By reflow I simply mean heat the solder joints up using a normal soldering iron and add a little fresh solder, if only for the flux, to make joints flow and "heal" again.

Yes, the board is double-sided with fairly few vias, but the board material and changing temperature plays a role in that the
connection breaks inside the via. Reflowing may not be a lasting solution and jumper wires is what B&O recommended in the tech note.
I seem to remember some boards that can't be reflowed, because there really aren't any solder pads at the vias, but merely a small hole.

Not unlike the Beogram CD X, where vias are actual metal rivets that has been riveted and then filled with solder. They develop cracks and
becomes intermittent, not around the rivet heads as one might expect but somewhere down through the hole.
The rivets provide ground potential from one side of the board to their respective circuits on the other side and, particularly in combinations, they can
cause the strangest faults.
The best (and sometimes only lasting) cure here is to suck out all old solder, feed a lead through the hole in the rivet (they are fairly large so
no problem) and solder the lead to tracks on both sides.
The same board principle can be found in some early 1980s car stereos - knowing that temperature changes has influence imagine the
problems they can cause here,

Martin

I can confirm first hand what Martin said.

Last year I was able to get several troublesome Beomaster 8000 Microcomputer boards functioning again. They had various failures similar to what you described. These type of things often cause owners to think the microcomputer ICs have failed. I don't think I have encountered a failed BM8000 microcomputer IC yet.  The problems have always ended up being the board itself - solder joints and vias per Martin's description.

I am also starting to think a good restoration step on these (Microcomputer boards and maybe other BM8000 boards) is to look at replacing the actual board connectors.  I have seen where other restorers have done that. I did that on one BM8000 Microcomputer board but just on the connectors between it and the Display board. One of those connectors was the final piece to getting one of my Microcomputer boards back to full functionality.

Initially I was also puzzled why several BM8000 Microcomputer boards of the same BM8000 type had jumper wires and others did not. Martin's information about the technical recommendation from B&O clarifies that. By now (2020) even boards that had some of that work have other connection problems so really the best solution is to carefully go over everything on those boards. I like Martin's suggestion of reworking the board vias with connecting wire to ensure layer to layer connection.

In the end I have always been able to finally get a BM8000 Microcomputer board working again so be patient.

-sonavor

Dillen wrote:

By reflow I simply mean heat the solder joints up using a normal soldering iron and add a little fresh solder, if only for the flux, to make joints flow and "heal" again.

💡 Thanks, now it's clear. It was just a language-specific misunderstanding. I'd just call that "re-soldering of the vias".
Yes, that process is fine!
However, simply re-heating the solder joint and adding new solder may be a tricky process for somebody with limited experience in soldering. You can't really see the inside of the via. If your soldering temperature is too low, you might not fully melt the solder in the via, if it is too high or if the heat application is too long, the pad & track might lift off the board.

If this were my BM8000 and if I planned to keep the unit for another 10+ years, I'd do following:
*) I'd remove all old solder from the vias with a good desoldering pump or high quality desoldering wick.
*) clean those vias with a Q-tip and IPA (to remove old flux)
*) I'd push a jumper wire through the via and solder it to the board on both sides with fresh solder.
*) clean off new flux from those joints with IPA

No thermal expansion or vibration will ever brake that connection from top to bottom layer again. This may be a tedious job, but it would be worth it for me, just for the peace of mind.  

I would definitely not replace the IC sockets, unless there are some defects visible on the sockets or solder joints.I also wouldn't touch the crystals or the 1nF capacitors on the +5V pins for now.

I'm not sure if I would replace the connectors automatically, it might be hard to get the right parts. It's true that connectors can be a troublesome source of intermittend problems (cold solder joints, oxidation on the pins, oxidation or bad crimping on the female contacts). But if I didn't see any visual hint of a problem there, I'd probably test the boards before starting to mess with the connectors.

rgds, manfy

This weekend I disassembled the microcomputer board to reflow the vias and replace the crystals. Here's the board with the shield desoldered. Btw I think someone else already had disassembled the shield, the shield itself was pretty badly warped and some of the solder pads for attaching the shield (this is the mk1 board) were missing.

