Beogram 4002 DC Platter Motor Restoration

I recently received the DC platter motor of a Beogram 4002 from a customer in the UK for restoration.

This shows the unit as received:

I disassembled it to get the shaft bearing out for an oil infusion under vacuum:

The bearings are the two small donuts on the black pad upfront. 

I immersed them in synthetic oil and pulled a vacuum. Immediately, strong bubbling started, making some foam on the surface of the oil:

After a couple of days, the bubbling stopped, indicating that the pores of the Oilite bearing material were filled again with oil. The vacuum pulls out the air from the bearing material, which enables the oil to penetrate. Once the motor runs, the oil slowly leaks out onto the shaft, keeping it lubricated until the bearing material is empty again. Most small electric motors have such bearings.

I reassembled the motor and installed a new Beolover DC Platter Motor Pulley for Beogram 4002 and 4004. 

Unlike some offerings on eBay, etc., Beolover pulleys have the proper crowning to keep the belt centered.

I installed the motor in one of my 4002s and ran a 24 hrs RPM stability test with my BeoloverRPM device:

In its 'slow' mode, the BeoloverRPM allows logging of the RPM in 10s intervals to the serial port of a computer. The ASCII data can be plotted with any standard graphing program. This is the curve I obtained for this restored motor:

This curve is as good as it gets with the DC platter motors. There is a little bit of thermal drift due to ambient temperature changes, commensurate with analog motor control systems.

This motor is ready for duty again!

In case you wonder why there is some 'noise' on this graph: This comes from the fact that the BeoloverRPM essentially measures the time between platter ribs passing under its sensor and then calculates the RPM. Since the platter ribs spacing is slightly varying around the platter due to manufacturing tolerances, the RPM measurements are slightly affected by these distance variations. This can be seen if the BeoloverRPM is used in its 'fast' mode, where it sends an RPM measurement to the serial port whenever a platter rib passes:

The screen shows a pattern that repeats every 24 measurements. There are 24 ribs on the platter... This is a curve measured in this mode:

The data in this graph corresponds to 35 platter turns. The repeating platter-related pattern is superimposed on a slower sine-wavy shape. This slower pattern corresponds to 'wow-and-flutter', i.e., short-term RPM fluctuations. These are mainly caused by the feedback-based motor control system that keeps the motor RPM constant based on constant RPM measurements.

Beogram 4002 (5513): A New Arrival from California - First Impressions

Today I started working on a Beogram 4002 (Type 5513) that was sent to me a while ago by a customer in California. It arrived safely packaged in a Beolover shipping container, and no shipping damage occurred. This shows the unit on my bench:

Like so many, it has a fairly scratched dust cover:

I removed the hood:

The aluminum surfaces are still pretty decent. The platter has a bit of scratching in the center:

I guess someone tried to install it upside down and missed the center pin. There are some other minor marks on it:

The keypad has the usual smudges in the coating from using it:

Luckily, these damaged 4002/4004 keypads can now be restored to a like-new appearance with a new keypad plate. See here for options if you are interested in getting your keypad fixed.

The wood plinth is toast. It has significant damage on the right front corner:

and should be replaced with a new, beautiful CNC-machined plinth made from solid wood.

I removed the aluminum plates and platter and had a look below deck:

It looks fairly 'original', but there was some previous human interaction, as is evident from the liberally spread black paint near the tracking sensor

and the platter bearing:

We will see soon how much of this mess I can remove.

The transport lock bushings are also gone. The locks are completely 'naked' around the orifices where the lock bolt goes through:

This will be fixed by installing a set of Beolover Transport Lock Bushing Set for Beogram 4000, 4002, and 4004. 

After this visual inspection, I plugged the unit in and pressed START. The carriage started moving, found the LP setdown point, and the solenoid engaged. The platter motor was also running.

All good signs for a restoration. I am confident this unit can be returned to like-new condition and performance. Stay tuned for my restoration report for this Beogram! 

