Beogram 4002 (5513): Complete Restoration and Installation of the Beolover Internal RIAA Pre-Amplifier

This post describes the work done during the restoration of a Beogram 4002 (Type 5513) destined for a customer in Australia (that is the reason this unit will receive a universal power supply that it can be plugged into any outlet at any voltage). My initial assessment of this unit is posted here.

Here is a picture of the final result of my efforts:

This unit is a bit special since my customer had me install a brown tinted hood as they were used on the Beogram 4000. The later 4002 and 4004 models received grey tinted hoods. I think my customer is really on to something with his choice. It looks pretty awesome together with the new CNC machined 4000c style solid walnut frame. A great advantage is that these earlier hoods are more transparent and so the beauty of the Beogram is also apparent when the cover is closed:

Allright! Enough ogling, let me discuss what it took to get there:

This shows the unit on my bench in original condition with the aluminum plates and platters removed:

As usual with DC motor Beograms I first removed the platter motor and disassembled it to get the shaft bearings out for oil infusion:

These bearings have a limited amount of oil in their porous material and they all run dry at some point. Essentially such motors have a finite life time, which runs out when the bearings lost all their oil. 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 indicating air being drawn from the porous bearing material:

The oil re-infusion process can take 2-3 days. It is complete when the bubbling stops, i.e. all air has been removed from the bearing material and the pores are full with oil again.

While this process was going on I focused on the remaining restoration tasks. First came the cleaning and re-lubricating of the moving parts around the carriage. This shows everything before I took it apart:

Lots of old hardened lubricants! I removed all the parts. While the carriage is 'loose' it is a good idea to place it on a soft pad to protect the fragile wiring at its bottom:

The solenoid of this Beogram looked suspiciously 'cooked'! The bubbled up insulation material indicated an overheating event:

I measured 1.4 Ohms as resistance and it was obvious that the wire insulation was damaged and some of the windings were in direct contact. A good solenoid coil measures about 9 Ohms:

This solenoid coil had to be replaced with a new Beolover Solenoid Coil Replacement for Beogram 4000, 4002, and 4004. I removed the plunger and checked its magnetism. As so often it was able to attract a small ferrous setscrew:

I used my tape head demagnetizer on the plunger until it no longer attracted the setscrew. This is an important step of a Beogram restoration since a magnetized plunger can cause delayed arm raising. During auto return at the end of a record this can result in the needle scratching across the record while the carriage returns home.

This shows all the removed components ready for the ultrasonic cleaner:

After 30 min in the cleaner they looked much nicer!:

I coated the cleaned carriage rods with Tactikel, a nano-particle based fluid that not only increases the slickness of gun barrels, but also of carriage rods for smoother carriage motion..;-):

I also installed a new damper gasket:

The original gaskets are often hardened and/or deformed. This can result in occasional high-speed arm drops, an hair raising experience when a newly rebuilt $800 cartridge is installed!...;-).

At this point I also replaced the incandescent bulb in the tracking sensor with a new Beolover Tracking Sensor LED Light Source (Beogram 4002 and 4004). This shows the original black bulb housing still in place:

I removed it, which revealed the tracking aperture:

This aperture permits more or less light on a photoresistor in the bottom of the assembly, which informs the carriage motor about how much to propel the carriage forward during record play. This shows new and old in direct comparison:

The small LED of the new assembly sits in the same spot where the filament of the original bulb is located.

This shows the part installed:

I set out to replace the solenoid coil. This shows the replacement coil together with the original assembly:

This was a solenoid assembly where the metal tube inside the coil was secured with a retaining washer. After removal of the washer I pressed the tube out using my arbor press. I used a small screw driver bit that has a smaller diameter than the orifice where the tube pokes through:

This shows the tube slowly coming out below the assembly:

I had to use a second screw driver bit on top of the first to fully get it out:

Here a picture of the disassembled setup:

I pressed the tube back in through the new coil and secured it with the retaining washer:

Then I re-installed the de-magnetized and cleaned plunger:

I usually secure screwed in plunger extension with a drop if white glue on the threads. This helps preventing the extension from coming loose during activations of the solenoid. Then I put the carriage assembly back together. This shows it re-assembled:

I also installed a nice new shiny aluminum carriage pulley and EPDM belt to replace the original cracked plastic pulley:

The final task on the carriage assembly was cleaning and re-lubricating the pivot point of the damper-to-tonearm linkage. It is located in-between the tonearm and the sensor arm assemblies. You can see it stick out from the v-cut in the small metal plate bolted to the end of the counter weight:

I removed the two screws that hold the sensor arm to the carriage and pulled it out:

I removed the linkage (watch out for the small spring that is under the retaining washer if you do this at home):

After cleaning and re-lubricating the linkage I put everything back together and epoxied the small copper plate back that usually is pretty loose due to degraded double sided tape:

After the epoxy had hardened I bolted the sensor arm assembly back into its place.

Now it was time to focus on restoring the circuit boards. First I worked on the main  PCB. It has two power transistors installed on its solder side. It is best to replace these while the board is still installed. That makes it easy to position the replacements properly. I always replace them since they often die when Beograms are put back into service after many years of storage. This shows 1IC1 that serves as voltage regulator for the 21V rail. It is usually a TIP120 originally:

I always replace such power transistors with stronger types. They cost almost the same but promise a longer service life. For a TIP120 a TIP102 can be used:

For some reason I took the above picture before I soldered the leads. Modern TIP Darlingtons in this circuit configuration need a 100nF capacitor at their emitter to prevent high-frequency oscillations superimposed on the power rail (the yellowish package connected to the northern leg of the new transistor. The other end is connected to a convenient GND solder pad nearby.

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

Here a detail shot of the original 'RPM control section' consisting of a Siemens RPM relay and the two RPM trimmers on its left:

I replaced all the electrolytic capacitors, power transistors, RPM relay and trimmers, and the sensor transistor 1TR3 (often out of spec causing the record detection to malfunction). This shows the rebuilt board together with the removed parts:

A detail photo of the new RPM section with a new Beolover Siemens Relay Replacement for Beogram 4000, 4002, and 4004:

I always replace the original RPM trimmers with modern encapsulated 25-turn precision pots, which make RPM fine-tuning much easier.

Next came the output PCB carrying the output relay and its delay circuit:

This shows the circuit magnified:

I installed another Siemens style replacement relay and also added a (red, next to the output jack) switch that allows connecting signal and system grounds:

Connecting signal and system grounds cures most hum issues, especially when connecting the Beogram using a DIN-to-RCA adapter. In some later Beogram 4002 and 4004 versions the grounds actually are permanently connected using a wire bridge in lieu of 8C2, so B&O realized at some point that connecting the grounds is a good idea. I like the switch, giving the user a choice.

This concluded my work on the PCBs and I moved on to the RPM panel located above the keypad. This shows the panel removed and flipped over:

After removing the bulb covers the bulbs are visible in their aluminum clips:

The above picture also shows the LED boards for replacing the bulbs. They solder directly to the solder points of the bulbs, i.e. form an extension of the original circuit board:

Here a detail photo of one of them installed:

The bulb covers can be replaced after the exchange. There is no interference between the new boards and the covers.

The fourth and last bulb that needed replacing was the bulb in the sensor arm. This shows the sensor arm compartment pulled out with the original bulb still in place. Next to it is the Beolover Sensor Arm LED Light Source (Beogram 4002, 4004 and 4000) and its alignment aid:

I removed the bulb and put the LED assembly in place:

Then I removed the floating chassis and the original reservoir capacitor, and the enclosure was finally empty:

I vacuumed it and then installed new transport lock bushings. This shows one of the original ones still in place:

The original ones already showed signs of deterioration, so I replaced them as a preventative measure. This shows a set of new replacements:

They are easy to install due to their two-part design. Simply stick one half into the orifice from below

And then the other half from the top:

Note that it is a good idea to secure the parts with a bead of superglue gel on the the vertical parts before inserting them. Some Beograms have slightly larger orifices, causing the parts to sit a bit loose. Next I installed a new Beolover Main Reservoir Capacitor for Beogram 4002 and 4004 (Types 551x/552x). This shows the original setup before I removed it:

If there are connections on both end of this capacitor, it is a dual capacitance unit. A number of Beogram 4002 came with these. They integrate a 4000uF and a 1000uF capacitor in the same can. Such capacitors are not manufactured anymore.

The Beolover replacement part is setup that it can replace both single and dual-capacitance reservoir capacitors. First the red alignment piece is inserted (the part comes with two different alignment piece styles matching the layouts of the two different enclosure designs that were manufactured):

Then the replacement assembly can be bolted in place with the same screw the original capacitor used:

The above picture shows the white and black wires soldered to the respectively labeled solder pads, as well as the green AC wires from the second transformer secondary winding that used the 1000uF capacitor in the original can. It is best to solder the orange wire after everything else is in place. Usually it is just long enough to wrap it once around one of the capacitors for a neat setup:

This shows the removed capacitor together with the small rectifier that is bolted to it:

These parts are no longer needed. The replacement part also features the rectifier for the green leads.

I installed everything back in the enclosure and then it was time for a brief test of the electronics. Since this unit had burned fuses and a fried solenoid, I went about it very carefully and did not plug it in right away. Instead I ran it from a regulated and protected bench power supply, which I connected directly to the white and black leads on the reservoir capacitor:

Then I ramped the voltage slowly while watching the current on the power supply display. Already at a few volts the supply went into protection mode since the current had surpassed more than 1 Amp. this indicated a short circuit somewhere. And indeed after a bit of poking around I thought I had found the problem: one of the green secondary wires had been squeezed very hard under the back left chassis spring, which damaged its insulation, i.e. it made contact to ground!:

I repaired the insulation with a small piece of shrink tubing:

Unfortunately, this did not fix the issue. The bench supply still went into protection like before. I pulled the power-in plug from the main circuit board and now plugged the Beogram into an outlet. A muffled pop was audible from the fuse department. Sadly one of the new fuses had burned out when I plugged it in. So I had to conclude that the transformer had an internal issue. It may have overheated at some point due to the damaged wire insulation and that led to short circuits between some of the windings of the coils.

Not a big deal in this case, since my customer wanted a universal power supply installed anyway. This unit will go to Australia where they use 240V instead of the the 110V this Beogram was originally setup for. 

This shows the Meanwell RS25-24 universal power supply ready for installation bolted to its 3D printed mounting adapter:

These fairly small supplies can provide 25W and the voltage can be adjusted up to 28V, which is enough to operate a DC-motor Beogram (they are rated 15W). Note that this should not be installed in an older AC-motor type since these are rated 50W due to their power-hungry synchronous platter motors. 

I unbolted the four screws that hold the transformer and the fuse box in place:

This shows the connections inside the fuse box:

The grey leads on the bottom go to the mains plug and the red and yellow wires connect to the primary winding of the transformer. I unsoldered the transformer leads

and made a connector for the new power supply. It seems a 130 mm jumper does the trick:

After cutting off the green and blue secondary windings wires about 50 mm away from the transformer I installed 3 mm solder lugs on the wires that can be bolted to the terminals of the RS-25:

Note that I combined one of the green and one of the blue leads into single solder lugs. Polarity does not matter here since these wires connect to the rectifiers on the main PCB and the new Beolover reservoir capacitor assembly.

Then I soldered the other end of the 130 mm jumper to the terminals of the fuse box. This shows how the wires go in from the bottom and feed out of the small enclosure together with the grey mains cable:

This shows how the wiring is connected to the power supply terminals:

The line-in terminals are on the left side of the terminal block and the DC output terminals are on the right end. Again, polarity does not matter for both in and out connections. Then I bolted the supply and fuse box back into the enclosure:

This shows the extracted transformer.

After a brief power up test that elicited the proper life signs from the Beogram, I extracted the platter motor bearings from the oil vat in my vacuum chamber:

I re-assembled the motor with the re-infused bearings and installed it in the Beogram for a 24 hour RPM stability test using my Beolover RPM device:

The BeoloverRPM has two operational modes. In 'slow' mode it measures the RPM in 10 sec intervals and relays the data to a serial port of a computer where it can be recorded with any generic terminal software. This allows plotting the RPM over long periods of time using Excel or similar software. This shows the result of my 24 hrs measurement:

This curve is fairly decent for an original DC platter motor! The slight choppiness will likely go away after a few tens of hours of runtime as the top bearing gets polished in its new orientation. The observed longterm drift is related to ambient temperature changes and basically a 'feature' of analog control systems.

I also measured this Beogram in the 'fast' mode of the BeoloverRPM:

In the fast mode it transmits a RPM measurement every time a platter rib passes under the sensor. This yields high-resolution graphs that show short term RPM changes ("wow and flutter") in detail. This graph shows a measurement covering about 60 turns of the platter, representing a run time of about 2 min:

 The zig-zag pattern is a measurement artifact that comes from the fact that the spacing of the platter ribs around the platter is slightly irregular due to manufacturing imperfections. This generates a repeating pattern every 24 measurements (there are 24 ribs around the platter), which is superimposed to the real RPM changes that are introduced by the feedback system that keeps the motor RPM stable over time. This real RPM change is essentially the sine-wave like pattern that modulates the zig-zag pattern. An evaluation of the wavy component yields a wow and flutter estimate of about 0.1%.

This is 2x of the 0.05% stated in the specs list in the service manual. This difference is most likely systematic due to the entirely different way wow and flutter was measured in the 1970s when these turntables were produced. I see it for all Beogram motors that I restore. Back then the measurement was carried out with a test tone on a test record. In these measurements deviations from the tone center were measured with an analog spectrum analyzer and then converted into a wow and flutter number. It should be pointed out that this discussion is pretty academic in the first place since humans start typically recognizing frequency fluctuations above the 0.7% threshold, i.e. the RPM fluctuations of this Beogram are well below this threshold, whether the number is 0.05% or 0.1%. This motor is definitely ready for duty again!

It was time to do the adjustments to align platter, arms and the subchassis etc...for this the plinth needed to be in place so the aluminum panels can be installed. My customer wanted a new CNC machined walnut frame with 4000c-style corner splines installed.

This required extraction of the metal fixtures from the original veneered rosewood frame. It is fairly easy to get the metal parts off the wood if the frame is 'baked' for 30-60 min at around 250-300F. Once everything is warmed up the metal parts can be cut free with a carpet knife or similar. The old glue can be removed by soaking the parts in Goo-Gone. I usually wrap them in Goo-Gone soaked paper towel sheets and then I put everything inside plastic wrap to prevent the stuff from drying out. After ~24 hrs the glue can be rubbed off with a bit of elbow grease. After washing the parts with dish soap in the sink they look almost like new:

The next step is applying 3M 300LSE double sided tape ands cutting it to size with a razor blade to match the cutouts of the metal parts:

This shows the metal parts glued in place inside the new walnut frame:

Since it was a rare rainy afternoon in Albuquerque, I decided to also install the new dust cover. The original one was badly scratched 

and while I could have polished it, it is so much nicer to install a new reproduction dust cover from the DKsoundparts store in Denmark. These hoods are perfect reproductions of the originals and are cast from a mold. Perfect!

The first step was removing the black metal hinge from the original hood. I separated the side parts of the aluminum trim from the hood with a razor blade:

After removal of the screws the hinge assembly can be removed. This shows the liberated hinge together with the new hood and a new aluminum trim strip:

After a few disasters where freshly installed dust covers cracked off their hinges after a few weeks of use I now always conduct a test before I install a new one to see if there is any interference between the metal and plexiglass parts.

So I bolted the hinge into the new hood and checked if the metal part touched the plexiglass anywhere along the way when the hinge was moved relative to the hood. The crucial part is the 'stabilizer bar' of the hood that sits under the area where the aluminum trim is applied on top. It can happen that this plastic bar chafes on the metal if the hinges are somewhat bent. After a while this stress can crack the plexiglass where the hood bolts to the metal part.

This hood passed the test. The gap was at least a mm during the entire motion range of the hinge:

I removed the hinge and corrected a small design flaw of these (otherwise wonderful!) hoods. For some reason there is a 'channel' under the area where the side parts of the aluminum trim bend down. This channel makes it difficult to bend the aluminum strip properly making a sharp corner since there is a void under the strip where it is bent. I came up with a solution and now epoxy small 3D printed plastic bits into the upper end of the channel:

After the epoxy is cured, I cut the plastic bits off with a new razor blade so I get a fairly precise edge all across the aluminum strip area:

After this corrective measure, it was time to do the installation: First I used my alignment tools to determine the proper centered alignment of the trim strip:

I usually mark the position with a pencil against a strip of blue tape applied to the hood:

The pencil rubs off easily after the strip has been glued in place. After making this alignment mark I bolted the hinge in again. This time I put a bit of white glue onto the threads. This helps keep the screws in place, which cannot be tightened very hard. Too tight screws are another leading course of hoods cracking off the hinge prematurely:

This shows the hinge bolted in on one side:

After also bolting it in on the other side it was time to apply the adhesive backed trim. First I glued the center part on top of the hood after carefully aligning it with my pencil mark:

And then came the fun part: Bending the trim down at the sides. I use a 3D printed plastic block that has the exact width of the strip. Since I add the little plastic bits I get pretty happy sharp corners:

There is the other side:

Before adding the plastic bits there was always a bit of a 'rounding' towards the front where the channel was under the strip. Beoprecision is Beolovely!...;-).

This shows the end result:

After installing the new walnut plinth it was time to finally do the adjustments.

First I did the arms parallelism. For this I recently designed a 3D printed adjustment tool that clamps to the carriage rails. 

The arms adjustment requires to first get the sensor arm base perpendicular to the rods, followed by adjusting the sensor arm to be also perpendicular to the rods, followed by adjusting the tonearm to be parallel to the sensor arm at a ~7mm gap between them.

My new tool has a 'lower section' that allows adjusting the base:

And then the arm can be adjusted with the 'upper section':

The gap between the sensor- and tonearms is adjusted by moving the sensor arm assembly left or right and bolting it down in a spot where the gap is 7mm. Once that is in place the parallelism of the tonearm relative to the sensor arm can be fine-tuned by bending the end of the damper to arm linkage a bit to move the v-cut a bit left or right until the arms are parallel.

After the arms looked good, I replaced the locking washer that holds the counter weight screw in place 

with a square nut and a washer:

This allows locking my calibration safely in place so it survives the rigors of shipping. I adjusted the counter weight that 1.2g on the weight actually corresponds to 1.2g on a tracking weight gauge:

The weight dial is notoriously imprecise, but most cartridges need 1-1.5g, so adjusting it to be precise at 1.2g makes it reasonably accurate across the needed adjustment range. Of course, generally it is best to use a digital tracking weight gauge and simply turn the dial until the desired weight shows on the gauge.

After the tracking weight was calibrated I adjusted the arm lowering limit:

The needle should be about 1 mm away from the lower parts of the black ribs once the arm lowers on an empty platter. This is a safeguard against a record detection circuit malfunction that could cause arm lowering on an empty rotating platter.

Once the arms have been properly adjusted the other mechanical adjustments can be done. 

First came the tedious process of aligning 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 an iterative process that takes a few passes to yield proper alignment. Sometimes I spend three hours, sometimes 30 min. It always depends on how off-spec the alignment is for a particular Beogram.

On to adjusting the tracking feedback. The platter should do about 2 turns after the needle hits the grooves before the carriage motor starts advancing the carriage. The little white trimmer on my tracking sensor LED light source is a convenient feature for fine tuning the feedback after a coarse mechanical adjustment has been achieved:

I still had to qualify the record detection circuit performance. The first step was adjusting the bias of TR4 to get 4V DC at the collector:

After this adjustment I moved the trimmer over to the component side of the board:

Then I put the platter on and measured the signal on the TR3 collector with an empty spinning platter under the sensor. I got a nice and juicy signal with an amplitude exceeding 7V:

Each of the drops corresponds to a platter rib passing under the sensor. This signal is used by the detection circuit to disable arm lowering when there is no record on the platter, or to 'find' a 7" single on the platter.

On the way to a first test spin of this Beogram I still had to replace the grimy old RCA plugs

with a couple nice new ones with gold plated terminals:

Beogolden!..;-).

I still had to replace the smudgy original keypad with a nice newly restored one. We offer this as a service (i.e. send your original keypad and we send it back restored), but we also have a keypad restoration kit for DIY for 'mechanically inclined' Beogram fans. This shows old and restored keypads together:

The original plastic clips for holding the RPM panel in place were cracked on both sides:

So it was time to replace them with new reproduction parts from the DKsoundparts store in Denmark. These nice plastic parts need the original metal springs installed. This shows the extracted springs togheter with the new plastic parts:

I installed the springs:

and then bolted the plastic parts into the restored keypad assembly:

this shows the left side:

Before I installed the restored keypad I replaced the original output PCB with the Beolover Internal RIAA Pre-Amplifier for Beogram 4002 and 4004. It is a simple plug-and-play process. The RIAA board has exactly the same layout like the original PCB: 

Simply pull the plugs and remove the board and install the RIAA board in exactly the same way:

I put the keypad in and installed the platter and aluminum panels and then it was finally time for a first test spin with this beautifully restored Beogram 4002! I selected a recent addition to my collection, Earl Klugh's 'Crazy for You' album. The album was released in 1981 on the Liberty label (LT-51133). I like listening to it while working on Beograms! A nice calm line-up of tracks with nice melodies. Here a picture of the album playing on the Beogram:

I really like the nice 4000c style walnut plinths. Perfect for the 4002!

After this successful test spin I bolted the new hood on and took a few nice pictures. I think I really like the non-spec 4000 style brownish tinted hood. It definitely looks better when closed since it is more transparent than the spec gray tinted hoods! Also the brownish color looks better with a real wood plinth.

I never understood why B&O changed the hood style for the 4002/4004 models to grey tinted. I guess 'out with the old, in with the new!' is a powerful motivator in marketing departments...;-).

Enjoy the pictures: