Beogram 4000: Installation of a New Precision Machined Solid Teak Frame

I recently received a lovely new CNC machined solid Teak plinth from Christian Hakansson, to be installed on the Beogram 4000 that I recently functionally restored.

Unfortunately, this Beogram had an original plinth that was damaged at the corners (please, watch out for door jambs when you carry your Beograms around!!...;-):

I removed the plinth:

If you do this at home, please, note that there is a 5th bolt under the keypad, i.e. the keypad needs to be removed before the plinth can be extracted. 

The deformed particle board under the veneer suggested that this plinth not only had damaged corners, but also had been in contact with too much moisture at some point in time:

The installation of a new plinth requires that the metal fixtures be removed from the old plinth to be transferred to the new one. I usually wrap the plinth in aluminum foil and then I put it into the oven for an hour at 200-250F. Once it comes out the metal parts can be separated with a suitable spatula or similar:

Once the parts are off, the residual contact cement

needs to be removed. I usually soak paper towel with isopropanol or goo gone and wrap the metal parts in it:

I put aluminum foil around it to prevent evaporation:

After 24 hrs the glue can be rubbed off with the drenched paper towels:

This shows one of the cleaned metal fixtures:

I use ultra thin 3M 300LSE adhesive tape for the installation. This makes the process much cleaner than with contact cement. The tape can be cut out with a razor blade to match the frame shape:

It is important to not cover the little spring tabs that are in the back of the frame:

Otherwise they can get glued to the wood and are not able to perform their duty (holding the plinth in place once it is pushed back) anymore.

Now comes the moment where the metal parts are glued to the wood. This is probably the most difficult task since one only has one try and the parts need to be placed precisely at the right level on the frame. This is greatly simplified by using the aluminum panels of the Beogram. They have the right thickness and can be used to align the metal parts with the frame.

It can be challenging to get the metal parts all the way into the corners since there is adhesive on both sides. I usually deal this issue by using some parchment or wax paper to prevent one side from sticking, while I push in the part all the way into the corner:

This shows one side glued while the other still has the paper in:

Sometimes the metal parts are not entirely conform with the frame in the corners. Many Beograms actually came with this issue 'factory installed'...;-). I sometimes use tongue-and-groove pliers to help seating the metal parts in the corners. Be careful not to damage the frame if you do this at home: 

This shows the new frame with installed metal parts:

And here a couple happy shots of the Beogram 4000 with the new frame installed. Beolovely!!:

Beogram 4002/4004 Keypad Restoration Service

One of our long-term projects has been to come up with a process to restore Beogram 4002, 4004 (and 6000!) keypads to their original state. We made several attempts over the past 3 years to replicate the original process and were met with many failures. Most steel will bend and just stay bent when buttons are pushed/flexed. Most cutting methods cannot create details as thin as the small gap between the keys. The coating is a challenge, to find something durable that has a similar overall effect.

We have finally landed on the best combination of materials and procedures to fully restore the keypads and are now offering this service in the Beolover Shop. Be sure to read this full article for information on what the restoration process is and how to remove your keypad to send it to us.

Many keypads will start off with keys in this condition:

We originally planned on a process to restore/re-coat the original steel plate, but there are several problems with trying to restore the original. The first problem is that if you strip away the old coating, the areas that had worn through will stand out in certain lighting conditions, even when they have been re-coated. The second problem is that you will likely have some waviness in the material from the extraction process, the glue especially in DC models is very tough and the plate can easily get bends where you don't want them. The third problem is that if anything doesn't go perfectly, you have no backup plan. Below is an extreme example after the coating was stripped, there is waviness everywhere:

Now that we have new keypad plates, we don't have to think about the originals, which is quite a relief. There is no difference between the original keypad with a new coating on it and our new keypad with the same coating. There are minor differences between the new coating and the original coating, which we feel are worth it for the quality of the coating. Below on the left is an original in very good condition, on the right is a restored keypad with a brand new plate (more photos in this post).

How to Remove Your Keypad

In order to send your keypad for restoration, you will need to extract it from your Beogram. All that is needed for this is a flathead screwdriver for the single screw that holds the keypad in place. The process is easy, anyone can do it (takes 5-10 minutes). The first part of the removal process is also discussed in the Beolover Commander installation video (from 0:45 to 2:15). The Beolover Commander product is highly recommended to extend the life of the restored keypad.

1. Unplug the unit and hold the start key to drain any remaining power.

2. Carefully remove the stylus cartridge, taking care not to touch the stylus/needle/cantilever. Store in a safe place!

3. Slide the wood plinth forward to unlock the top plates. If it does not easily slide forward, locate the two black metal tabs at the back inside edge of the plinth (see photo below) that may need to be carefully nudged outwards while sliding the plinth forwards. These tabs are between the case and the plinth on each side.

4. Carefully begin to lift the right side of the right cover plate, taking care not to scratch the paint on the case or the keypad display module. Underneath, it is secured by a copper U-shaped piece and the left edge is tucked under the platter surround plate. Once it is slightly raised, it should slide out to the right.

Note: when re-installing this top plate, it is sometimes necessary to lift the top right of the platter surround plate so that this plate can slide back into the copper support pieces. Some earlier models have two pegs on the keypad module that this plate plate aligns with.

5. Lift the platter surround plate out, taking great care not to damage the case, the platter or the plate itself.

6. Place a cloth on top of the platter and carefully lift the gauge/display module out of the keypad by rotating it backwards slightly while pulling up on it gently. It should click out of place, then you can rotate it to the left and place it upside down on top of the platter.

7. Under the display module you will find a single flathead screw that holds the keypad in place, use a screwdriver to remove it.

Note: When you get your keypad back and have to replace this screw, it can be a little tricky to find the screw hole on the case, it will require a moment of patience as it is on the other side of the audio output board which may not be perfectly aligned with the hole as it moves around.

8. Lift the keypad gently, if you have a DC (551x/552x) model then you can disconnect the keypad harness from the main board. For an AC (550x) model, place a piece of paper or cardboard underneath the keypad as we will be removing the circuit board and don't want it touching the case or audio output board.

9. On back side of the keypad module, locate the triangular copper shape at the top center of the circuit board and lift it slightly so that the pin at the top is above the keypad frame, then slide the board sideways out of the keypad (and for AC models, let it rest on the cardboard/paper you placed previously).

Your keypad should now be free from the unit and ready to mail to us. It is best to use a 8x8x4" box and some foam padding if possible, though there are many ways to pack it safely.

Our Restoration Process

Each keypad sent to us will go through the same procedure to restore it. The first step is to warm up the keypad, to a level that will not affect the black paint of the screen printed lettering, but will loosen the glue and make it easier for us to remove the original pad and back pieces. Once it is warm, we can start with a large flathead between two of the keys, a small turn and they will break loose:

Lifting the original steel pad is challenging, especially on DC models where it can be very stubborn, eventually it will let go:

Nest, we have to clear all the old glue:

We re-paint the black edges under the keypad:

Optionally, upon request, we can also take this time to paint the area below the keypad so it cannot be seen from above when looking down at the Beogram:

We then use high bonding strength adhesive on both the keypad frame and the steel pad and join them. Here is the new keypad plate before it has been glued down:

Replacement Back Pieces

The last step is to install the back pieces. The original back pieces are not ideal for several reasons. One reason is that they are not quite thick enough, especially the 550x models which have 2.85mm back pieces. The later 551x/552x have 3mm back pieces. The original pieces have a tendency to "pull the keys down" below level, here you can see an original keypad - exactly as B&O left it; the START key (and all the other keys) are actually "below" the level of the sides:

We wanted to fix this so the keys would always be level and so we could use a consistent-level high bonding strength adhesive tape, rather than contact cement (which introduces variability to the level). We have created new back pieces which are 3.15mm thick and leave the keys perfectly level every time:

We were also able to reduce key travel slightly so that the buttons are closer to the electrical contracts. The new back pieces are made from black nylon and look like this (not that anyone will ever see them!):

Optionally, upon request, we can return the keypad without any back pieces installed if you wish to glue the originals back.

Finished!

Once you receive the keypad back from us, simply follow the keypad removal steps in reverse. The final product, once installed in your Beogram:

Beogram 4002 (5513): Update of a 2018 Restoration to the Current State-of-the-Art and a Test Spin with the Beatles

One of my customers recently offered me a Beogram 4002 (Type 5513) that I had restored for him in 2018 for buyback. The unit still looked as pristine as when I sent it to him with an almost perfect original keypad, great aluminum surfaces, an almost flawless platter and a pretty good original rosewood plinth. So I needed about a microsecond deciding to want it! No better deal than buying an original Beolover restoration!...;-).