Beogram 4000: Update to the Current 'Beolover State-of-the-Art' of a Unit I Restored A While Ago

A customer recently sold me his nearly pristine-looking Beogram 4000 due to an upcoming move to France, which required him to 'thin the herd'. I restored this unit for him about 10 years ago. It was the second Beogram 4000 I ever worked on at that point. In the meantime, I did about 25 more, and my process has evolved quite a bit. Therefore, to get this unit ready for sale, it needed quite a bit of updating to the current Beolover state-of-the-art.

This shows the unit after I was done:

Beolovely! Check the 'Restored B&O for Sale' section at the DKaudiolover store for currently available restored B&O.

Let's see what it took to get there:

This shows the unit as received on my bench:

It looked quite nice! The hood is a replacement made from laser-cut plexiglass panels glued together. So it has seams. After updating the Beogram, I replaced it with a new reproduction dust cover made with injection molding, like the originals (see below).

The unit has an almost pristine platter

and the keypad plates are in very nice shape without any dings around their perimeters:

The oak frame is also in very nice condition with sharp front corners on bothsides:

My customer broke the MMC cartridge tab off when he removed his nice original MMC20 CL cartridge for transport:

Sadly, this is a common issue with the Beogram 4000. The tabs of the 4000s were made from a stiffer, more brittle plastic compared to the later 4002 and 4004 models. When I restore a 4000 these days, I always replace the MMC mount with a Beolover MMC Cartridge Mounting Tab for Beogram 4000, 4002, and 4004 made from resilient nylon.

This shows the MMC20 CL cartridge with the stuck tab:

Luckily, it was easily poked out using the sharp corner of a razor blade:

Since the flexible printed circuit board that furnishes the electrical contacts usually remains on the arm due to weak 50+ year old glue, the stuck tab is a bit thinner than normal, which makes it easy to move it as long as there is some plastic sticking out where the razor blade can grab it. So my recommendation is: Pull the cartridge out a little bit before you break off the tab!...;-).

I stuck the cartridge on the Beogram 4002 (Type 5503) that I recently restored, and it turned out that it is in very good shape, putting out a very balanced and clear signal.

I removed the aluminum panels and had a look 'below deck':

It featured an early version of my main capacitor replacement assembly based on modern big cans and a 3D printed part to hold everything together:

During my update, I replaced this with a Beolover Efficient 24V Power Supply and Main Capacitors for Beogram 4000 (see below). It is a much cleaner setup, and the part also significantly reduces energy consumption and heat emissions.

I also found an early 'exploratory' version of my Beolover Tracking Sensor LED Light Source for Beogram 4000:

Ah, down memory lane!...;-).

Let's get to work! First, I focused on the mechanical aspects of the carriage:

I removed the red position indicator assembly, and then I removed the carriage

This shows it flipped over:

The removed parts had accumulated a bit of dust and patina since I did it last time:

I put them into my ultrasonic for a good cleaning.

A look at the solenoid-activated switches revealed that some corrosion had returned:

Back then I just learned how to do gold plating, and it seems my process was not yet perfect. Luckily, nowadays one can buy a set of contact terminals from the dksoundparts store in Denmark. Their kit allows replacing both the contact fins as well as the contact 'bridges' that form the other side of most of the contacts:

First, I replaced the solenoid switch that is responsible for reducing the solenoid current after it activates. The original one broke in two when I removed it back then, which I repaired by soldering a small piece of brass foil on top:

I removed the old part and installed a new replacement:

Beogolden! The DKsoundparts terminal fit perfectly!

