Featured Post

Beogram 4002: Restoration of DC Motor Video Published - Check It Out!

By popular request (really, I got quite a few emails about this!...;-), I finally completed my Beogram DC motor restoration video! It demon...

Tuesday, November 20, 2018

Canada Beomaster 8000: The display board is ready

The display board out of the Beomaster 8000 from Canada is completed now. It was the final piece before starting the functional testing phase of the restoration project. Once this board is installed back in the Beomaster I will recheck the power supply voltages and the output amplifier assemblies.

After installing new smd LED devices in the segmented display assemblies and bench testing them I sealed the display modules back up and checked them one more time on the bench.

They look much nicer now that no stray light is able to get out.

Before I installed the segmented displays onto the display board I changed out the four indicator lamps on the board with four Beolover LED indicator assemblies. The change to the Beolover LED indicator assemblies requires the removal of four 220Ω resistors on the display board (R34, R36, R38 & R40). They are highlighted in the picture.

Now the display board can be reassembled for installation back in the Beomaster.

Saturday, November 17, 2018

Canada Beomaster 8000: Rebuilding the displays

I have the new smd LED components installed in the Beomaster 8000 segmented display boards and I am running a burn-in test on them.

It is a little difficult to capture (in a photo) what the rebuilt displays look like without the top lens cover on but the purpose of this test is to make sure all of the segments work. I usually run this burn-in test for twenty-four hours. That gives the new components a decent amount of time to fail if they are going to. I would hate to bundle everything up then have to disassemble the displays again.

To get to this point I removed the display modules from the Beomaster 8000 display board.

There are white and red tabs on the rear of the display board modules that hold the light guide and lens cover in place. Those tabs were originally part of the plastic sub-assemblies and melted down to secure the full display assembly. Since those cannot be removed and reused I have to just cut them off and use a different way to re-assemble the display modules after I rebuild them.

Here is the channel balance display board with the light guide and lens cover removed.

You can see these Bang & Olufsen display modules were very custom and the tiny, smd-like LED components cannot be replaced. We could locate the LEDs that have failed and just change those. However, that would result in a noticeable difference in the red illumination because the new smd LEDs are not exactly the same as the originals. They are very close but not the same. So it is best to change all of them. Another equally good reason to change all of the LEDs is because you can bet that if you only changed the failed LEDs the remaining good ones will likely fail as well. It's definitely better to get everything done while the Beomaster is open.

Here are some pictures from previous Beomaster 8000 display repairs that I use as a reference for this task.

This first picture shows the new smd LED component next to the original B&O LED.

This picture shows identifier markings on the replacement smd LED component for reference.

The next step in the display rebuild after removing the light guide and lens cover is to cut out the original LED component.

I gathered a number of my Bang & Olufsen display modules that I have been wanting to rebuild and hope I can take the opportunity, while I am display repair mode, to get them done as well.

To prepare each display board for the new smd LED components I cut off the small solder pad that the original LED cathode rested on. That makes the gap between the anode and cathode pads better for the new components.

It is important to "tin" the solder pads before mounting the smd LED components. That will ensure a good, reliable connection.

Now for the actual mounting of the new smd LED components.

I have to say that this task is one of the most time consuming restoration tasks and requires a lot of patience to handle the smd LED devices while soldering them in place. The technique I use (and recommend) is to mount the smd LED components just on one lead. The reason is because the LED component must fit just right in the plastic light guide of the display module. There are always several LED components that I have to reposition slightly to correctly function inside the light guide. That also means the soldering process is to first position the smd LED with tweezers, solder one side to secure it, check the fit with the light guide, reposition if necessary and finally, solder in the other side. Quite labor intensive and hard on the eyes.

Here is a fit check of the smd LED in the light guide.

Once all of the new LED components have been fit checked I finish soldering the LEDs to the board.

Here is the completed FM tuner frequency display board.

On to the remaining three modules.