Here's the board after replacing the crystals (with new 18pf capacitors) and 2 capacitors replaced. 

I desoldered the vias and applied some fresh solder. On this board there are wires connecting the 2 sides, so I'm not sure if these older boards have the same common faults. I did notice some suspect solder joints that I reflowed as well and cleaned up a lot of flux residue.    

Today it was time to assemble the Beomaster and do a power up test without the dim bulb tester. First I checked the voltages on the power supply board, the 5V, +15V and -15V lines measured all fine.

When connecting to AC, I noticed the standby red dot. The BM did not show the erratic switching behaviour like before so that was an encouraging sign. Finally I dared to tap the TP1 key and well... success! The Beomaster switched on (relays clicked) and did not automatically shut down anymore. All keyboard buttons and the volume dial were now working. The microcomputer self test (pressing 0 + Monitor) was also successful.

I measured around 13mv on the output board resistors so still need to finetune the no-load current and DC offset.  

The only issue I found was that when putting the BM in standby and pressing TP1 again, in some cases the relays did not engage even though the display was on. Any ideas on what could be a possible cause for this?

krais wrote:

The only issue I found was that when putting the BM in standby and pressing TP1 again, in some cases the relays did not engage even though the display was on. Any ideas on what could be a possible cause for this?

You need to wait until the supplies are bled down before you can power up again.
A 20-30 second wait is not unusual.
You will also notice the S-meter splashing around a bit as the voltages are going down.

Martin

From service note:

Starting problems

When re-starting shortly after switch-off (St.by) without load on the output, no mains voltage can be connected to the output.

The reason is that the protecting circuit is triggered by a DC shift on the output during switch off. The

protection circuit is not reset unil +/- 15V drops to 0V. Until then the RL1 and 2 cannot be activated (15 - 60

seconds after switch-off).

The problem is solved by raising the trigger level of the protection circuit adequately to avoid that the DC shift

on the output triggers the protection circuit.

This may be obtained by changing 5R254 from 120 kOhm ro 390 kOhm.

Dillen wrote:

You need to wait until the supplies are bled down before you can power up again.
A 20-30 second wait is not unusual.

Thanks Martin! I guess this means that the microcomputer is working fine. I also checked the remote and that works without issues. 

Now for the bad news...

I did a quick test hooking up an audio signal to tape 1 input RCAs to see if there would be any sound from the headphone output... nothing. When I checked the connectors on the preamp and tone control boards I noticed that a ground pin on P27 on PC4 had come loose. After reinserting that pin firmly in the connector housing and switching on the BM all of a sudden... magic smoke.

Not a big bang, just a puff of smoke around the starting up circuit board. I've narrowed it to the left bridge rectifier (one of the internal diodes has shorted).

Wondering if this is likely to be a normal part failure or could there be an underlying cause that I need to investigate? Any advice on a replacement part (or specs)? Thanks!

At first, congratulations to your success with the uProcessor board.

krais wrote:

Not a big bang, just a puff of smoke around the starting up circuit board. I've narrowed it to the left bridge rectifier (one of the internal diodes has shorted).

Wondering if this is likely to be a normal part failure or could there be an underlying cause that I need to investigate? Any advice on a replacement part (or specs)? Thanks!

Smoke coming from electronics is never really a good thing! There must be a reason for it and that GND connection you re-attached to PC4 is the first thing I'd look at.
Without disconnecting any wires from the board I'd check all the supply rails that were served by that rectifier. Check if there's any short circuit with a multimeter.

Which rectifier blew up? 30D1 or 30D2?
The service manual shows "KB 62-B125/110-25" for both rectifiers but I couldn't find any datasheet on the net. Check the component marking on those bridge rectifiers. It will help to find a replacement.

PS: After some more googling it shows me a Diotec bridge rectifier 400V/25A, square metal case with 6.35mm connectors. If that's the one installed, then overcurrent is unlikely - that current would have blown the fuse and/or some other components on its way to heaven. That supports your random component failure idea. Nevertheless, I'd also do the short circuit checks described above just for peace of mind.