Next came the muting switches (left) and the carriage motion enabling switch. Original:

And new:

Here is an impression of the oxidized original terminals together with the new replacements:

With the carriage 'up' it was the perfect moment to attend to the broken-off MMC cartridge mount. A look at the underside of the tonearm revealed a screw:

This meant that this Beogram 4000 had the original arm setup that makes its electrical connections with spring-loaded contacts that allow easy removal of the arm without the need to unsolder wires. Hence, the arm can be removed by simply removing the single screw that is at the end of the arm

Followed by pulling the arm forward, which reveals a small T-shaped plate with a hole for an excenter

The removed screw bolts the arm to the T-plate. The excenter allows adjusting the arm length, or rather the extent the arm sticks out towards the front. I pulled the arm off, which revealed the small circuit board that forms the male part of the spring-loaded contact scheme:

Which revealed a spring:

The removed arm with the screw:

The screw holds the part with the spring-loaded female contacts inside the arm in place. I removed it:

The insert inside the arm is usually glued in place. The glue can be softened by 'cooking' the arm in boiling water:

After cooking it for about 30 min I was able to push the insert out from the back by forcing it with a suitably sized plastic stick.

The female spring-loaded contact piece is on the right side in the above photo. This shows the part together with the Beolover replacement part:

I removed the lower part of the MMC tab end of the assembly, which revealed the original wiring:

I started transferring the wires to the Beolover part. First came the green one:

This shows all of the wires transferred:

At the contact terminal end I usually need to glue the metal parts into their proper places with some superglue:

The original adhesive has usually decayed, and the contact terminals are often loose. Then it was time to push everything back into the arm tube and put the screw back in:

Then I installed the new Beolover Tracking Sensor LED Light Source for Beogram 4000. This shows my old experimental part still in place:

I removed it, which revealed the tracking sensor aperture:

This shows the new part in place:

Next I tended to the arm lowering solenoid:

Their plungers are often magnetized, and that can cause erratic arm lifting behavior. A magnetized plunger is usually the culprit if the needle drags across the record during the automatic carriage return at the end of a record. I disassembled the plunger:

Then I tested the magnetization of the plunger with a small 3 mm ferrous set screw. As often, the screw got attracted by the plunger, a sign of existing magnetization:

After treating the plunger with my tape head demagnetizer, the set screw did not get attracted anymore, even at very small distances:

This shows the ultrasonically cleaned mechanical parts:

Before putting everything back together, I installed a new Damper Gasket for Beogram 4000, 4002, and 4004:

Next came the replacement of the plastic solenoid arm extension with a new Beolover Solenoid Arm Extension for Beogram 4000. This shows the new part together with the original setup:

It is a good idea to replace this plastic part since it often cracks at the rivet. This one was still o.k., but I replaced it anyway. If it breaks off, it can be fatal for the solenoid. Without the extension, the solenoid switch is not activated anymore when the arm lowers, and that causes the solenoid to burn up within a few minutes and magic smoke arises!...;-).

This shows the new part installed on the arm:

If you do this at home, always secure the attachment and the screw and nut with some epoxy in addition to the provided screw. This prevents the potential loosening of the extension after a number of solenoid activations.

Then it was time to put everything back together:

I also put in a new Beolover Aluminum Carriage Pulley for Beogram 4000 and a new carriage belt:

This concluded my work on the carriage

Next in line was rebuilding the carriage position detection PCB located under the carriage assembly. This shows the original setup:

It still had the original (red) solenoid resistor installed. Back when I restored it for the first time, I did not yet understand that these resistors are a bit underdimensioned and sometimes can burn out. I removed the plastic piston assemblies for ultrasonic cleaning:

While they ran in the cleaner, I replaced the oxidized contact terminals and installed a modern 2W resistor:

Beogolden! I really like these new contact terminals from the DKsoundparts store in Denmark! Awesome job!! I highly recommend buying them for any Beogram 4000 restoration. It is important that these contacts all work perfectly. This shows the restored board reinstalled with the cleaned plastic plunger assemblies back in place:

Then it was time to replace the light bulb in the sensor arm. When I worked on this unit the first time, I did not have an LED solution in place yet for this bulb. The bulb is revealed when the small compartment at the end of the sensor arm is pulled out:

In this early version, the 4000 sensor arm setup is also removable without soldering. This shows the unplugged compartment together with the Beolover Sensor Arm LED Light Source (Beogram 4002, 4004 and 4000) components:

And after installation:

This shows the insert plugged halfway back into the sensor arm tube:

This concluded my work on the carriage.

Next, I removed the AC platter motor and the big reservoir and motor capacitors. This shows the setup I created during my first restoration of this unit:

After removal:

This shows the removed AC motor:

I drilled out the rivets and took it apart:

These motors often have dry Oilite shaft bearings, causing excessive 'knocking' when the motor runs. The remedy is to re-infuse the bearings with fresh oil under vacuum. As far as I can tell, there is no way to remove the bearings without causing some significant damage to the motor. Therefore, I usually resort to submerging the entire motor housing with the installed bearings in oil.

This shows the parts right after pulling a vacuum, showing some initial air bubbling from the bearing in the center:

After a few minutes, it usually looks like this:

The oil starts foaming as the air is drawn out of the bearing and some other pores in the motor housing. Once the bubbling stops, the bearings are ready for duty again. In my experience, this can take up to 3 days.

While this process was underway, I focused on some of the other remaining restoration tasks. I unbolted the main PCB and flipped it up. This revealed the transistors of the platter motor push-pull stage. They are bolted to the enclosure bottom, which serves as their heat sink:

This unit still had its original TIP31/32 matched pair installed. When I worked on it the first time, I did not yet understand that these transistors are prone to failure. Depending on the motor current dialled in, they can get a bit warm. Therefore, I now always replace them with their stronger cousins, TIP 41/42. It is convenient to unbolt the transistors and temporarily bolt them to one of the mounting posts of the PCB for unsoldering the leads:

Soldering the leads to the replacements is also best done while they are mounted to the post:

This shows the new TIPs bolted to the enclosure:

Then I updated the main PCB. This shows it flipped over to reveal the component side:

Back then, I still did not replace the power transistors. The board also had a preliminary homemade version of my Siemens-style replacement relay. I replaced the transistors and also installed the current version of the relay:

Next I replaced the sole power transistor (silver can) on the small power supply board next to the transformer:

Next I focused on updating the keypad cluster. It is held in place with a single screw:

After removing the screw and pulling the assembly up, it is a good idea to take a photo of the way the wiring harness is folded underneath:

This helps with getting the keypad back in properly later on. This shows the removed keypad assembly in service position next to the Beogram:

It had an early version of my carriage position indicator illumination LED boards installed:

Back then I used red-green LEDs tuned to 'incandescent color' with appropriate resistors. I abandoned this practice once I realized that these multicolor LEDs can have varying output intensities between different batches. This caused trouble since the light of newly manufactured batches could be too red or too green, and the resistor values had to be manually adjusted to get the proper light color.

Nowadays, I use white LEDs with glued-on color filters. This shows a set of the Beolover Scale Illumination LEDs for Beogram 4000 installed:

The above picture also features a new set of Beolover RPM Trimmer Illumination LEDs for Beogram 4000. I now use different LEDs and resistor values for slightly stronger brightness so that the selected speed can be discerned in a normally lit living room. The original bulbs are very dim.

Next came replacing the oxidized contact terminals of the keypad switches. This shows the terminals before

and after exchange:

Beogolden!!...;-). Note that the new symmetrical contact 'bridges' seem to be slightly different from the original ones in terms of vertical dimensions. I had to push them out slightly to get the <</>> keys to work properly:

A small complaint! On the positive side, this tolerance allows adjusting the sensitivity of the fast <</>> functions by adjusting the gap.

I also extracted the plastic plunger assemblies

and put them through the ultrasonic cleaner. They work best when they are clean!

While the keypad was removed, it was the perfect moment to replace the cracked original plinth guidance washers. This shows one of the original washers:

This shows a set of Beolover Plinth Guide Washer Set for Beogram 4000, 4002, and 4004:

I installed them:

All good in the plinth department again!

In the meantime, the oil infusion process of the AC platter motor bearings had concluded. I broke the vacuum and removed the clamshell housing from my oil vat. After thoroughly cleaning off the excess oil, I put the motor back together:

I use (red) 3D-printed parts that hold M3 nuts in place for the tilt adjustment screws (to replace the drilled-out threaded rivets).

This shows the motor installed together with the new Beolover Efficient 24V Power Supply and Main Capacitors for Beogram 4000 board that replaces the big capacitor cans. This design also replaces the original wasteful linear voltage regulator of the 24V rail with a modern, energy-efficient buck converter-based design. This made this Beogram run much cooler, and it also significantly reduced the transformer current, increasing its life expectancy.

At this point, the Beogram was pretty much back together. So I plugged it in and turned it on to adjust the sensor arm transistor bias to get 1.8V at the collector:

After this adjustment I moved the (blue) trimmer to the component side:

Then I adjusted the platter motor amplitude and frequency. These oscilloscope traces were measured on the purple motor lead. First 33.33 RPM:

and 45 RPM:

Very nice sinusoidal signals! This platter motor was running very nicely!

Then I soldered a measurement lead to the collector of the sensor arm transistor and measured the sensor signal. This is always a good idea whenever the sensor bulb is replaced to make sure the sensor works properly. This is the curve I saw:

In the Beogram 4000 the signal needs to be at least 2.2V. This unit exceeded this value nicely with a 4.4V peak-to-peak amplitude. This sensor is in excellent working order!

Next, I put a belt on the platter and ran a 24 hrs RPM stability test with my Beolover RPM device:

The BeoloverRPM features two different operating modes. In "slow" mode (pictured above), the device takes an RPM reading every 10 seconds and outputs the data via serial connection to a computer. Any standard terminal program can capture the stream, making it straightforward to plot long-term RPM trends in Excel or a similar spreadsheet tool. Here is the data from a 24-hour run:

That is a very nice, stable result, but nothing surprising for AC platter motors. Since they are synchronous motors driven by a Wien oscillator, they are 'naturally' stable. 

Following that run, I logged higher-resolution data using the BeoloverRPM’s "fast" mode:

In fast mode, the unit transmits an RPM value each time one of the platter's 24 ribs passes the sensor, yielding 24 discrete measurements per full revolution. This lets us observe immediate, short-term speed variations ('wow and flutter'). The plot below covers roughly 35 revolutions (around 70 seconds):

The distinct zig-zag pattern visible in this curve is actually an artifact caused by slight manufacturing variations in rib spacing around the perimeter of the platter. Because those spacing changes repeat identically with every revolution—like a mechanical fingerprint—they ride directly on top of the actual RPM shifts produced by the platter drive. The genuine speed variations correspond to the superimposed, slight sine-like wave modulating that zig-zag. Analyzing the wave reveals an estimated wow and flutter of around 0.1%.

While that figure is double the 0.05% factory spec listed in the original service manual, the discrepancy is largely systematic. The testing methodology used during the 1970s was completely different: technicians played a test record with a steady tone and fed the signal into an analog spectrum analyzer to calculate wow and flutter from center-frequency drift. Moreover, the distinction is largely academic, as human hearing generally doesn't detect pitch variations below 0.7%. Whether the true figure sits at 0.05% or 0.1%, this Beogram falls well below that threshold. This motor is ready to go back into service!

After completing these tests, I started the process of aligning the arms, platter, and floating chassis to get the platter parallel to the arms across their travel, and the platter flush with the surrounding aluminum panels. This can be a slightly tedious iterative process that takes a few passes to yield proper alignment.

After all this was in place, I did the arm adjustments. First came the adjustment of the arm lowering limit. When the needle is down the tip should be ~1 mm above the lower rungs of the platter:

 Then I checked the weight at the 1.2g weight dial setting:

My calibration seems to have survived the decade since I worked on this unit pretty well!

At this point, I thought I was pretty much done. So I carried the unit over to my listening station and plugged it in. I put on a record and pressed ON. Everything occurred according to plan, but when the music started, I realized it was only playing on the left channel. On a hunch, I wiggled the output cable a bit at the cable gland where it exits the enclosure. And bingo, I was able to restore the channel, but only very briefly. So it turned out that the cable had been pulled with some excess force at some point, and that must have ripped the signal wire for the right channel apart.

Back to the workbench! These days, pretty much the only option for finding a suitable cable is to get a decent DIN5 male-to-male jumper, remove one of the plugs, and then install it. I still had a few DIN5 cables from Sounds Heavenly in the UK from when Steve still shipped overseas. So I grabbed one and removed one of the plugs:

It turns out that Steve seemingly also had a hard time finding cables set up like the originals with an external shield. 'They' simply do not make them anymore like this. He settled instead for a 4-channel RCA-type cable with 4 signal leads immersed in individual copper braids. These copper braids were not insulated from each other, unlike in the original Beogram cables. The workaround with such cables is to use one or two of the signal leads as signal return for the two stereo channels and use the copper braid as the outer shield. This shows the original cable together with the prepared-for-installation Sounds Heavenly cable:

I used the red and white leads for the signals, the green and yellow leads as common signal return, and the copper braid as the outer shield.

This shows how the original cable was connected to the output terminals:

The grey leads are the left channel, the black/red wires are the right channel, the combined green and yellow wires are the signal return, and the black/braid wires are the enclosure/cabloe shield.

This shows the corresponding connections with the new cable:

I carried the deck back to the listening setup. I selected one of my favorite CTI label records, Stanley Turrentine "The Baddest Turrentine" (CTI 6048). This album came out in 1974, a perfect contemporary companion for this lovely restored Beogram 4000! Here is an impression:

What great cover art! When men were still men!...;-). The sound is also pretty awesome. Rudy van Gelder was the recording engineer! The Beogram played it perfectly! A happy day for the Beolover!

After this successful test, it was time to install a new reproduction dust cover with new aluminum trim.

The hood this Beogram came with was a state-of-the-art replacement about 10 years ago when I restored it. It had been provided by Classic Audio in Denmark together with a laser-printed plastic trim that looked like aluminum from a distance. Back then, these hoods were made by cutting plexiglass sheets with a laser and then gluing the panels together. This looked pretty good from a distance. This shows the removed hood:

But closer inspection immediately revealed the seams from the gluing process:

The labels on the plastic trim were a bit too bold

This shows the current state-of-the-art. The dust covers are now made by injection molding, exactly like the originals! Another great 'parts coup' by the DKsoundparts store! The replacement trim is now also made like the originals from brushed aluminum strips.

Whenever I install one of these nice new dust covers, I first check if there is any interference between the plastic and the metal hinge assembly when the hood moves on the hinges. The design of these hoods is a bit 'tight' at the back end where the stabilizer bar is (to prevent sagging in the center between the hinges). So I always bolt them in for a quick test. This hood passed barely, but there was still a sufficient gap:

If the hood chafes, it will break at some point due to the strain caused by the interference.

Once I was certain this dust cover was ready for bolting in, I glued small plastic bits into the strange channels they designed into the recess for the aluminum trim in the back of the hood:

After the epoxy was hardened, I trimmed the bits to match the plexiglass edge:

These plastic bits make it possible to bend the aluminum trim tightly around the edge. Without the bits, there is a hollow, and the trim will invariably make a round bend instead of a sharp one.

The next step was centering the new aluminum trim. I do this with my 3D-printed alignment aids:

The blue tape in the center is used to make an alignment mark once the strip is centered

With the strip alignment established, it was time to bolt the hood to the hinges. I usually put a bit of white wood glue onto the threads to fix the bolts in place. One is not supposed to tighten them very much so as not to stress the plexiglass too much:

With the hood bolted in, I removed the backing from the adhesive and placed the trim, minding the alignment mark precisely:

Then I used my bending block to bend the sides down:

Beolovely!:

And here the final result:

I bolted the hood assembly to the restored Beogram 4000 and carried it over to my photo studio (locally also known as 'garage'...;-).

Here are some of the pictures I took. Enjoy!: