installed:

The 3D printed ABS part is more stable than the original cracking prone plexiglass part.

During the installation of the floating chassis I replaced the disintegrated transport lock bushings using a new Beolover Transport Lock Bushing Set for Beogram 4000, 4002, and 4004: 

My bushings come in two parts with makes installation very simple. Just put in one half from the bottom:

And the other from the top

And that is it:

When I put the keypad assembly back in I realized that both fixtures that clamp the RPM panel down were broken out or had already cracks:

This frequently happens when the RPM panel is not removed in the proper way. Luckily there are replacements available from the Danish Sound Parts store: This shows old and new next to each other:

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

And then they can be bolted in:

Beolovely!

After I put in the main PCB I adjusted the bias of the sensor transistor so it would have the specified 4V at the collector:

This is important for getting a proper signal when no record is on the platter. After setting the trimmer I installed in below deck on the component side:

It is always a good idea to check the sensor signal after working on the circuit. I hooked up my oscilloscope and measured this trace at the collector of TR3:

A pretty strong signal! Definitely a passing grade! Every dip corresponds to a black platter rib passing under the sensor.

The replacement of the incandescent light bulbs came next. First I replaced them in the RPM panel where they are used for back illumination of the RPM adjustment scales. This shows the board flipped over:

The two bub covers need to be removed:

Then the Beolover RPM Panel LED Backlights for Beogram 4002 and 4004 can be soldered in. They directly solder to the pads that held the bulb wires and can be extracted as board extensions:

This shows a detailed picture of one of them:

These LED boards do not obstruct the bulb covers in any way and they can be installed again after the transformation:

The third bulb is in the sensor arm. This shows the small compartment pulled out from the end of the arm together with the Beolover Sensor Arm LED Light Source and its alignment aid:

And after replacing the bulb:

The fourth and final bulb to replace was in the tracking sensor. This shows the original bulb housing still in place:

I removed it. One of the mounting tabs had broken off. This shows old bulb and and the new Beolover Tracking Sensor LED Light Source in direct comparison:

Here a glimpse at the aperture that adjusts the amount of light that falls on the photo resistor in the lower part of the tracking sensor assembly:

The LED assembly directly replaces the original bulb housing:

After this I tested all the functions of the Beogram. It became clear quickly that not all was well in the electronic control department. Whenever I pressed START on the keypad the arm lowered immediately and the carriage stopped when I let the key go. Both behaviors are not normal. The carriage should move after pressing Start (<<) until <, >, or up/down is pressed. The same behavior occurred when pressing STOP (<>). I spent an afternoon figuring out the root cause of this issue. It turned out to be a bad spark snubber diode 8D2 on the output board that subsequently killed 8IC1 on that board. The dead 8IC1 Darlington is responsible for driving the output relay. This shows the measurement of the diode with my transistor tester. It essentially had become a resistor:

Without a working  diode the Darlington cannot survive the magnetically induced voltage spike that occurs whenever the relay could turns off. This post explores this unusual fault in more detail. This shows the new components implanted:

After fixing this I thought I was done with operational issues, but it turned out that < and > sometimes worked and sometimes not. Intermittent issues are often caused by bad connections in these Beograms. In this case it turned out that the board header that takes the keypad plug had cracked solder joints:

I re-flowed the solder points and that solved the issue:

Now it was finally time to do all the adjustments. In these adjustments the arm travel across the platter needs to become parallel, the platter needs to be in the proper distance from the arms, and the platter needs to be flush with the surrounding aluminum panels. This can sometimes be a pretty iterative slog. Due to the little space the floating chassis has for movement there is not much room for 'improvisation' and everything needs to be perfect.

After this process was completed I did the arm adjustments. first I replaced the flimsy locking washer that holds the counterweight screw in place

with a square nut and a washer:

This allows locking the tracking weight calibration in place, so my calibration survives the rigors of the shipping process. Then I did the arm lowering limit adjustment. The arm should stop lowering that there is still about 1 mm space between the needle and the lower rungs of the black platter ribs:

This is a fail safe in case the arm should ever be lowered onto an empty spinning platter due to a record detection circuit failure. The next step was to calibrate the tracking weight:

I usually try to get it adjusted in a way that the weight scale on the small adjustment wheel is accurate around 1.2 g, which is the weight that most B&O cartridges prefer.

Then I adjusted the tracking sensor feedback:

There was one more electrical issue to be fixed: The power LED had a broken off lead:

This is a common failure with these decks since the leads that are soldered to the LED are not protected by shrink tubing. I installed a new 3 mm red LED with a 20k resistor in series to lower its intensity to something similar like the original LED produced:

Modern components are often more powerful in their performance than original vintage parts.

After this I checked on the fuses. It is always a good idea to do this with these old decks since fuses degrade over time. Often their metal ends are not fused anymore to the glass body and as soon as one takes them out they disintegrate. In this case one of the fuses was already burned out:

This means this deck had suffered an 'event' before I received it and did not work anymore at 220/240V.  But it worked on my bench since I switched it to 110V, which uses both main sets of primary windings in parallel. If one set of windings is disconnected via a broken fuse the current doubles in the other winding but the deck still works, all else equal. Of course the remaining winding gets a bit hotter than it would normally since it now handles all the current drawn by the Beogram alone. 

I installed two new 250mA T fuses:

All good now in the transformer department!

At this point the bubbling around the motor bearings had stopped and I extracted them from the oil:

After re-assembling the motor I performed a RPM stability test with the BeoloverRPM device. The BeoloverRPM is able to log the RPM in 10 s intervals for extended periods of time:

This is the curve I measured after about 24 hrs:

This is a pretty decent result for a DC platter motor. With these motors there is always some long-term drift, usually caused by thermal effects, especially now in winter when I need to run the heater in my workshop during the day. The slight choppiness of the curve is related to the the top bearing of the motor being polished by the shaft as the motor drives the platter. Since the re-infused bearing usually sits in a somewhat different orientation compared to where it was prior to extraction shaft and bearing have to get 'used' to each other again, which can cause some small changes of the RPM. This effect usually vanishes after playing a bunch of records. At any rate the changes are much too small to be discernible by most humans.

The BeoloverRPM has a second measurement mode ('fast') where it logs a measurement every time a platter rib passes under its sensor. Such curves show a repeating spiky pattern which is a measurement artifact. The peaks do not relate to actual RPM changes, they rather reveal the manufacturing tolerances during production of the platters. The black ribs are not entirely equidistant around the platter and this translates in the detection of 'RPM variations'. This shows such a curve reflecting about 60 platter rotations (=~120 sec of measurement):

The sine-wavy pattern superimposed to the zigzag 'platter pattern' represents actual RPM changes related to the DC motor feedback process that keeps the motor spinning at a constant RPM. The variations are about 0.1%, which is a typical result for Beogram 4002 DC platter motors.

Finally, I replaced the as usual oxidized DIN5 plug

with a nice new all metal plug with gold plated pins:

Beogolden!...;-). 

Aside from putting in the Beolover SyncDrive again, my customer asked me to also install the Beolover Commander remote control system for Beo4 and Beoremote One remotes:

The Commander is perfect for protecting keypads from further deterioration. It can control all functions of the deck and adds auto repeat functionality.

I also installed the Beolover internal RIAA pre-amplifier board:

The Beolover internal RIAA pre-amp is perfect for integrating a Beogram 4002 or 4004 into more modern systems where the amplifier does not have a dedicated phono input anymore. The installation is completely reversible. The board simply replaces the original output board. It can also be configured to act as a standard output board by simply plugging the in- and output plugs into the respective sockets on the board.

The final result:

This picture shows all the currently available Beolover upgrades together! Beolovely!

On to a first test spin! I selected a recent acquisition: "American Express" by Harry Thumann. This somewhat obscure record was recorded in 1979 on an Italian label, Baby Records. The disco hit on it is "Underwater", which was used in the video game Grand Theft Auto in 2008. This made the record popular again, and that is how I found it: For some mysterious artificial intelligence reason, Spotify suggested a playlist based on the GTA game. I curiously clicked on it, and it turned out that this was a pretty cool music mix containing a lot of tracks from the 1970s and 80s I often listen to, but also Underwater, which I did not know. I liked it so much that I went finding a vinyl copy of it. Mine is a release from Canada that was probably pressed in 1980. Anyway, here an impression of this lovely album (the hair!!...;-) together with this fully restored Beogram 4002:

I will now listen to this deck for some more days and if nothing additional comes up it will be time to send it back to my customer in Singapore!