Featured Post

Beolover SyncDrive: DC Platter Motor Replacement for Beogram 4002 and 4004 (Type 551x and 552x)

Late Beogram 4002 and the 4004 (Types 551x and 552x), which have DC platter motors instead of the earlier synchronous AC motors usually suff...

Sunday, October 31, 2021

Beogram 8000: Wrapping up the Workshop Beogram with speed sensor replacement testing

It is time to wrap up testing with the Workshop Beogram 8000 for now.  
The test platform has helped me restore and repair five sets of Beogram 8000 circuit boards along with using the quick connect/disconnect socket on five uC (2IC1) devices.

The Workshop Beogram also gave me a way to test the speed sensor (OPE2).
I discovered that I had two faulty speed sensor devices in my collection of Beogram 8000 parts so this last Workshop Beogram 8000 blog post will show the result of my search for a speed sensor replacement device.

The original two types of speed sensor (OPE2) devices in the Beogram 8000 are these two Bang & Olufsen part numbers - 8330007 (the early type sensor with resistors R46 & R47 on PCB 1) and 8005067 (the later type sensor with resistors R2 & R3 on the speed sensor assembly).
I don't know the story with those part numbers but the 8005067 assembly is what is called out in the later model Beogram 800x turntables (BG8000 & BG8002).  Maybe B&O had reserved that OPE2 part number but didn't release it until later.

Both sensor devices were similar slot type optical switches.  They consist of an emitter that is an IR LED light source and an NPN photo transistor sensor. They are a single assembly with four leads.

Here are the two types of Beogram 8000 speed sensor assemblies.

Here are the wiring diagrams for both speed sensor assembly types.

I was unable to find data sheets for either type of speed sensor that B&O used for these two types of slot type optical sensors.

However, I did find a few good candidates that are the same size and have the same pin configuration (same layout of the leads).

To check these sensors out I have one more bench test mode for my Workshop Beogram 8000.  That is an empty floating chassis with a platter hub/tachometer and the mount for a speed sensor.
Electrically I run the wires from the speed sensor I am testing to a breadboard setup to measure the slot optical sensor output (with an oscilloscope).  The breadboard contains necessary current limiting resistors (R46/R47 or R2/R3).  I use a bench power supply to provide the +15 VDC power for the circuit.

The first tests are for a B&O 8330007 speed sensor device.

Because this is an 8330007 B&O speed sensor I am using current limiting resistors per the schematic. It also matches what was installed on the PCB 1 board for this type of sensor assembly.

Applying the +15 VDC power to this circuit and manually rotating the platter hub here is the oscilloscope measurement.

This 8330007 device is a working speed sensor that I have also functionally tested with the Workshop Beogram 8000 turntable.
The amount of current this sensor circuit draws with the +15 VDC power applied is 30mA.

The second test setup is with a B&O 8005067 speed sensor assembly. It has built in current limiting resistors (R2 and R3).

Since the 8005067 speed sensor assembly has current limiting resistors built into the assembly it isn't necessary to connect up the test components on the bread board like the 8330007 speed sensor.  I can just connect my scope probe and DC power supply directly to the sensor assembly.

With +15 VDC applied to the 8005067 sensor assembly I manually turned the platter hub again and measured the speed sensor signal at the photo transistor collector.  This sensor drew around 20mA while it was operating.

Just like the first sensor, the 8330007, this 8005067 sensor is also a fully working unit that I tested with the Workshop Beogram 8000 turntable.  

In the third test I tried out one of my replacement candidate slot type optical switches.
The device I chose for this test is the TT Electronics/Optek Technology OPB370N55.
Here is the test setup with that device.  It will require the use of the breadboard to provide the current limiting resistors.

To get a usable signal from this new sensor device I had to adjust the trimmers to around 2.4KΩ for R46 and around 6.8KΩ for R47.
Using those values I get a good, healthy speed sensor signal that uses about 10mA with the +15 VDC power applied.  That is half the amount of current used by the 8005067 devices and a third of the amount of current used by the 8330007 devices.

Now that I can see what values I need for the two current limiting resistors I can try out this replacement sensor device in a failed 8005067 speed sensor assembly and test it with the Workbench Beogram 8000 turntable.

Here is my test Beogram 8000 speed sensor with the new slot type optical switch installed.
I am using the 8005067 assembly for the test and I am using 2.2KΩ for R2 and 6.8KΩ for R3.

Before testing the new sensor device under full control of the Beogram 8000 I used my P4 test harness to do a manual check where I rotate the platter hub and measure the sensor signal coming off the sensor and at the microcomputer, 2IC1 pin 40.

That looks perfect.  The raw signal off the OPB370N55 collector is a nice strong signal that feeds the 1IC1 Op-amp before being routed up to the uC pin 40 input (via the signal conditioning flip-flop).

My final check is to connect up the new speed sensor to the Beogram 8000 P4 harness normally and see how it works under full control of the Beogram 8000.

For this test I utilized the Beogram 8000 "Turn" function button.
The signal at the uC pin 40 still looks great and the Beogram 8000 control circuitry has no problem locking in the platter speed.

It looks to me like the TT Electronics/Optek Technology OPB370N55 Slotted Optical Switch is a good replacement for the Beogram 8000 speed sensor device.

If you try it keep in mind that you must change the resistors (either 1R46 & 1R47 or R2 & R3) according to the version your Beogram uses.

There are two other Slotted Optical Switch devices by TT Electronics/Optek Technology that I will try.
OPB871N55 and OPB825.  They will require testing to determine their current limiting resistor values but I believe they could be used as a replacement device as well as the OPB370N55.

This concludes this round of Workshop Beogram 8000 projects.

I will be retooling my workbenches to begin some restoration work on some Beomaster 4400 units next.

Beogram 4002 (5503): Repair of a Cracked Solenoid Lever

I recently received a request from a customer to help with a cracked solenoid lever from a Beogram 4002 Type 5503. The AC versions of the 4002 use a mechanical switch to limit the solenoid current after it has been activated during the arm lowering process. This means that the solenoid arm is elongated with a tab that presses the switch down after the arm is lowered. In some later AC motor versions this extension is achieved by a plastic extension that is riveted to the solenoid lever.

I checked my photo album of earlier AC motor Beogram restorations, and I found several instances where this plastic lever extension was glued back together with epoxy. Here is an example (check the area under the solenoid wires, and the yellowish blob of epoxy is apparent):

Earlier AC-motor versions use an all-metal lever:
I guess the 'streamlining' of the manufacturing process went in the wrong direction with this plastic feature...;-)
It is important that this lever does its job, since without pulling the switch down, the solenoid will overheat after a few minutes and can get permanently damaged.

Anyway, this is the cracked part I received in the mail:

The crack is basically a stress fracture due to the rivet hole in the plastic. I carefully drilled the rived out
And designed an 'improved design' 3D printed replacement part. I made sure the part was fortified around the small bolt I used to fix it to the metal lever:

This shows the part installed in one of my Beogram 4002s. First in the 'arm up' position
and then in the 'arm down' position:
Beolovely! Let's see how long this plastic part will last! Time to send it back to the Beogram it came from and find out!..;-). Life is a learning curve!

Wednesday, October 27, 2021

Beogram Commander Remote Control: Successful Test of New 4002 (550x) AC Motor Version

I received a few inquiries about making the Beolover Commander remote module work with AC motor Beogram 4002s. My recent update of the Commander for Beogram 4002 and 4004 only works in DC motor Beograms since AC motor versions have no keypad plug socket on the main PCB. Their keypads are hard wired to the PCB. So I thought it would be nice to offer an AC version. But this made a redesign of the Commander module necessary.

This shows the redesigned version:

The main difference between DC and the presented AC version is an additional small breakout PCB (the long and narrow item on the picture) that can be soldered to the keypad wire-to-board solder pads. It feeds the keypad signals into a small connector that then connects to a matching connector on the Commander board. That is what the green jumper is for. The final version will come with a custom made jumper that has an appropriate length.

I thought it would be perfect to celebrate the arrival of the AC Motor Commander by implementing it for the first time in a rare early 5501 version (the first 4002 series they ever made after discontinuing the original 4000). This 5501 Beogram recently revisited my work bench from Australia due to the malfunction of the solenoid transistor and a non-working carriage motor. My customer agreed to let me delay return shipment to try out the AC Commander on his precious deck! Thank you very much!

The first step of the installation was to solder the breakout to the keypad PCB. This shows the original PCB after extracting it from the keypad assembly:

The wire harness from the main PCB connects to the pads up front from below. The two resistors on top are 22k pull-ups for the 33 and 45 RPM transistors, which seem to have been added after the fact due to operability issues in this early version. In later AC motor versions the PCB was changed and the resistors are placed on dedicated solder pads on the central part of the PCB. 
Unfortunately, on the 5501 they are a bit in the way of installing the Commander breakout board. When I tried to push them slightly out of the way for removing some of the solder on the pads, one broke apart:

I replaced them both with 20k resistors that I had in my stash. The 10% different value is not important, they only pull up the RPM switch transistors to make their ON/OFF states more reliable.
This shows the two resistors in place. They now allow some room for Commander breakout PCB:

The next step was to solder the breakout across the pads, which was easily done by putting some flux on the pads and then touching their orthogonal contact area with the soldering tip carrying a bunch of solder:

This is how it looks when the PCB is inserted back into the keypad:

The next step was bolting the Commander onto the main PCB and connecting the two connectors with the green jumper:

Then came the moment of truth, when I powered the deck up. And hurray, no magic smoke arose!

So I flashed the chip. The next step was pairing the Apple remote to the Commander board.

On the DC version Commander this involves pressing two buttons for ~6 min until the 33 RPM light comes on and the platter moves very briefly, indicating the completion of the pairing process.

I did the same here: And nothing happened! Panic arose, but pressing the Start button on the remote set the carriage in motion! So it seemed to have paired properly! 

And then it dawned on me: The AC versions does not have the 'record cleaning function', which is activated by pressing the 33 RPM key with the carriage in its home position. On the DC versions, pressing 33 RPM simply sets the platter in motion until the key is let go again. This is very convenient for wiping the platter before playing the record.
B&O probably introduced this function with the DC motor since the DC motor has a much lower torque than the AC motors, causing strong RPM fluctuations during swiping the platter. On AC motor units one can wipe the platter while the music plays, and there is almost no RPM slowdown. But with the DC motor one has to do it before actually pressing START for listening.

Anyway, so far so good! V1.0 seemed to work!

But the above made me think: It would be nice to have a record swipe functionality like on the DC models! And of course the pairing process feedback needed to be changed.

So I changed the code:
  • Instead of the 33RPM/platter twitch response to a completed pairing process, now the auto-repeat LED briefly lights up.
  • Now one can press the 'select' button on the remote for platter swiping like on a DC Beogram. This was done by replacing the absent swipe circuitry by a combination of activating START and then immediately "<" to stop the motion of the carriage. So it only moves a few mm until the 'shut off (SO)' switch is cleared that the deck can run the motor. Subsequent release of the select button activates the STOP function and the carriage goes back home and the platter motor goes off. What would we do without microcontrollers these days!!...;-)
Beolovely!! I estimate the AC Motor Commander should become available in 2-3 weeks when I will hopefully have received the proper jumpers for connecting keypad breakout and Commander board. Stay tuned!

Beogram 8000: Testing speed sensors with the Workbench Beogram 8000

The Workshop Beogram 8000 unit has already aided me in restoring a number of Beogram 8000 PCB 1 and PCB 2 boards.  One set was extra troublesome as it had some broken traces that were not visible.

This photo shows the boards finally working as I finished off the testing by playing a record.

One thing I discovered in my board testing was that a problem with any part of the speed sensor circuit could potentially cause damage to other components of the platter drive system.  I am referring to the speed sensor itself and the path the signal takes to the Beogram microcomputer (2IC1) pin 40.

Having a problem with the speed sensor is similar to attempting to operate the Beogram 8000 with the rotor removed (sometimes referred to as the sub-platter).  In both cases when the Play or Turn buttons are pressed the Beogram uC will start sending commands to turn the platter.  The uC expects a speed signal as feedback to regulate the platter drive.  When the speed sensor signal is missing (i.e. broken or the rotor is not in place to turn the tachodisc) the uC will keep trying to drive the platter.  Typically what happens is the F1 fuse (300mA, slow for USA model Beogram 8000 turntables) in the transformer compartment will blow and shut things down.  However, damage can also occur to the platter drive transistors 0TR2 and 0TR3.  It would probably also be possible to damage some other components in the platter drive circuitry.

A symptom of a broken sensor signal is that the platter will immediately ramp up to a very high speed when either the Play or Turn button are pressed.  If power is not removed quick enough there can be some component damage.

This problem existed in the PCB 2 assembly of this troublesome set of Beogram boards.
The speed sensor signal was good all the way into the PCB 2 assembly but failed to reach pin 40 of the uC IC due to a broken trace. 

The route of the speed sensor signal to the uC pin 40 can vary depending on when the Beogram 8000 was manufactured or whether it received a service modification.

This photo shows the two types of PCB 2 assemblies found in Beogram 8000 turntables with regard to the speed sensor signal and the uC pin 40.

In the early model Beogram 8000 units the speed sensor signal went from the sensor at the platter hub through an OpAmp (1IC1) then on to 2IC1 (uC) pin 40 via pin 2 of the ribbon cable that connects PCB 1 and PCB 2.

Later in production B&O engineers found that tolerance differences between 1IC1 and 2IC1 could cause periodic variations in the platter speed.  So to solve that problem a 4013 flip-flop circuit was introduced to stabilize the signal.

This makes a difference when trouble-shooting a suspect speed sensor issue.
The early model PCB 2 assemblies had a short trace from the ribbon cable pin 2 over to the 2IC1 (uC) pin 40.
The later models routed the incoming ribbon cable pin 2 over to the new 4013 flip-flop by a yellow wire then returned the conditioned signal to a solder point next to 2IC1 (uC) pin 40 by an orange wire.
So there are more places to lose the signal in the modified PCB 2 assemblies.

In the problem PCB 2 I had to repair, the short trace between the conditioned signal (orange wire) and 2IC1-40 was open. The break was not visible to me but it measured open circuit and the board's primary problem was no speed sensor signal.  Adding a jumper from the conditioned signal directly to 2IC1-40 solved the problem.

I found that it is easy to check 2IC1-40 with an oscilloscope and see if there is a good speed sensor signal. 
However, if the signal is not there then there will not be time to check for the signal and deal with the platter drive having no feedback.

The solution I came up with to make the easy measurement and protect the Beogram circuit components is a new speed sensor test harness for the platter drive cable.
That is the cable that connects to PCB 1 via the P4 board connector.
Disconnecting the P4 cable allows operating the Beogram 8000 without the platter drive in operation.
It is a handy mode when doing things like adjusting the record tracking sensor and setting the tracking force.  

I decided to make a P4 harness that only connected up the three wires for the speed sensor assembly.
This allows me to operate the Beogram 8000 in the mode where the platter drive is disabled except I have the speed sensor signals available to test.

Here is that harness ready to use with my Workshop Beogram 8000.

I have two oscilloscope probes in place.
One for the raw speed sensor signal right at the sensor.  The other will be connected to 2IC1 (uC) pin 40.

Now I can press the Play button or the Turn button and manually rotate the platter hub to check the speed sensor.

Here is a closer look at what the oscilloscope sees at those two measurement points for a properly operating speed sensor.

I now pre-test the speed sensor functionality of a Beogram 8000 using the speed sensor only P4 harness before trying out the full P4 harness.  That way I will get an early warning that there is a problem before risking the loss of a component.

Saturday, October 16, 2021

Beogram 4002 DC Platter Motor: Some Microsurgery on Disconnected Pickup Coils

I recently received a DC platter motor from a Beogram 4002 in Norway for some TLC. The customer already had done the 'oil infusion' process for the Oilite bearings, but after he put the motor back together and into the Beogram, the platter speed was much too high and could not be regulated anymore.

This shows the motor as received without pulley:

Runaway speed of a DC platter motor is a fairly sure sign (in absence of issues in the motor control circuit) that one or both of the RPM feedback pickup coils in the motor had been disconnected during the extraction/re-installation of the bottom bearing. This can easily be checked: Just connect a Ohm meter between the red and white leads on the plug, and if you measure an infinite resistance then one or both  coils are damaged:
If everything is alright the measurement should yield something between 35 and 45 Ohms. This varies across the various motor generations that were used. I opened the motor up and extracted the brush carrier that also holds the two pickup coils:
If you look closely, you will see the small disconnected wires on either side. Once this happens, a bit of microsurgery is needed. Basically one needs to unwind some wire under a microscope until about 15-25 Ohms can be measures between the two ends of the coil. Unfortunately, in this case this seemed to have been already tried, and the coil windings were cut to some depth. So I spent a bit of time with unwinding, tinning the end with solder (the wire is 'magnet wire' i.e. there is a coating on them that needs to be evaporated and replaced with solder before one can measure resistance reliably), measuring OC, unwinding until the wire piece fell off since it was cut off, and then anew. After a while I finally reached the still connected end and measured the proper resistance:
The wiring that I removed:
On the other coil it went similarly. Once this one was recovered, too, I put them back on the brush carrier.
This is the resistance I measured between red and white wires:
Then I reassembled the motor, installed it in one of my Beogram 4002s and measured the RPM with my BeoloverRPM device. This is the curve I measured:

This is as good as it gets with the Beogram DC platter motor! This motor is back in business!

Friday, October 15, 2021

Beogram 4002 DC Platter Motor Restoration

I recently received the DC platter motor of a Beogram located in Massachusetts for restoration. A quick bench test revealed the usual squealing indicative of dry Oilite bearings. This shows the motor:

I took it apart to get the bearings out:
The bearings are the two small donuts on the black pad. I immersed them in motor oil and pulled a vacuum. Immediately, strong bubbling started:
This shows that the vacuum pulls the air from the empty oilite pores making room for oil to diffuse into the bearing material. After about 48 hrs the bubbling stopped and I removed the bearings from the oil:
Then I reassembled the motor and installed it in one of my Beogram 4002s for a 24 hrs RPM stability test with the BeoloverRPM device:
This is the curve I measured. It shows the RPM in 10s intervals over 24 hrs:
This is as good as it gets with the Beogram DC platter motor. This motor can be sent back to its owner for another tour of duty in his Beogram!