beolover: Beogram 4004: Complete Functional and Cosmetic Restoration, and a Test Spin with Freddie Hubbard

This post describes the restoration of a Beogram 4004 that I received a few months ago from a customer in Virginia. This post discusses my first assessment of the unit.

The customer ultimately decided to sell it to me rather than have it restored for himself, i.e. this unit is now available via the DKaudiolover store. This picture shows the final result of my efforts:

Beolovely! Let's see how this unit was 'resurrected' from its fairly sad state when I received it...;-):

This shows the unit as-received with the panels and platter off:

As usual with the DC motor Beograms I started my process by removing the platter motor:

Most of these DC motors have dry motor bearings, which need to be re-infused with oil under vacuum. The removal of the bearings requires a complete disassembly of the motor:

The two small donuts on the black pad upfront are the bearings. I immersed them in synthetic oil and pulled a vacuum. Immediately strong bubbling started:

The bubbling is indicative of air being drawn out of the porous bearing material. Once the air is out the oil can diffuse into the bearing and replenish its oil reserves. This process usually takes two or three days. The end is reached when the bubbling stops.

While this process ran its course, I focused on the remaining restoration tasks. First came the restoration of the carriage assembly. This shows it in its original condition:

The arm lowering and carriage moving mechanisms need to be disassembled for removal of the original and usually hardened lubricants. This shows the assembly with all moving parts removed:

It is good to protect the underside of the carriage with a soft foam pad or similar. This shows the removed parts ready for cleaning:

After an hour in my ultrasonic cleaner things definitely looked brighter!...;-):

A crucial restoration item is the demagnetization of the solenoid plunger. I use a standard tape head demagnetizer on it. A small iron setscrew can be used for testing if all magnetization has been removed. If it does not get attracted all is well:

Another important item when restoring the arm lowering mechanism is the installation of a new Beolover Damper Gasket for Beogram 4000, 4002, and 4004:

The original damper gaskets are usually hardened and/or deformed and that frequently causes inconsistencies in the arm lowering speed. This shows everything back in place:

There is one more component of the arm lowering mechanism that can only be accessed by removing the sensor arm assembly. This component, the connecting linkage between damper and tonearm, can be seen here sticking out from the v-cut in the metal part that is bolted to the back of the counter weight assembly:

I removed the sensor arm

and extracted the linkage:

After putting everything back together it was time to replace the tracking sensor light bulb with a Beolover Tracking Sensor LED Light Source for Beogram 4002 and 4004. This shows the original setup still in place:

I removed the original bulb housing. This shows it in direct comparison with the Beolover LED part:

The LED is in the same geometric location as the filament of the light bulb. This shows the part installed:

When disassembling things I had noticed that the carriage motor was wiggling inside its plastic enclosure. This usually means the foam damping material has disintegrated. And indeed this was the case here, too:

I cleaned off the mess and replaced it with a new piece of polyurethane foam:

Then I re-installed the motor and also put on a new Beolover Aluminum Carriage Pulley for Beogram 4002 and 4004 and a new EPDM belt:

This completed my work on the carriage assembly and I moved on to restoring the main PCB. The PCB has two power transistors that are installed on its solder side. It is best to replace them while the board is still installed. This simplifies their placement. This shows 1IC1 that stabilizes the 24V power rail:

Originally it is a TIP120 power transistor. I usually replace it with a stronger TIP102:

For some reason modern TIP transistors in this circuit configuration tend to generate a high frequency oscillation superimposed to the power rail. This can be quenched by installing a 100nF capacitor between the emitter and GND (small yellow capacitor in the above picture).

After replacement of 1IC4 with a TIP107 I removed the board for replacement of the remaining components that need attention:

This shows a detail shot of the 'RPM section' consisting of the two RPM trimmers and the RPM relay, a National type in this case:

I replaced all electrolytic capacitors, power transistors, RPM section parts, and the sensor transistor (they often have lost gain impairing the record detection function). This shows the rebuilt board:

This shows the RPM section with a new Beolover National Relay Replacement for Beogram 4002 and 4004 as well as two new multi-turn RPM trimmers in place of the original single turn units:

Next came the output board that I extracted after removing the keypad assembly. This shows it in its original condition:

This board is the main difference between the Beogram 4002 and 4004 types. In addition to the common output relay circuit, the 4004 board also contains circuitry that allows remote control of the Beogram 4004 via a Beomaster 2400 and its remote control. This circuit occupies the left 2/3rd of the board. The output section is on the right 3rd, as shown here magnified:

I replaced all electrolytic capacitors and the output relay:

I also installed a switch (red) behind the connector for the output cable that allows connecting system and signal grounds in case there is a hum in the signal when connecting to an amplifier:

Next I replaced the incandescent light bulbs in the RPM panel above the keypad. This shows it already removed and flipped on its back:

The bulbs can be accessed after removing the two bulb covers:

The above picture also shows the two Beolover RPM Panel LED Backlights for Beogram 4002 and 4004 (Types 551x/552x) in front of the panel ready for installation.

