Pete wrote:

Just out of curiosity measured voltage across left and right channel speaker output - both at 45v dc !?

For what it's worth, I think you cleaned up your soldering or desoldering mishap(s) quite nicely!

What you did forget in your reactivation attempt was measuring, measuring, measuring! Once you noticed that you couldn't adjust the no-load current, the ONLY logical thing to do would have been to compare the nominal voltages that are given in the circuit diagram with the actual voltages of all those points! 
The results would have given a clear indication for what to do next - if not for you then certainly for some members on this forum.

Based on your description so far, I'm suspecting that you have a ground problem on that board. You measured 45V from speaker output (before the output cap C108, I hope!?) to GND and that suggests that some tracks that should be GND are floating (ie. not really connected to power GND). The fact that the board behaves so differently when the preamp is connected supports this, because now some parts of the board are suddenly connected to GND via the preamp.

All of this is fixable, but it needs a systematic approach that starts with measuring and then making sense of that measurement. Trial and error component swapping alone doesn't cut it. A power supply with adjustable output current would be a great help, but well, life is not always fair. There are other ways to limit the current.

Anyways, it may be a good idea to take a rest from this project for a while. You'll probably not be able to find the problem quickly, but now you're too eager to get it working, it seems - and that leads to "shortcuts" and mistakes in the repair process.
But don't throw the board out just yet! Putting in another different output amp is not gonna be plug-and-play either.

rgds,

manfy

Many thanks for your help..

manfy - 'You measured 45V from speaker output (before the output cap C108, I hope!?) to GND'

I measured it from the end of L100/200 on the 4pin central white connector to the other terminals on the white connector. At this time the board was out of the case so I'm guessing not connected to earth (via heatsink?) and then as you say got earthed as soon as i plugged the preamp output to the board - which was when the frying started!

as for voltage measurement at other place i wasn't sure what voltages I'm looking for as it was noted that some given in the circuit diagram are incorrect. Could the circuit diagram be wrong so as to stop people copying it? naa I'm just going bonkers!

Giving this whole episode a rest is a good idea, a very very good, big, great idea (as trump would say)  tho I wish I could find a photo of a working undamaged board taken from below so i could double check my track repair... but I haven't seen any pigs flying past my window these last few weeks 😕 

Pete wrote:

Could the circuit diagram be wrong so as to stop people copying it? naa I'm just going bonkers! 

😁 Haha, naah I got a better explanation: The BM1700 was made some time in the seventies and back then, Cannabis was probably not yet a controlled substance in Denmark...and the designer of the docu and circuit diagram consumed particularly much of it....😎

I even have proof....aaand an explanation why your offline tests didn't work out:
According to the circuit diagram in the service manual, there's only one type of GND on that board. But when you look at the layout drawing, you can see that you actually have 4 (in letters: FOUR) separate ground planes: power ground left, power ground right, signal ground left, signal ground right. 😠

Granted, all those ground planes are galvanically connected at a star point in the power supply, but they are physically separated on the amplifier board! No wonder your offline test didn't work!

In any case, my recommendation of taking a break from this project still stands. Get some mental distance from it for a few days. You'll still have to build a current limiter before you can start actual tests.
Can you still get incandescent light bulbs in London? If so, try to get a hold of a 25, 40, and 60W bulb! ...and an old bedside table lamp!
Wire it up so that you can use it in series with the BM1700. This way you can limit the maximum current that can flow through the transformer of the Beomaster. 25W = max 0.1A at 230V, 40W = max. 0.18A, 60W = 0.26A

If you only can get one, aim for 40W. That should be able to handle the normal no-load current of the amp and limit any excess current while you're searching for the actual fault.

Thanks manfy,

After more than a few days off and a few night dreams about the B&O amp pcb/circuit I've dug out a 40W bulb to wire up in (series?) on the live 240v, yet know not what voltages to test for on the power amp stage on the pcb.

The earth configuration is indeed a crazy, specially since it takes the placement of the tiny pre-amp output plug to this board to earth despite there being two earths from the power supply going via this board to the headphone and speaker outputs! Those earths literally come to this board only to be routed to the speakers via the white four pin connector without touching anything else on the board!! (purple and brown wires)

So in my further quest to find an accurate pcb track diagram after my solder digging/mining disaster and perhaps for more voltage info, a printed service manual has been ordered in the hope the tracks are more visible than on the photocopied one downloaded some time ago + a variety of small capacitors (non electrolytic) and a better transistor tester.

Rebuilding the amp on euro vero is a final option. Yet due to likely B&O cannabis influence in the past, confidence in the manuals' circuit diagram is V-low... for example ceramic C102/202 is 1nf on circuit diagram yet a ceramic marked 3n3 is placed on the board.  Obviously everything worked originally but if I want to replace components with better alternatives what values do I choose?

This begs my final question... has anyone on this planet ever repaired this board with 21st century equivalent components (especially those pesky Te1088/89's) after a speaker shorting of one channel ?  If anyone has, evidence of said repair is thin on the ground, including here! 

eg This thread ends abruptly with no solution: link and even illustrates electrolytic capacitors placed the wrong way round on the board - a fully understandable mistake considering the circuit digram and screen print on the pcb!

Those pre-drug-control era engineers were certainly having a good time

If anyone one this forum has one of these Beomaster 1700 (type 1701's) please can they open it up and photograph the underside of this board + transistors in situ on top so i can triple check again and get some sleep 🙁

muchios apreciatum 

 

Just to set the record straight: Keeping signal GND and power GND separated really IS a good design practice. It avoids noise in the signal path and it's still a common thing to do in modern electronics. But when this happens you normally use different GND symbols in the circuit diagram.
So, basically it was my mistake to tell you that you could test the board by disconnecting input and output and supplying power with a DC power supply and then start measuring.