Another tip is to have a reference picture or (even better) a spare display board to use as a reference when mounting the new LED components. This task is very tedious so be disciplined to check the polarity carefully with each LED mounting.  It is also another good reason to only solder one side initially.

Here is the completed LED replacement on the Beomaster 8000 selected source display board.
Note: The glare from my lighting off the no-clean flux residue makes a couple of the solder connects appear shorted.

The volume level display board is next.

Finally the channel balance display board. This board is the easiest to position the smd LED components on because it doesn't have pads where the LED mounts. However, the narrow trace makes soldering in the LED components a little tricky and this board is usually the one to have a connection problem.

Once the twenty-four hour burn-in test is complete I can reassemble the four display components and wrap up the display board. Then it will on to functional testing with this Beomaster.

Saturday, November 10, 2018

Texas Beomaster 8000: Testing the Output

I left off with the Texas Beomaster 8000 playing music for several hours as a check of the amplifier rework and a burn-in test. That test play of the Beomaster went great. I ended up running the Beomaster amplifier for eighteen hours without a hiccup.

Before connecting up some test equipment to measure the Beomaster output amplifier total harmonic distortion (THD) at 1KHz and maximum output I needed to take care of a couple of minor things.

The first thing was to adjust out the "clicking" noise that was occurring when altering the volume level. This is something that can happen due to DC offset voltage in the volume control circuit. The Beomaster has separate left and right trimmer resistors to remove that offset. I typically have to adjust those trimmers after replacing components on the Filter & Tone Control board (as part of the restoration).

Along with the left and right trimmer resistors for the DC offset there are two fixed resistors per channel that can be used or removed to extend the range of the offset adjustment.  In this case I had to remove the R145 resistor in the left channel and the R243 resistor in the right channel.

I verified the Beomaster 8000 remote control functions work so I re-attached the metal shielding box for the remote control receiver circuitry on the power supply board.

It was also time to finish re-assembling the Filter & Tone Control board. It is fairly common to find broken plastic tabs that hold the push button board. I repair those using black hot glue. That glue will do the job of securing the board but can be peeled off later if needed.

On the problem with the FM tuner not functioning I found a loose wire on the cable between the FM board and the microcomputer board. An easy fix and the tuner is working great again.

Now to check the output amplifiers and their rated maximum 100W output across my 8Ω dummy loads.

For this test I will use a signal generator to apply a 1.5Vrms, 1KHz sine wave to the Beomaster TP1 input connectors (L & R).  The speaker outputs will go to my 8Ω dummy loads where I will measure the output voltage with a bench DMM and the THD with an audio analyzer.  I also connected my Fluke multi-meter across the emitter resistors to see what the DC voltage levels do there when the amplifier is driving the speaker load.

The THD is measured using the QuantAsylum QA400.  Here is the left channel output amplifier THD with 9 Vrms output across the dummy speaker load. That is about 10W. That was with the Beomaster volume control set to 5.0.  The THD was about 0.013% which is a good number for this amplifier.

Continuing on I increased the Beomaster volume until I got 28.2 Vrms across the speaker load. That corresponds to about 100W which is the rated output of the Beomaster.

I get just under 0.05% THD (not weighted).  While I can't precisely duplicate the methods Bang & Olufsen used back in 1980 to make these performance measurements my results appear to be in line with what this amplifier is advertised to do.

The right channel measures pretty close to the same.

Friday, November 9, 2018

Texas Beomaster 8000: Readjusting the output amplifier assemblies - Part 4

It took a few detours but the Texas Beomaster 8000 finally reached the start of some functional tests.  After the stand alone bench testing of the left channel output amplifier assembly...

I installed the tested output amplifier assembly back into the Beomaster 8000 cabinet.

I rechecked the no-load current and DC offset adjustments of both the left and right channels one more time. 
After verifying those settings were good I got back to playing some actual music through the Tape 1 (TP1) source of the Beomaster.  Both channels sound good.  I will do some real measurements a little later but for now I want to run a twelve hour burn in test of the amplifier assemblies playing music.

I connected a music loaded ipod nano to the Beomaster TP1 source input and started with a volume level of zero. I increased that to a volume level of 4.3 then 5.0 ... playing through my trusty workshop Beovox S55 speakers. 

This is another major milestone.

There are still some restoration tasks though. I hear some typical clicking in the speakers when adjusting the volume. That is due to DC offset in the volume circuit. There are trimmers for the two channels that will have to be adjusted. The FM tuner also does not work correctly.  I will have to determine if that is due to a cable or if there is a component problem on the FM board.

Thursday, November 8, 2018

Texas Beomaster 8000: Readjusting the output amplifier assemblies - Part 3

The left channel output amplifier assembly on the Texas Beomaster 8000 is working again.

I removed the heatsink assembly darlington output transistors (TIP 141 and TIP 146 devices) and replaced them with spares I had in stock. I decided it was better to replace the full set of output transistors and I decided to replace the two output transistor emitter resistors.

The replaced parts are outlined in the photo below.

Here are the replacements installed.

Now it was back to the bench to check and readjust the no-load current and DC offset.
Here is the test setup for that.

There were no problems on those adjustments.

It is worth noting that when output amplifier components are replaced the no-load current is often way off. After I ramp up the ±55 VDC rail voltages the current draw on those supplies should be around 30mA.  As I turn the output amplifier on by slowly ramping up the +15V I monitor the current draw of the ±55VDC bench power supplies. When the no-load current trimmer is way off the current draw on the ±55 VDC supply rails will sometimes go beyond 130mA way before the +15VDC control voltage is reached.  If that happens I turn no-load current trimmer to bring the ±55VDC supplies current draw back down to 100mA.  Eventually I get to the full +15VDC control voltage with the ±55VDC supply currents at around 90mA to 100mA and the voltage across the emitter resistors at 18mV.

Now for what I learned on this whole exercise...
It is worth doing a quick functional check of the output amplifier assembly to make sure it can actually amplify an input signal.

The setup for this test is to remove the jumper I temporarily installed across the audio input connector and plug in a cable to my audio tester (the HP 8903A audio analyzer in this case). The output amplifier ± speaker connections are connected to my 8Ω dummy load where I measure the voltage with an oscilloscope (through a differential probe).

I begin the test with the audio analyzer outputting a zero amplitude sine wave.  I raise the amplitude on the signal in 1mV increments to 20mV while monitoring the output on the scope.

Here is the test setup again. I have marked the input signal and output signal.

Finally the left channel output amplifier is showing a good, amplified sine wave at the speaker output.

Now comes the fun part of re-installing the output amplifier assembly back inside the Beomaster cabinet.  Then I can get back to functional tests on this unit.

Wednesday, November 7, 2018

Beogram 4002: DC Motor Restoration and RPM Stability Test

A Beogram 4002 DC motor arrived from Australia for restoration. It exhibited the usual RPM variations due to dry bearings. This shows the motor as received:
I extracted the bearings for oil infusion under vacuum. This required disassembling the motor:
The bearings are the two small donuts on the black pad.
I immersed them in engine oil and pulled a vacuum:
Immediately vigorous bubbling started from the bearings, indicating that the vacuum drew the air from the porous Oilite brass material to make room for the diffusion of fresh oil into the bearing.
These bearings were particularly 'thirsty' and it took about 72 hours until the bubbling stopped. After the bubbling stopped I extracted the bearings 
and put the motor back together for an RPM stability test. I used my 3D printed assembly tool for installing the top bearing:
I installed the motor in one of my Beogram 4002s for measuring the RPM stability over 24 hrs using my BeoloverRPM device:
This is the curve that I measured:
This is as good as it gets with the Beogram 4002 DC motors. This motor is ready for its trip home to Australia!