Beomaster 8000: New uProcessor Crystals, Improved Decoupling of the uProcessors, and New IC Sockets

A Beomaster 8000 that I restored in 2014 recently returned to my bench due to some erratic behavior. Apparently it turns itself on spontaneously once in a while. While I was not able to reproduce this behavior for several weeks, I experienced it once with one of my own Beomaster 8000s, and there are some sporadic reports on the internet about similar issues.

Since I was not able to find anything wrong with this Beomaster, I am left with a hypothesis, namely that the uProcessor once in a while suffers from a power fluctuation or that one of the clock crystals is going bad, and that puts it into an unexpected state.

So I decided to replace the crystals and add 100nF decoupling capacitors to the power supply pins of the uProcessor IC, which only carry 1 uF from the factory. These days most of the time one finds 1nF in parallel with 100nF, which allows filtering a broader frequency range.

This shows the inside of the uProcessor can in original condition:

I removed the precious processor ICs from their sockets to ensure that they would not get damaged during the surgery. The crystals and also their capacitors can be charged with a high voltage out of the box, and that can damage the ICs. Therefore, I normally discharge all components against GND and I remove the ICs before working on this board. 

This shows one of the crystal setups in more detail:

I removed the crystal and their two 22pF oscillator capacitors and implanted modern crystals with their specified 18pF capacitors:

I also replaced the IC sockets, while I was in there:

This shows the an original and a new 2 MHz crystal in comparison:

When putting in the new 18pF capacitors, one needs to remember to solder one of the pins from both sides, since it serves as a via.

The final task was to add the 100nF capacitors. This is an easy thing to do since the 1uF cap is soldered between the 5V and GND pins on the back side of the PCB. This shows the original setup for IC3:

I simply soldered the 100uF cap in parallel across the 1uF cap:

After this step I put the board back in, and the Beomaster still worked...let's hope the new parts help suppressing the sudden-on phenomenon. Only the future will show...

Beomaster 8000: Replacing the Processor Crystals

A Beomaster 8000 that I restored a few years ago for an Australian customer unfortunately had to come back to my bench due to a malfunction of the processor board. It exhibited rapid-fire power on/off events, which are a sign for a dead crystal on IC4 (slave processor). See here for more detail on how I first ran into this issue on the 8000 that is in my living room. Anyway, here are a few pictures of how I fixed this one. This shows the original setup:

As you can see someone in the past soldered jumper wires directly to the pins of the microcontrollers (savages!) to bridge the often failing vias that connect some of the processor pins to other components on the board. I usually re-solder the vias instead of putting in jumper wires. The reason that these vias fail is poor initial soldering. In the 80s electronics manufacturing just started becoming fully automated and some outcomes were not as high quality as one would desire...

Anyway, here I had to deal with these wires since it is good practice to take the microcontrollers out before replacing the crystals. They can be charged with high voltages when you take them out of the packaging, and this can fry the ICs...and that essentially would mean that a donor Beomaster 8000 would need to be procured. I was able to remove IC3 (right) without trouble but IC4 on the left was stuck. When the wires were soldered on a lot of solder was used and some of it penetrated the lower regions of the socket. I unsoldered the affected legs of the socket and then pulled the IC out. It came out including the three soldered-on pins from the socket. Once it was out it was easy to clean the pins up since I was able to get to the solder without having to heat everything for a dangerously long time...much better for the survival of the precious IC.

I implanted a new IC socket

And then exchanged the crystals and oscillator capacitors (the original ones are 12 pF while the modern crystals need 18 pF caps):

I elected to not solder the jumper wires back on to IC4. Instead I fixed the vias. I left the wires on IC3 in place since every time one heats up a pin on an IC there is a slight chance that damage can occur...

Once it was all back together, I fired the Beomaster up for the first time, and it came on normally, no more relay rattling etc...so I hope this did the trick.

Beolovely!

Beomaster 8000: Microcomputer Board Reworked

Today was the day to work on the Beomaster 8000 microcomputer board. On an earlier post I noted that this Beomaster is an earlier model unit so it has the first generation Beomaster 8000 microcomputer board. The later model boards are direct replaceable but are definitely different. I will show an example of the late model microcomputer board at the end of this post.

One thing I don't like about these first generation boards is the metal enclosure around the processor ICs. It is necessary for shielding of course but these early boards are difficult to open up. B&O soldered metal tabs to connect the upper and lower sections of the metal shielding box. The problem is the tabs also solder to the ground plane of the the microcomputer board. Because tabs are soldered to the ground plane a lot of heat is required to melt the solder for removal. When people try to open the shield box up to get to the components they usually damage the traces on the board where the shield box mounts to it. This Beomaster has been serviced before and falls into the category of having damage to the mounting points.

The following picture shows the two different Beomaster 8000 microcomputer boards. The board for this project is the bottom board. Notice that the metal shield box on the top board is pressed to fit while the bottom board is soldered together (red circles show some of the solder locations).

Here is the microcomputer board for this project as it was before restoration.

Here are the covers removed.

The early model boards used a metal bar for a heatsink on the two main processor ICs (IC3 & IC4).
The places circled in red show the damaged places on the board where the mounting tabs for the case originally were. The last person to work on this unit was able to re-seal the box with solder but I won't want to do it that way again as I don't trust the grounding with those missing tabs.

This is similar to what B&O used in the metal shield box for the Beogram 8000 and 8002 turntable processors. A strip of special tape is used to keep the messy thermal paste off the ICs themselves.
Here is the component side of the board with the thermal tape removed. The restoration task will be to replace the 22uF electrolytic capacitor to a 105°C, high reliability type. Also the 1uF tantalum. Following Beolover's lead I will also replace the two oscillators for the processor ICs (X1 & X2). Those will require changing two capacitors on each to new ones that match the new oscillators (18pF). Last, I will also reflow the solder joints on all of the connects and board vias as those have been known to have hidden problems.

For the oscillator capacitors, this original board has two 12pF capacitors for each oscillator. One oscillator has its two 12pF capacitors on the component side. The other oscillator has them on the trace side. 

The later model microcomputer boards have mounting holes for the oscillator capacitors on the component side.

Here is the microcomputer board with its updated components.

CAUTION: As Beolover noted in his Beomaster 8000 restoration, messing with the crystal oscillators could be a risk to the two processor ICs (IC3 & IC4). As a safety measure I shorted the two oscillator leads together and removed the processor ICs from their sockets (using an ESD grounding strap of course). With IC3 and IC4 safely out of the way I performed all of the rework to the board.

After doing the tedious solder reflow work it is time to put this board back together. I re-inserted IC3 and IC4, then prepared the heatsink thermal tape and compound so I could remount the top and bottom covers.

Now to solve the broken ground points where the metal covers mount. I have solved this problem before by using copper tape the way metal bands are wrapped around crates for shipping. The copper bands fit perfectly in the board slots where the mounting tabs go. I solder the bands together on the top and bottom of the metal box. I also solder the copper bands to the metal box so the grounding will work as it is supposed to. To tie the box to the board ground planes I run a copper tape strap to the main ground lug (both on the top and bottom side of the board).

The result is not the prettiest thing (rather ugly really) but it is solid and works. The copper straps securely keep the box together and I can measure good ground continuity anywhere on the board. If I have to do this type of repair again I think I will apply strips of copper to the lid pieces first (to use as solder pad anchors if you will), then tie the bands in with those anchors.

Removing the assembly should be pretty easy.  I didn't solder the copper straps on the sides where the top and bottom pieces meet. To open this box up again I only have cut the straps along the sides, de-solder the two main ground straps and the box will open up. In the picture below I marked the cut points with the red arrows. Note the green arrows. Those show the previous repair where someone fixed the broken ground connection by scraping away some board coating at the ground plane and soldered that right to the metal housing. While that should work I think it is risky because it is susceptible to the solder joint cracking if the metal box is stressed. I like the straps better because they provide a much more reliable connection to the ground plane of the board. The grounding strap also has a little give in it so it can withstand any movement.

Now it is time to test this updated board in the Beomaster 8000. 

I am happy to say that works great. Now it needs exercising for a while to make sure there aren't any hidden problems.

For reference - Here is a Beomaster 8000 later model microcomputer board I restored earlier.

These cover plates are easy to remove as they just press-to-fit. Not the rectangular tabs in the top cover plate. Those make contact with the two main processor ICs (IC3 & IC4) to transfer heat away from the ICs and to the metal shield box for disapation.

Here is the component side (restored). Both sets of 18pF capacitors for the new crystal oscillators are on the component side of this board.  Also notice that this board has a frame for the top and bottom covers permanently mounted to the board. This frame provides all of the structure. There is no strain on the board layers and no soldering of the shield box. Way nicer for servicing.

Here is the trace side of the board. All I did on this side was reflow solder joints.

Now I need to get back to the assembly of those Beomaster 8000 display modules so I can complete the last board.

Beomaster 8000: Step Six - Rebuilding the uProcessor Board

*********************************************************************************

ACHTUNG: Please, note that the crystal exchange procedure described here can damage the microprocessor chips on the uProcessor board. It is recommended to remove the chips before replacing the crystals and the capacitors. Due to the inherent capacitance of these devices, high voltages can be present between their terminals, which upon release, can burn out the gate that forms the oscillator together with the crystal. Make sure you short circuit the leads of the parts before installing them.

*********************************************************************************

After completing the display board rebuild with new SMD LEDs in the displays themselves, as well as replacing the incandescent light bulbs in the indicator cabinets with SMD LED based assemblies, it was time to do the uProcessor board.

The uProcessor PCB is the main source of operational trouble due to a few shortcomings in its design. In the 80s double sided PCBs were only beginning being standard in consumer electronics, and the earlier 8000s have boards without through plated vias. This made it necessary to insert metal plugs into the vias and solder them. Many 8000s have startup trouble due to these solder joints cracking. Luckily this Beomaster 8000 already had an updated board where the via plugs were replaced by soldered in colorful wire jumpers. This shows the processors, the jumpers and the crystals:

Another source of trouble are crystals that stop working reliably. This can cause severe malfunctions in the start-up circuit potentially damaging the main transformer. I started replacing them as a standard restoration item since I had a malfunction in one of my own Beomaster 8000s that caused the uninterruptible power supply (UPS) I run this Beomaster on to go in protection mode. This indicated that very high currents were flowing in the main transformer while the malfunction occurred. More here about this phenomenon. When the exact same condition happened in another Beomaster that I restored a few years ago, I took it as a sign that this is an issue that may be prevalent in many 8000s.

So here we go. This shows the old big crystals replaced with smaller modern 2MHz units. 

You can pretty much use any 2MHz crystal, as long as you also install the correct capacitors. The original crystals use 12pF capacitors to resonate, while modern units usually have 18-22pF. If you look closely, you will see that I also exchanged the two small capacitors directly underneath the crystals. They are now 18pF. 

I also replaced the two electrolytic capacitors on the board with high quality 105C grade Japanese units.

Here are some more photos: This is a detail shot of crystal X1

and a picture showing the solder side:

This shows the solder side of X2 for reference:

After this I plugged everything back together, and it was time for a test run! This shows the display board in all its glory:

Beautiful! Very Beolovely! By the way: One can 'provoke' the clipping light (left most) to come on, too, by setting the tuner to a station and then cranking up the volume. Above about 5.0 the clipping light usually comes on (make sure you do not have speakers connected during this test...;-).

On to replacing the speaker switches!

Beomaster 8000: Rattling Main Relays and a Broken 2MHz Oscillator Crystal

*********************************************************************************
ACHTUNG: Please, note that the crystal exchange procedure described here can damage the microprocessor chips on the uProcessor board. It is recommended to remove the chips before replacing the crystals and the capacitors. Due to the inherent capacitance of these devices, high voltages can be present between their terminals, which upon release, can burn out the gate that forms the oscillator together with the crystal. Make sure you short circuit the leads of the parts before installing them.
*********************************************************************************


Oh well, nothing lasts for ever!...The fully restored (2011) Beomaster 8000 that I am using in our living room recently broke with a rather strange fault. I woke up at 4 am in the morning to the beeping of an uninterruptible power supply (UPS) in distress. First I thought we had a power outage, but it turned out that the UPS turned itself off due to a malfunction in the connected Beomaster 8000 (I run all my B&O equipment on UPSs to reduce the risk of damage from grid voltage spikes and the like). I restarted the UPS and immediately the main transformer power relays of the 8000 started going on/off in rapid sequence. And then the UPS caved in again, turned itself off and started beeping again. This immediately pointed to a severe fault in the 8000 since a UPS only shuts down if the current drawn exceeds the rated 10 or 16 amps. This suggested to not try running the Beomaster again before having a look inside.

I shut everything down and went back to bed. The next morning I swapped the 8000 out with one of the other 8000s I have around the house in less important locations. A few days later I opened the malfunctioning unit up and ran it. First everything was normal but after a while it started to rattle again. I switched it off immediately and started wondering what might cause such behavior. First I thought there is a problem in one of the outputs triggering the protection circuit. The reason was that I initially detected 15V at the output of the protection circuit (collector of 6TR15) during relay rattling phases. The 15V pull up the base of 6TR11 via 6D12, which then causes the relays to go off, cutting power to the main transformer. This is usually caused by malfunctioning output transistors or overheating of the outputs. However, in this case the 15V were rather a consequence of an entirely different root cause.

After some playing around with the unit and enjoying a few more UPS shutdowns and relay rattling events, I finally figured out that the crystal oscillator of 9IC4 on the main processor board had some issues. I figured this out due to two phenomena:

1) After one of the rattling episodes I was not able to turn the Beomaster back on. The standby LED was lit, but no more reaction to the keypad. Panic ensued since I thought I had accidentally fried the processor or some other disastrous event occurred. So I did a processor self-test by pressing the Monitor key first and then additionally the on/off bar on the keypad. This initiates a self-test sequence. And I finally saw my first error code on a 8000: TE8!

The manual yields a rather cryptic cause for this error:

"Error TE8:    Defect IC:9IC4(RAM)      Or short pin to chassis: 10-11"

Initially I thought I had really fried the slave processor that is responsible for the relays...but then ,after my initial panic subsided and normal brain functions kicked back in, I thought that my maxim "silicon usually dies last" should apply here too, since even a complete main power supply failure and dramatic short circuit etc...would have a hard time killing anything on the processor board. There is simply no direct connection. Furthermore, when I tried to turn the unit on again after an hour or so, it worked normally again indicating a healthy state of affairs on the processor end of things (silicon either is alive or dead...rarely there is an intermittent state in my experience). Also, once it worked again, the error code went away and I got this from the self-test:

A happy Test Passed (TP).

2) The second indication toward the oscillator issue came from an oscilloscope measurement I made after the error code episode:

The probe was hooked up to the base of 6TR11, which is connected to Pin 16 of 9IC4 via some resistors. At this point the unit was working and there was no rattling from the relays at all. But you see some type of digital random oscillation of the signal that normally should be just near-0V when the unit is on. The digital character of the signal already suggested to me that there might be a processor issue...I watched this for some time and then it changed to a more aggressive behavior:

A much stronger deviation from 0V and still pretty random and close to switching 6TR11! And then it went completely crazy and the relay clicking started again as 6TR11 did its job. I did not save the measurement at that point since I was busy turning off the unit as fast as I could...;-). This finally put me on the right track, however, and I started looking into 9IC4. The same random signals were also directly visible at Pin 16 of 9IC4, just on a 0-5V scale since at Pin 16 we are directly at the source of the signal and there are no resistors over which the voltage drops. This told me that indeed the processor was going 'crazy' occasionally. This combined with the error information above that alternatively to a processor malfunction also Pins 10-11 could be 'grounded' finally made me see the light since the oscillator crystal is connected to Pins 10 and 11!

I wiggled the crystal a bit and indeed, I was able to cause the Pin 16 signal to change from completely quiet 0V to the above shown random signals and to 'crazy' causing relay rattling. So I guess what happened was that the clock of the processor started to have random hiccups causing timing issues while processing its firmware, which then caused the normally constant output at Pin 16 to become random triggering the relay circuitry.

I ordered a few 2MHz crystals from Newark (21M6819) along with matching 18pF resonator capacitors (46P6436) - the original crystals run on 12 pF capacitors, i.e. they need to be exchanged along with the crystal to get a proper oscillator signal. After a few days I received the parts and put them in. This shows the original crystal with the two capacitors (brown, right below):

This shows the board with the parts removed:

This is the new smaller crystal in comparison to the original one:

Like most components modern ones are considerably smaller than the original ones...and this finally shows the new units implanted (I also exchanged the 9IC3 components assuming both the original crystals were from the same batch...;-):

The exchange is straight forward, but the right capacitors are difficult to remove since one of their legs is fed through a hollow via, which is a bit of a pain to unsolder due to the very small space between lead and via insert. One has to hold the solder tip to the via to liquefy the solder and then pull the cap out...three hands would be great for this process...;-)

One more thing: It is crucial to cut the leads on the solder side of the board very short to prevent short circuits to the EMI can that encloses the processors:

I had them too long initially, and I got some really spectacular readings on the displays when I turned the unit on. Essentially, it became completely unresponsive showing some random zeros on all of the displays. Another near-heart attack...;-). Cutting the leads solved that problem...live and learn...;-) 

I am running the unit now for a couple days back in the living room and it seems the issue has gone away. Also wiggling the crystal did not cause the issue anymore. So I am assuming we are back in business with this lovely Beomaster 8000!

Remarkably, a few days after this happened to my 8000 I received an email from an Australian customer whose 8000 I restored a couple years ago telling me that his Beomaster developed the exact same symptoms (I am sending him a couple crystals and the capacitors plus some instructions...this is Beolove!).

So I am thinking that exchanging the crystals should become a standard part of any Beomaster 8000 restoration...two at the same time is a rather strange coincidence, which points to a systematic issue with these crystals. Probably the leads are delaminating from the crystal material after 30-35 years...

Time to get back to my other restoration projects!