We will still do an offline test of the amplifier board, but you will have to prep the board first. So please do not try to power up the amp before we've done that. 

Regarding the light bulb, yes, you have to connect it in series into the mains lead. The light bulb acts as a cheap power resistor. If everything is fine with the board, it will only draw a small current, the voltage drop will be small across the bulb and it  will be dim. If there's a short circuit or similar problem on the board, it will try do draw a high current. Now also the current in the mains line will increase, i.e. the light will be brighter and there will be a bigger voltage drop across it, which means the trafo voltage will go down which also lowers the current drawn.
It's quite simple actually but if you want, you can search the net for "dim bulb tester" for additional details.

Generally you can assume that the circuit diagram in the service manual is correct. But don't take that as a guarantee. Mistakes can happen. Especially for vintage products made before computers were mainstream, such typos or mistakes may have been intentionally overlooked because scrapping and reprinting of 10,000 service manuals was jsut not a viable option. They were probably just reported internally with separate change notices.
Just keep in mind the legal disclaimer: "Subject to change without notice." and if something looks really starnge, just double-check the circuit diagram with the actual board, layout and components and if you find a discrepancy, you have to figure out what is right and what is wrong.

There's one more thing I want you to check and confirm: The power transistors on the heat sink need to be isolated from the heat sink! Normally there's a mica sheet between the metal part of the transistor and the aluminum heat sink. Check the resistance between the collector of every power darlngton and the heat sink. If there's a short, fix it.

Last thing: terminology. Earth (or PE [potential earth]) and ground are 2 distinct concepts in electronics. To avoid confusion, forget about "earth". Let's call it GND from now on.

More later!

Thanks for this manfy, I've checked heatsink isolation after putting fresh compound between all the mica/transistors - isolation is good - from everything (-: I also understand your points about GND and separation.

All suspected newly blown components will be removed later today and hope to replace in the next few days when the latest deliveries arrive; polyester film capacitors, transistor checker and a few 39 pf ceramics - In the meantime, a 'dim bulb tester' will be created😕

Fingers crossed that the manual showing a clearer view of the PCB track layout will arrive soon also.

Cheers

Pete  

linkHi Pete,

Here's how I would approach this fix now:

1) visual check of the board based on the service manual
2) quick check with a multimeter in continuity test mode around all the suspicious spots on the board (that means all the spots that were affected by the soldering massacre)Once that is fine, I'd prep the system for the power-on test of the amplifier board:- the dim bulb tester with a 40W bulb is required for this- to make sure that the other electronics of the BM1700 doesn't affect my test, I'd disable it by unplugging P5 connector from the power supply board 22Since I want to test left and right channel amps separately, I'd prep the board this way:- I'd start testing the right channel because in your case there's a higher chance that it works
- desolder the wire for 50V supply and 'power GND left' from the left amp (alternatively you could also unplug connector P6 from the power supply and run test wires with croco clips to the right output amp)- solder a jumper wire from 'power GND right' to 'signal GND right'. You don't have to worry about 'speaker GND right' because the speaker MUST BE DISCONNECTED for this first test anyway!!- solder another short wire to 'power GND right' which we will use as a clearly defined GND point for the multimeter. All voltage measurements will be done against this GND point and this point only!- IMPORTANT: turn trimmer R112 to maximum, i.e. turned toward R111. Confirm this by measuring the resistance between base of IC100 and R111. It must be zero or less than one ohm!
Please note: Since '50V supply left' and 'signal GND left' is now disconnected, we also disabled the startup circuit R1/C1 for the right amp! That's fine as long as you do not connect any speaker to the amplifier output. I'll show you later why this is so.
Now we're ready for the power-up test. Our setup in short:- only the right channel is connected to 50V supply and GND; signal GND right and power GND right are shorted on board 21   - amp input and output, i.e. the speaker is diconnected! 
- trimmer R112 is set to maximum- primary side of transformer is connected to mains with a 40W light bulb in series
Once you press the power button on the beomaster, you should see a short bright flash from the dim bulb tester because of the inrush current and then the bulb will go dim. If there's still bright light after 1 second or if you see smoke coming from your circuit, switch off immediately -- but this is very unlikely provided the dim bulb tester is wired up correctly.
You should measure and log all voltages marked with a red box (if they are off I'll need those actual values to assess what's wrong):     

Login to view image

linkIf everything's good on the board, you'll see something like this in the first 6sec after power on:

Login to view image

The idle current through R121 is below 1mA, supply voltage close to 50V, the voltage at TP101 (speaker output) is about 24V and you will be able to adjust the idle current with R112.

If there's an open at IC100, you'd see good 350mA at R121 (which is not a problem because of the dim bulb tester) Login to view image

But without the dim bulb tester and the same fault on the board, you'd see 2.5A, i.e. 2.5W on a 1W resistor, ergo R121/122 would have every right to smoke! 😎

Login to view image

 

 

Brilliant ! Thanks manfy.

I'll set everything up on Monday as I've promised the wife I'll spend time with her on her weekend off,  then I'll be able to take each step you've provided at a dedicated slow pace.

Pete

PS Monday is going to be a long day !

 

Photos below are not for the faint hearted - LOL

 

Login to view image

attempted work-around ruined pcb tracking !!!!

 

Login to view imageAfter everything...starting from scratch to make one channel of the amp onto double sided vero is the final option feared!

Once all is in place there'll be testing and measurement as advised by manfy and if it works out another channel will be built exactly the same

There was never any intention to restore and sell this B&O, I just wanted it as a living (working) reminder of my uncle. Everything works apart from the output stage - bit like life really (-:

Yes, I'm quite mad!