But when I consulted my pictures and posts that I made during the restoration of this Beogram, I realized that I had made quite some progress in the last 6 years, and that my current restoration process is considerably more evolved than in 2018. I worked by now on about 100 Beogram 400x, and I think I developed quite an idea about what they need to be reliable again and perform like new.

So I decided to update this Beogram to my latest state-of-the-art before offering it for sale via the Beolover Store (It is for sale now, please, see here). Read below what I upgraded.

This is how this beautiful deck looks like now:

This unit still featured a sensor arm LED implant based on a flex PCB that was hand made:

I removed it. This shows the flex PCB next to my current LED assembly (left):

One of the issues of the 'folded flex board' approach was that sometimes the LED would be misaligned. My current design comes with an alignment aid. It is simply inserted into the bulb compartment before soldering the LED board in:

The LED board rests on the alignment piece and so the LED position becomes well defined:

I also made some progress with the LED assemblies for the RPM panel above the keypad. This shows it removed, flipped on its face and the bulb covers removed:

The two small boards in front of the panel are my current LED designs. Out with the old

And in with the new. The new boards solder directly to the solder points that were used for the original bulbs:

The boards basically become extensions of the PCB that supplies the bulbs with current:

Of course the bulb covers still fit after this implantation:

When I took out the RPM panel for the LED exchange I could not help noticing that one of the plexiglass mounting pieces had deteriorated:

Luckily, there are newly made replacements available. All that needs to be done is transferring the metal springs to the new part:

Then it can be installed:

I realized that in 2018 I still did not replace the power transistors by default. I guess I still had not yet experienced enough failures at that point...;-). It is best to replace the two Darlingtons that are on the solder side of the main PCB while it is still installed. This shows IC1 that regulates the 21V power rail. It is usually a TIP120:

I replace them with their higher current cousin TIP102:

One needs to add a 100nF capacitor at the emitter when installing new Darlingtons in this circuit configuration. Otherwise there may be high-frequency oscillations superimposed on the power rail that can trick the record detection circuit into believing there is no record, even if there is one. This disables arm lowering.

The other TIP package on the solder side is IC4, normally a TIP 125:

I usually replace them with stronger TIP107s:

Then I took the board out and replaced the H-bridge transistors, the platter motor transistor and the sensor transistor. I also gave it a 5MOhm biasing trimmer. 

Another item that I updated to a more professional design is the main reservoir capacitor. I used to replace the double capacitance types that a significant fraction of Beogram 4002 featured with a 3D printed assembly containing two separate capacitors (the original dual-capacitance cans are not available anymore):

My current design is based on a circuit board. It can be used for both single and dual-capacitance setups, i.e. works in all DC motor 4002 and 4004. This shows it implanted as a dual-capacitance set-up:

This board also replaces the separate rectifier for CD-4 board power supply that is bolted on to the original  capacitor can.

Another item that I improved are the carriage pulleys. In 2018 I still bought them from someone else, but they were often a bit wobbly. Since then I started using my own manufacturer, who seems able to hold tolerances a bit better. This shows one of the Beolover pulleys installed:

Next came the transport lock bushings. My current design is a bit 'leaner'. This shows the removed old bushings (back) together with a set of the new ones (front):

The new design leaves more room for the movement of the floating chassis. It can be difficult to adjust it in the first place, and even if perfect centering is achieved, chassis movement still can result in banging into the surrounding structural components. The new bushing design maximizes the translational range in all three dimensions before such interferences happen. This shows one of them installed:

Their two-component design makes for an easy process: Just push one in from the bottom and one from the top (note that some of the orifices in the chassis are a bit larger than others, i.e. it can be advantageous to add a couple drops of superglue gel to the inserted surfaces to ensure proper seating).

Another item that I did not understand fully back then was that often the solenoid plunger is magnetized a bit, and that can lead to delayed solenoid action when the arm is lifted. That can be a bit ugly during the automated carriage return at the end of a record. Typically, in such cases the needle gets dragged for some distance across the record before the arm finally obeys and lifts up.

This shows the installed solenoid:

I removed it and disassembled the plunger:

Testing for residual magnetization is easy: Just use a (non-stainless) set screw and see if the plunger can move it without touching it. In this case the plunger did attract the screw:

I treated the plunger with a tape-head demagnetizer until the screw was not attracted anymore.

On to the damper: Back in 2018 I did not understand yet that the original gasket in the damper that seals it during arm lowering is often hardened and/or deformed. This often leads to inconsistent arm lowering speeds, including total loss of damping, and that can be somewhat hair-raising when the new $800 restored cartridge crashes onto the platter!...;-). So I started replacing these gaskets whenever I restore a Beogram. This shows the new gasket installed:

This shows everything back in place:

Another interesting upgrade I introduced since 2018 is always replacing the original fuses if they are still in there:

These cartridges often lost their integrity and this can lead to fuse failures without reason. This shows the removed original fuse cartridges:

One of them came apart. I installed new 250mA slow blow units:

At this point I thought I was done with the upgrades and so I adjusted the sensor transistor bias to the specified 4V

and then did a 24-hours RPM stability test with my BeoloverRPM device:

In its 'slow' mode it is able to log the RPM every 10 seconds into any serial port of a computer. This is the curve I had measured after about 24 hrs:

This is as good as it gets with the original DC platter motor of Beogram 4002 and 4004. The small longterm drift is a result of temperature changes in the ambient environment, typical of analog feedback control systems.

The BeoloverRPM has a second mode, 'fast', which generates a RPM measurement every single time a platter rib passes under the sensor. This allows the measurement of 'wow and flutter' (i.e. short term RPM fluctuations during individual turns of the platter. The first thing one notices when running the BeoloverRPM in fast mode is a repeating patter on the screen:

This pattern repeats exactly every 24 points which immediately suggests it is a measurement artifact generated by the platter itself (they have 24 ribs...;-). When measuring this pattern for ~60 turns the graph looks like this:

It is evident that the 24 points pattern is superimposed by a slower 'wavy' pattern. This wavy pattern corresponds to the actual RPM variations introduced by the analog feedback system of the platter drive. The DC motors have pickup coils that produce a small induced current every time a rotor pole passes over them when the motor spins. This signal is picked up by the control electronic and turned into a change in the motor voltage: Whenever the motor is below the setpoint speed, the voltage is increased a bit and whenever it is too fast, the voltage decreases accordingly. This results in a 'meandering' RPM behavior around the setpoint, hence the wavy pattern.

An evaluation of this pattern typically yields variations around 0.1%, which is 2x of the stated value stated in the manual of 0.05%. But we need to keep in mind here that back in the 1970s this was measured in an analog and likely less accurate way. So the difference may just be that, a discrepancy between measurement results of the same value obtained with two different methods.

At any rate these variations are way too small for detection by humans, who usually have difficulties discerning variations smaller than 0.7%.

Next I tested the carriage mechanism. I pressed start without a record on the platter. The expectation is that the arms sweep the entire platter in search of a record and then return home to base. 

This return home is caused by triggering the 'end switch (ES)' that is mounted on the small PCB that is underneath the plexiglass 'ruler' with the black bars that is bolted to the carriage assembly. 

In this case nothing happened and the carriage motor tried to push the carriage further than the switch. An inspection yielded a defective switch:

So I installed a new Beolover Carriage Position PCB for Beogram 4002 and 4004 (Types 551x/552x):

This board fixes all switch issues and it also puts the carriage position sensor on a modern footing: The board features a modern monolithic IR interrupter that is read out by an op-amp and Schmitt trigger based circuit. This yields a clean digital signal that reliably triggers the circuitry on the main PCB. The board also has a built-in LED that can be switched into the circuit for easy testing of proper carriage position detection. So one does not need an oscilloscope anymore for this task.

And now the deck seemed in perfect like-new condition again, and it was time for a test-spin!

I selected a recent addition to my collection, "The Beatles/1967-1970". A great sampler of their later work. I received this record (together with the matching "Red Album") from a gentleman who sold me his Beogram 4004 a few months ago. This album is a bit 'down the memory lane' for me since I last listened to it when I was in high-school!...;-). But I have to say I really enjoy listening to it again! Strawberry Fields Forever!

This shows this nice Beogram 4002 playing this contemporary album:

The Beogram seems to perform perfectly. A happy Beolover moment!

This deck is for sale now at Beolover.com.

Here a few impression of its looks: