Beogram 4000: An Interesting Tracking Issue and Installation of a Beolover Efficient 24V Power Supply and Main Capacitor Array

I recently received a Beogram 4000 back that I restored in February 2024 for a customer in Massachusetts due to an issue with the tracking system.

The phenomenon described was that the carriage occasionally would not track after the needle was lowered into the run in groove. I tried for some time unsuccessfully to reproduce this issue until it finally happened after I put a new record on the deck. It became clear to me at this point that it depended on the particular record played since the problem was fully reproducible with this particular record, i.e. happened every time I started the deck.

After a bit of head scratching I figured out that when facing a 'fast' run in groove, the arm would move so fast to the left, that the carriage motor was not able to keep up moving the carriage along before the tracking sensor aperture had already completely passed past the photo resistor opening in the sensor, in effect turning off the motor. The motor voltage is directly controlled by the resistance of the photo resistor and if it is dark the motor stops running.

The big question at this point was why this would happen with this particular Beogram 4000 and not with all the other ones I restored so far. 

I set out to measure the motor voltage with my oscilloscope. This trace shows the evolution of the motor voltage during regular play of a track on this Beogram 4000 as I received it:

What we see on this graph is that the motor voltage 'oscillates' and that there are larger jumps every few oscillations. It turns out that the oscillations correspond to the rotation of the platter: Note the 4s per unit time constant and that there are two oscillations per scale unit. It takes about 2s for a rotation at 33 RPM...

The voltage fluctuations are essentially a result of the eccentricity that most vinyls have to varying degree. The graph means that the carriage only weakly moves during ~6 or 7 rotations and then in one large burst after the voltage gets big enough for the motor to overcome stiction. Then the carriage drives forward until the arms are almost parallel and then the process starts again.

This suggested that the amount of light falling on the photoresistor was not enough to generate a high enough motor voltage through the H-bridge.

I decided to modify one of my standard Beolover Tracking Sensor LEDs to give the LED a bit more current to make it light up more:

After I installed the modified light source I re-calibrated the tracking feedback gain to start moving the carriage after about 3-4 rotations and measured again:

This time the curve looked like this and the carriage reacted speedily after hitting the run in groove of the record that caused the above issue. During play of the same track as above it gave me this trace on the oscilloscope. This shows that the carriage is now basically moving every rotation of the platter.

There are still a few variations between the peaks, which probably indicates a variation of the friction the motor encounters as the spindle rotates. Most spindles are not completely straight and so there is some change in friction during a rotation. It appears that each rotation of the spindle corresponds to maybe 4-5 turns of the platter. Of course this depends on the particular record since the groove density varies from record to record.

It seems this issue is fixed, but of course I am wondering why this particular Beogram 4000 needs a higher light intensity. It may be that the carriage motor itself is a bit different and turns less easily than others at a given voltage. Another possibility is that the photo resistors in the sensor are different, or from a different batch. The Beogram 4000 has two photo resistors in the tracking sensor that allow the carriage to actually track both forward and backward (this was discontinued in the later 4002 and 4004 models which only have one resistor for forward movement).

This unit seemed to track about the same in both directions when the arm was moved to the right or left manually, so I think it is not an issue with one of the resistors. If it were, I would expect that only one direction be affected. B&O has a history of making small circuit changes over the manufacturing run of a design. Maybe they simply used somewhat different resistors or there was variation between batches of them. Possibly the potentiometer for adjusting the voltage into the bulb that is normally found on the original configuration of the tracking sensor of the 4000 is an indicator that they already had this issue when they produced them, and someone calibrated the bulb intensity before Beograms were shipped out.

We may never know!...;-). Anyway, let's hope this tracking mechanism is fit for duty again!

While the unit was on my bench my customer decided to let me install one of my new efficient 24V power supply and main capacitors boards for Beogram 4000:

These boards elegantly replace the big reservoir and motor capacitor mess that is normally found in Beogram 4000s. It also updates the old-fashioned linear regulator based 24V power supply that causes a lot of power loss and heat emission in Beogram 4000s. This board will make any 4000 run more efficiently and the unit will get much less warm.

This Beogram of course already had my previous capacitor replacement kit implemented:

Out with the old

and in with the new! This shows the board bolted in. The solder pads are clearly labeled and in the right spots for easy connections:

This shows all the leads soldered to the board and the motor re-installed:

All good now with this Beogram 4000!

I will play it a bit more and then it will be time to send it back to my customer!

Beogram 4000: Full Functional Restoration and a Test Spin with the Bill Evans Trio

This post describes the restoration work done on a Beogram 4000 that I recently received from a customer in Massachusetts. The initial assessment of this unit is posted here.

This shows the unit with the aluminum panels and platter removed:

As usual, I started my process with rebuilding the systems on the carriage. This shows the arm lowering mechanism and the solenoid-activated switches:

I removed the rods on which the carriage travels, the spindle and the linkages of the arm lowering mechanism, as well as the damper. These parts often have hardened lubricants, which need to be removed and replaced with modern synthetic lubricants:

This shows the removed parts ready for the ultrasonic cleaner:

A special feature of the setup in the Beogram 4000 are the solenoid activated switches. Their contact terminals are usually oxidized and this can create operational issues. This shows the original state of the switches indicating previous issues that someone tried to correct by bending the terminal tabs:

The removal of the contact terminals requires extraction of the small vertical boards from the base PCBs. This shows the carriage upside down, which reveals the solder points of the vertical boards:

Here in magnification:

Sadly, the board that holds the switch that limits the solenoid current after the arm is down was cracked:

This shows the vertical boards after removal, as well as the cracked piece:

With the carriage up, it is the perfect moment for also extracting the terminals of the carriage position switches. This shows the carriage position PCB still in place:

Removal of its two screws and disconnecting the carriage motor leads allows a peek at the switch terminals:

I removed all the terminals of the solenoid and carriage switches to get them ready for nickel/gold plating:

I also removed the plastic actuators of the carriage switches for cleaning:

They are usually full of old lubricants since many owners think the actuators are the issue when operational issues develop, and their easy accessibility encourages the application of generous amounts of lubrication. Luckily, there are ultrasonic cleaners!...;-). It is also a great idea to replace the solenoid current limiting resistor that also resides on this board. It is usually semi-fried due to the fact that it is a bit under-dimensioned for the application:

After a thorough cleaning I coated the contact terminals first with nickel to ensure good adhesion, followed by a layer of gold-cobalt alloy (which is the industry standard for contact plating). The cobalt alloy is much harder than pure gold, while still having an excellent corrosion resistance. This shows the freshly plated terminals:

Beogolden! I re-installed the terminals in their respective boards:

and also put in a new 3W solenoid resistor:

With the carriage board removed it is usually the perfect moment to replace the solenoid transistor, which is originally a TIP41 power transistor. In this Beogram 4000 I found that it was already replaced with a modern BD809, which is an excellent replacement choice. Higher power and higher voltage capability. I left it in place:

Before I put the solenoid switches back together I glued the current limiter base PCB back together with a generous application of epoxy:

This shows the switches back in their place:

With the carriage up it is also a great moment to replace the cracking-prone MMC cartridge mount. For this the tonearm needs to be removed and the MMC mount assembly extracted. This is done best by 'simmering' the arm in hot water for 30-60 min. Then the assembly can usually be pushed out with a suitable rod that fits into the aluminum tube:

This shows the Beolover MMC mount replacement soldered to the original wiring:

The Beolover MMC mount is a 3D printed nylon part laminated with a flex-PCB that has gold plated traces that match the cartridge terminals. I installed the part making sure that cartridges will snugly fit without a significant gap to the aluminum tube. This was achieved by using a special tool that I developed for holding everything in the proper place while the glue hardened:

After this process it was time to re-install the arm. This was a later model 4000, so the MMC mount connected via soldered wires. It is a good idea to fixate the arm for the soldering work:

After the arm was back in place it was time to install the cleaned mechanical parts:

As usual, I replaced the damper gasket:

The original gaskets are often hardened and that causes issues with the arm lowering speed. Sudden drops or not lowering arms can be the result of retaining the original gasket. A new gasket ensures a consistent arm lowering speed.

This late model Beogram 4000 was already outfitted with a solenoid arm with a plastic end piece. Unfortunately, these tend to crack around the rivet that holds them to the main metal part: 

I removed the cracked part and installed a Beolover solenoid arm extension:

This shows everything put back together:

The final linkage that needed re-lubrication is the one that connects the damper to the actual arm assembly. You can see it stick out in the V-groove in the lateral arm in this picture:

To get to the pivot point of this linkage, the sensor arm assembly needs to be removed. This shows the linkage already removed from the assembly (make sure you do not loose the little spring that is under the retaining washer on the pivot point if you try to do this at home):

As usual, the small copper plate that aids the lateral arm movement in up position came loose after a light tug with my tweezers:

I glued it back in place with epoxy and re-installed the sensor arm with the re-lubricated linkage.

The next step of the carriage restoration was the installation of a LED based tracking sensor light source. This shows the original setup with the black bulb housing in place:

I removed the bulb housing, which revealed the aperture that governs the tracking feedback process:

Then I installed a Beolover tracking sensor LED assembly for the Beogram 4000. It is a drop in ready part that directly replaces the original bulb housing:

The final step was the replacement of the cracked carriage pulley. They always crack at the square nut that counters the set screw:

I replaced it with a precision machined aluminum pulley:

This concluded my work on the carriage and I began working on restoring the electronic circuitry and the platter motor. This shows the original main and motor capacitors, as well as the AC platter motor:

I removed everything:

This shows the original motor and reservoir capacitors

and the extracted motor:

Since oil infusion of the bearings can take up to three days, I disassembled the motor first:

I immersed the motor enclosure halves in motor oil to re-infuse the integrated bearings with motor oil. After I pulled a vacuum strong bubbling occurred indicative of the oil infusion process:

While this process was under way, I focused on the remaining restoration tasks. This shows the main PCB in original condition:

I replaced all electrolytic capacitors, power transistors, the RPM relay and the RPM trimmers with modern parts. I also replaced the record sensor transistor and its biasing resistor. This is a frequent trouble point since the original transistors often loose part of their gain, which can disable the record detection circuit. This shows the rebuilt board together with the extracted original parts:

I moved on to the power supply PCB next to the transformer. As found in many Beogram 4000 the 6V supply fuse had disintegrated.

I installed new fuses on the PCB

and also in the main fuse cabinet, where the fuses had the same issue 

Then I replaced the two 1uF electrolytic capacitors and the BC142 power transistor on the PCB:

In the meantime the oil infusion process had completed and it was time to put the main capacitor array and the platter motor back together. But first I installed a new 3W 24V Zener diode:

The original Zeners are under-dimensioned at 1W and tend to get pretty hot, which drives their voltage out of spec causing the 24V rail to actually have 25-27V. Not great. The installation of a stronger Zener solves this issue and the voltage is reliably set to 24V.

This shows the new capacitors and the re-assembled motor installed:

While the main PCB is up, it is a good moment to replace the power transistors of the push-pull stage of the AC motor driver. These transistors are usually TIP31/32 pairs. They are bolted to the enclosure for enhanced cooling:

I replaced them with stronger TIP41/42s:

Before I completed my work on the main PCB, I adjusted the sensor transistor bias with the newly installed 5MOhm 25 turn trimmer to the specified 1.8V DC at the collector:

Then I moved the adjusted trimmer to the component side:

After completing the main PCB I moved on to the keypad cluster. I removed it and put it into service position:

This revealed the four incandescent bulbs that back-illuminate the user-accessible RPM trimmers as well as the carriage position scale

and the switch terminals that are controlled via the keypad:

I removed all of the terminals for nickel/gold plating. This shows them in their original oxidized condition:

And after the plating process:

I re-assembled the switches:

Beogolden!...;-). Then I installed BeoloverLED boards to replace the carriage position back-lighting bulbs and regular 5mm LEDs with suitable resistors to replace the bulbs of the trimmers. You can see the 33 RPM LED lit up in this picture:

The final incandescent bulb that needed replacement was the one in the sensor arm. This shows the small sensor compartment pulled out from the end of the sensor arm together with the LED board and positioning aid:

I removed the original bulb and placed the LED assembly:

This restoration was coming to a close and so it was time to focus on the signal path. First I installed a switch that allows connecting signal and system grounds:

Connecting system and signal grounds is usually a good way to quench any hum issues that may arise, especially when connecting a Beogram to a RCA input amplifier. 

The next step was replacing the original as usual corroded DIN5 plug:

I installed a nice all-metal modern plug with gold plated terminals:

Before I started on the adjustments, I replaced the cracked plinth guidance washers. This shows one of the cracked original ones:

I replaced all five with Beolover nylon guide washers:

Then it was time to do the adjustments of the floating chassis, platter to arm distance, arm parallelism and platter height. This process can be a bit iterative until the sweet spot has been found. Don't give up if you try this at home. Beopatience!...;-)

Then I installed a nut on the counterweight adjustment screw. In its original setup it is just held in place with a retaining washer that sometimes is glued in place:

I removed the washer and put in a square M3 nut which allows tightening the adjustment screw in place. This ensures that my weight calibration survives the rigors of transport:

Next I adjusted the arm lowering limit to stop the needle from going beyond about 1 mm above the lower parts of the platter ribs:

This important adjustment ensures survival of the tip should the arm ever get lowered on an empty rotating platter in case of a record detection circuit malfunction.

After this I calibrated the tracking weight with a digital scale to yield 1.2g at the 1.2g scale setting:

The weight adjustment dials in Beogram 400x are notoriously inaccurate over their entire scale. Therefore, it is best to calibrate them to be accurate around the 1.2g setting, which is the tracking weight of most B&O cartridges.

Now it was time to adjust the tracking feedback to ensure a proper arm alignment during play:

Before playing a first record I measured the RPM stability of the platter drive with the newly re-designed BeoloverRPM device:

It allows a precise adjustment of the RPM as well as logging the RPM in 10s intervals. Perfect for detecting intermittent RPM inconstancies.

The serial port connection simultaneously transmits all data points to the connected PC, which allows  detailed monitoring of the RPM over extended periods of time.

This graph shows a 24 hrs measurement:

This is pretty much as good as it gets. The slow drift of the RPM is a result of the profound temperature change in my work shop between winter days and nights in Albuquerque, NM.

I also measured the motor voltage with the oscilloscope. A nice sine curve together with high motor torque indicated that the motor department was working well:

The final measurement was the record detection sensor response. In the 4000 the amplitude should exceed ~2.2V, which was easily achieved:

And now it was finally time to enjoy this fully restored Beogram 4000 by playing a nice record. My customer sent me a present along with this deck, an audiophile pressing of the seminal Bill Evans Trio recording "Waltz for Debbie", which was recorded live on June 25, 1961 at the Village Vanguard. What a perfect combination with this great looking deck! Like all newly pressed records I cleaned this vinyl ultrasonically to remove any release agent residues before I played it. This was done on a CleanerVinyl ProXL setup combined with a UC-8360 high frequency 132 kHz ultrasonic.

My favorite track on this album is Milestones, a famous Miles Davis composition.

While I was enjoying this restored deck, I worked on replacing the original scratched up hood with a new one from the beoparts-store in Denmark.

But first I needed to fix the hood hinge that had started to corrode:

I sanded the corroded areas and then spray-painted them with satin black. The next step was to align the new aluminum trim with the hood to be centered. I usually place a strip of blue tape and then align the strip with two 3D printed alignment tools before I mark the position with pencil:

Then it was time to bolt in the re-painted hinge part:

This shows the new hood after glueing the aluminum trim into place:

I will now play this Beogram 4000 for bit longer to make sure there are no intermittent issues, and then it will be time to send it back to its owner!