After removal of the light bulbs the two small PCBs solder directly to the points where the bulb wires were attached:

Here a detail picture of one of the installed boards:

There is a fourth incandescent bulb that needs replacing with an LED assembly. It is in the small compartment at the end of the sensor arm. This shows the compartment pulled out, which reveals the bulb:

This shows the Beolover Sensor Arm LED Light Source (Beogram 4002, 4004 and 4000) installed:

At this point pretty much everything removable was removed from the enclosure except the floating chassis. This made it the perfect moment for tending to the completely degraded transport lock bushings. This shows one of the locks with the top part removed. The orangish fragments to the left are remnants from the original lock bushings:

I disassembled the other two locks and removed the floating chassis, leaving an empty enclosure full of bushing fragments:

It is important to remove all the fragments since they can get lodged under the floating chassis impeding its motion. This shows the vacuumed out enclosure:

The next step was the installation of a Beolover Transport Lock Bushing Set for Beogram 4000, 4002, and 4004. The installation is easy. Simply push one half of the bushing from below

and the other from top:

Please, note that some floating chassis can have slightly larger bushing holes than others. In such cases it can be appropriate to put a bead of superglue gel on the bushing parts before pushing them in to hold them in place more firmly. This shows one of the bushings with its transport lock screw already in place:

After completing the three locks it was time to install a new Beolover Main Reservoir Capacitor for Beogram 4002 and 4004 (Types 551x/552x). The original one is shown here:

These old electrolytic cans are often degraded with reduced capacitance or they are leaking electrolyte, and it is a good idea to replace them when restoring a Beogram. This shows the Beolover part in place:

It bolts to the same mounting hole used for the clamp of the original capacitor:

In this case the original cap was a single capacitance unit, i.e. all that I needed to do in addition to bolting it in, was soldering the white and black lead to the respectively marked solder pads on the PCB.

Before making adjustments to the floating chassis and bolting the keypad back into the enclosure the damaged wood plinth (it had a shaved off corner in the front) needed to be replaced. I decided to replace it with a new CNC machined mahogany wood frame fitted with Beogram 4000c style corners with splines. For its installation the first step was the removal of the metal fixtures from the damaged original frame. After baking the frame for about half an hour at 300F it was easy to cut them loose by pushing a carpet knife between the metal and wood parts.

Once the metal parts were liberated the old glue needed to be removed. A good way to do this is soaking the parts for a day in Goo Gone Pro-Power. I usually spray the parts liberally with the stuff and then wrap them in cling wrap:

After ~24 hrs the old glue can be rubbed off with paper towel and the parts look like new again. I usually clean the remaining Goo Gone off in warm water with some dishwashing detergent followed by a wipe with isopropanol. Beolovely!:

Once the parts are clean, double sided tape can be applied:

Once the tape is on, the parts can be glued into the new wood frame. They need to be flush with the recess machined into the bottom part of the frame. A good way to do this alignment is to use two Beogram aluminum panels. They have a thickness exactly matching the recess. The difficulty is to get the metal parts fully into the corners while being sticky on both 'legs' after removal of the protective paper on the tape. I usually use strips of parchment paper to keep one side from sticking while gluing in the other:

This shows the final result:

After installing the frame I replaced the cracked plastic part that holds the output PCB and cable in place with a new Beolover Lead Holder and Output Board Bracket for Beogram 4002 and 4004. This shows the damaged original (right) and new replacement together:

And here bolted in place:

Then it was time to install a newly refurbished keypad. This picture shows the original (left) and new keypad next to each other:

Keypad Restoration Service for Beogram 4002 and 4004 is available via the Beolover's Dkaudiolover store. Simply send your original damaged keypad and get it refurbished.

Before installation the cracked original plastic parts that clip the RPM panel in place needed to be replaced with reproduction parts from the DKsoundparts store in Denmark. This shows the damaged original parts:

I removed the old parts. This shows them together with their new replacements:

The new parts come without the metal springs. I liberated them from the original parts

and installed them in the new ones:

And then I bolted them into the restored keypad:

After installing the main PCB it was time to adjust the sensor arm transistor bias to yield 4V at the collector:

In the meantime the oil infusion of the motor bearings had concluded. I extracted them from the oil vat in the vacuum chamber

and re-assembed the motor. Then it was time to do a 24 hrs RPM stability test with the BeoloverRPM calibration tool set to 'slow' mode:

In 'slow' mode it is set up for recording a RPM measurement every 10 sec and sending the data to the serial port of a computer. This is the curve I measured after about 24 hrs:

This is pretty much as good as it gets with a Beogram DC platter motor! There is a bit of drift caused by temperature changes, but this is normal for analog feedback systems.

I also did a wow and flutter measurement using the 'fast' mode of the BeoloverRPM:

In fast mode it will send an RPM measurement every time a platter rib passes under the sensor. This high resolution data allows seeing the feedback control mechanism in action. This graph shows a measurement over about 36 rotations of the platter (~1 min runtime):

The observed graph is a superposition between a slow sine-like wave and a fast repeating pattern. The fast pattern is a measurement artifact caused by the slightly varying distance between the ribs around the platter. It repeats exactly every 24 data points, corresponding to the number of ribs.

This is a manufacturing 'defect' that can be regarded as a 'finger print' that is unique to each platter. This particular platter has a double peak pattern meaning that the variation of the rib distances goes up and down twice around the platter.

The 'sine-wave-similar' superimposed pattern reflects actual RPM variations. They are caused by the feedback-based control mechanism that continuously regulates the speed around the setpoint. An evaluation of this line shape reveals RPM variations of about 0.1%. This is about double the <0.05 number stated in the service manual, but this may well be a result of the entirely different measurement methods used here and back in the 1970s, when digital technology was still in its early beginnings and wow and flutter was measured with test records and analog frequency analysis. In any event these variations are much too small for humans to notice. Only RPM variations above about 0.7% can be detected by most people.

While the 24 hrs RPM stability measurement was underway I exchanged the scratched original hood with a new reproduction hood for Beogram 4002 and 4004 from the DKsoundparts store in Denmark. This shows the original hood still bolted to the metal hinge part:

It is easy to remove the plexiglass cover: Just use a razor blade and cut the side parts of the aluminum trim loose so the mounting screws can be accessed:

After removal of the screws the hinge is liberated and ready for the new hood. This shows the new hood and aluminum trim (the original trim usually cannot be re-used since it gets damaged when removal is attempted):

Installation of the trim after bolting in the new hood requires carefully centering it. I designed special 3D printed parts that clip to the hood, which allow centering the trim prior to its installation:

Once the trim center position has been found

a small pencil mark between the trim and a piece of blue tape stuck to the hood records the center position:

Then the hood can be bolted to the hinge:

If you do this at home, make sure not to over-tighten the supplied bolts. There is a risk for cracking the hood hinge areas if the screws are tightened too much.

Once the hood has been bolted in, the trim can be installed:

This shows the final result:

Beolovely!

And then it was time to do the adjustments.

The first step is usually aligning arms, platter and subchassis to achieve a 23 mm top-of-the-arms-to-platter distance and to get the platter and arms parallel, while leveling the subchassis to get the platter flush with the surrounding panels. This can be a pretty tedious and iterative process, depending on how out of alignment things are. I usually start with getting the aluminum panels properly aligned and settled on their alignment pins together with the keypad. Then I get the arms parallel and in 23 mm distance with the aluminum panels. Then the platter can be adjusted to be in the same distance. If everything is done right at that point everything is parallel and in the proper distance.

Once this was squared away the tonearm adjustments could be done. I usually start by replacing the flimsy locking washer that holds the counterweight adjustment screw in place

with a square M3 nut and a washer:

This allows locking the calibration in place to better withstand the rigors of shipping. I adjusted the counter weight to get 1.2g tracking weight at the 1.2g setting of the adjustment wheel:

Then I adjusted the arm lowering limit that the needle missed the lower parts of the platter ribs by about 1 mm:

This is an important adjustment that will save your tip in case the record detection mechanism would ever fail and allow arm lowering onto an empty platter.

Then I adjusted the tracking feedback:

And then it was finally time to give this lovely Beogram 4004 a test spin! I selected a record I recently found while I was hunting for some other records on Discogs: "Splash" by Freddie Hubbard (Fantasy F-9610). He recorded this album in 1981 and it is pretty disco!...;-). Not his most important record, but still pretty enjoyable! Of course this album was ultrasonically cleaned using a CleanerVinyl ProXL setup!

Here an impression of the Beogram 4004 playing this album: