Help with diagnosis of a broken sound system?

Hello all from Copenhagen,

This is my first post. I found the audiokarma.org website searching for some help to do a DIY repair my old Bang & Olufsen system. I am an electrical engineer, but have never worked on audio systems before. So when my system went crazy yesterday, I thought this might be a good opportunity to open it up and learn a good bit about musical systems.

The problem is, I don't even know what my problem is! And so I was wondering if there's anyone out there that might be interested in giving me a hand with a virtual diagnosis.

The system in question is a B&O Beomaster 3500 (amp+tuner), and two B&O Beovox CX50 (speakers). I also have a B&O CD player plugged into the Beomaster, but I suspect that is irrelevant to the problem at hand. I imagine that on this forum my system is nothing special but I am extremely fond of it, and look forward to getting it back to work again!

What happened yesterday was as follows. I have a 3.5mm jack connector coming out the back of the Beomaster so that I can occasionally plug in an auxiliary music source such as a portable mp3 player. My girlfriend committed the cardinal sin of unplugging this jack from a laptop to plug it into her iPhone, without first switching off the equipment. This naturally resulted in a couple of loud sound explosions emanating from the speakers. You know, the kind that hurt you deep down. She's done it before (I've done it before if I'm going to be honest), and everything's been ok. But this time, something really bad happened, and now the system will only make nasty noises. I've uploaded a couple of wav recordings of the sound that comes out of the speakers, as I thought this might be a good first step to determining what the problem is:

1) This is the sound coming out of the speakers when I switch the system to the auxiliary input (the 3.5mm jack), with the jack not actually connected to anything. It's just lying on the floor:
https://drive.google.com/file/d/0BzDUei7aobABNE9sQmpXZDZJUDQ/edit?usp=sharing

2) This is the sound coming out of the speakers when I switch to the CD input and play a song:
https://drive.google.com/file/d/0BzDUei7aobABYzNXeExvUlFuekU/edit?usp=sharing

I suppose I'm king of hoping that this is a really characteristic sound that only happens with a specific type of problem. But if its not, then any help whatsoever to get me started on trying to find the root of the problem will be very helpful. Even just a couple of Google search terms would be great! I'm pretty handy with a soldering iron and have easy access to replacement electronic components.

I almost forgot to mention. My girlfriend said that there was a funny smell when it happened. Unfortunately I didn't arrive on the scene for a good couple of hours afterwards so I wasn't able to smell it myself. But when I asked her if it was a burning smell, she said no, it was a "funny" smell. Which makes me think... blown capacitors?

Thank you very much in advance for any help anyone can offer!

Eddie

Eddie,
Unfortunately, I am not familiar with the equipment you are having a problem with, so I can only generalize. Often pulling the plug on an input while hot will result in an oscillation that will destroy the output transistors. The transistors almost always fail short, which causes the emitter resistors to burn up. Sometimes this destroys the speakers by putting high current through the woofer coils, but usually the damage is contained to the amp. Or it could be something at the input side of things.
What test equipment do you have access to? A DMM? Oscilloscope?
A good place to start is to search for a schematic on the web.
Hopefully someone who knows this equipment will jump in with some suggestions, but lacking that, here is what I would do:
1.) Disconnect the speakers. I'm assuming that all of the input sources (tuner?) are also not playing, so I am focusing on the amplifier output. But obviously that is just a guess. The sounds you recorded would make me think that something in the input might have been damaged, but best to disconnect the speakers till you get to the bottom of this.
2.) Measure the DC voltage across the output terminals with the volume turned all the way down. It should be near zero if all is well. If it is tens of volts, then there is a good chance the speakers are damaged. And that would be a clue as to where to check next.
3.) Open the unit and look around. If it smelled bad, there is a good chance you can spot the burned part.
4.) The output transistors can be checked for shorts with an ohm meter.
5.) Sometimes there are fuses inside the unit. Often soldered in (pigtail fuses).
The rail voltages could be +/-45 volts or so (depending on the amplifier power rating) with lots of current available (several amps), so be careful probing around.

Terry

I too know nothing about your system (and not much about home audio). But I do believe that disconnecting plug jacks (while powered-up) can sometimes result in a momentary short between the pos. (+) and the neg. (-) contacts. With luck, this will only result in a blown fuse. But it can sometimes cause a cascade of failures (and hence the smell your girlfriend mentioned).

Here's a manual:
http://beophile.com/wp-content/plugins/download-monitor/download.php?id=140

There are a few fuses you should look at.


Terry

It sounds like a bad transistor or amplifier IC to me... It could be some IC without power too. (check fuses first...)

And try to discard an speaker problem ! Perhaps the amp is still ok...

Is the sound coming from both speakers, exactly the same?

There are several ICs in the unit, it would be good to isolate the problem, perhaps measuring voltages at every IC pin, following the signal path (bold line in the schematic, from page 18 of the PDF).

I'd extract or inject signal before and after some coupling capacitors (capacitor C51, C104, C43, at page 19 of the PDF file) to separate the preamp stages, to see if the problem is at the pre-amp, or at the power amp.

If you have an oscilloscope, you could measure at every preamp - power amp stage.

Or, listen to the signal at different stages, through another amplifier. Ask here if you want to go that route.

But if the problem is inside the amp, you need to measure inside the unit, and for sure unsolder some components.

A fast way to check the speakers themselves to verify if they are working, use a flashlight cell zinc carbon or alkaline (Edison cell) are best as they are least likey to cause any harm, and connect the speaker leads to the cell. They will make an audible bump sound, and a momentary contact is all that is needed to check each woofer, each tweeter and each midrange. If you stand with your ear beside the midrange and tweeter, you should hear them as you make and break contact with the cell.

If you hear scraping, the voice coils are likely deformed, and this can contribute to poor sound.
No sound from them when testing witha good battery indicates the voice coil of that speaker may be burned open, a crossover component may have opened, or if there is a protective device, that device may have opened.

Try a different pair of speakers. At low sound levels, almost any speakers will work. If the stereo sounds okay through the other speakers, the issue is the original speakers. If the amp sounds bad on the other speakers, the issue is in the amp, and likely in the output stage most likely. The input stages, while they make noise when something is shorting momentarily across the inputs, that in itself is not an issue as often volume and level controls will take an input stage down to the ground potential.

If you can use a set of headphones let us know what sound you get out of them.

It looks like there is circuitry on the power amp outputs to protect the speakers. If the amp output went bad, then the speaker protect relay would open. That seems to be the case on speaker output 1, but not speaker output 2. I would think it would also display an error message since POWER_FAIL signal goes to the CPU.
So it looks like maybe the problem is at the input side of things. Definitely check the fuses and supply voltages first, but then you might have to chase some signals. If you don't have access to an oscilloscope, some people use any available audio amp with headphones to probe along the signal path. An old cassette deck or similar is good for this. Make a probe from a cable and a capacitor. The capacitor is so that no DC voltage is applied to the caseete deck input. A small value is sufficient (0.1uF or so). Feed the signal (through the capacitor) into the tape deck input and monitor the signal on the meters or with a headphone. Always start with the input level control turned all the way down. Using this, you can trace the audio signal along the signal path pretty easily.

Terry

Hello all,

Thank you very much for your extremely valuable tips! It sounds like its going to be a good bit of fun. Unfortunately, stuff's come up that mean I won't be able to try all this out until the new year - but I will definitely post the results of my tinkering then.

Happy holidays!

Eddie

Hi all,

Sorry there's been a bit of a delay before I could start to act on your advice. I've been away. But I'm back now and have had a decent tinker.

The manual, with its circuit diagrams, is a great find so thanks for that link. And thanks for the tip to check the voltage across the speaker ports - I was able to verify that the voltage was safe to try on some spare speakers and I was therefore able to conclude that my speakers remain undamaged.

So I opened up the amp; unfortunately nothing was obviously damaged, and all the fuses were intact. So I generated a 440Hz tone with my laptop and put it in to trace it along with a digital oscilloscope.

For the below findings, please refer to the circuit diagrams and screen shots that I've uploaded to the following zip file: https://drive.google.com/file/d/0BzDUei7aobABZzhKUVpoVG1KRzA/edit?usp=sharing

The 440Hz tone is shown in PP1.png, a triangular wave with 2V amplitude. What comes out the other end is the mess shown in PPOUT.png. Interestingly, the output signal is completely unaffected by the volume setting, which gives us a nice hint of where to look.

So, I started tracing the signal from the input along the trail shown highlighted in blue in Circuit_diagram.pdf. Each red dot corresponds to a point where I applied the oscilloscope probe and screen shot the output. So as you can see through PP1.png up to PP4.png there's not a whole lot going on, except for a voltage offset being applied, and a bit of rounding off the lower corners. Still looking good. But as soon as we get to the output of the op-amp LF353 (pictured in PP5.png), its starting to look really ugly.

What do you all think? Is this normal? I suppose it could also be that my input signal is too loud, and so the op-amp is getting saturated at the upper ends; I don't really know how much signal distortion is to be expected as normal.

The reason I stopped there is because I can't seem to find the next IC along the trail on the board (LC7537). I have a suspicion its lying under a big metal shield which can only be removed by pulling out the whole PCB and committing myself to a long and difficult reassembly, so I was kind of hoping I'd found the problem, and won't have to go there! But I don't know, any thoughts so far? If what I've seen so far is normal, then let me know and I'll start tearing out screws and connectors everywhere to find that blooming LC7537 and keep up the hunt

Please note that the trail shown on Circuit_diagram.pdf is for the right channel only, but that I also followed the same trail along the left channel, with identical results.

Many thanks in advance for any pointers/suggestions. I look forward to finding the culprit!

Best,

Ed

It certainly looks like the LF353 is causing the problem, but before you conclude that, be sure to look at the power supply rails that supply it (the +12 and -12 volt inputs to it). If the -12 volt supply was missing, it might explain what you are seeing.
I don't think the DC offset that you see on almost all of the points should be there (point 2 and on). I would have expected to see the signal staying centered around zero volts. But it looks like a failure of the LF353 could cause that by raising the DC level all along the path. The LF353 should be a pretty easy part to get ahold of.

Terry

I don't like what I see at point 2... Can you disconnect the suspected ic to see what happens there?

with the music playing it sounds like a bad transistor to me .

Great to hear that you guys think we might have found the culprit already. I'm going to order a bunch of parts on Farnell anyway for another project so I'll just stick a couple of LF353's on there.

Petehall347, would you suggest I throw in a couple of transistors into my order as well just in case? Which one do you think would be suspect?

hard to say what transistor ... last time i heard one like that it was a driver transistor .. could be any chip or transistor though . diagnosis is the best way before ordering stuff and shot-gunning it .

Be sure to check the +/- 12 volt supplies before you go swapping parts, but it looks like it should be the LF357 (IC7). There is not much else that would cause what you see on BOTH channels. All of the other possiblities require more than one part to fail. But IC7 is used in both channels.

Terry

TerryS, you're absolutely spot on. I measured the V+ and V- pins on the LF353. The V+ pin is at +12V, but the V- pin is not far off ground!

Where do you think I should look to solve that?

TR8 is the -12V regulator pass transistor. It is on page 2-4 of the schematic. I suppose it is possible that the LF357 (IC7) is still responsible and is shorting out the supply voltage, but if that is the case I would expect IC7 to be pretty warm to the touch.
At any rate, just carefully measure the voltages around the -12V supply and it should be pretty obvious.

Terry

Ok, thanks to your guidance I've definitely found the culprit now. The 100R resistor on the collector of the -12V regulator pass transistor is dropping 30V! It also looks blackened. So hopefully I can just replace that with one that matches the specs given in the manual (10%, 0.3W) and all will be fixed.

Unless... I've not known high-power resistors to just fail like that, especially in a supposedly high quality piece of kit, so I wonder if there's some underlying cause? The voltages all look to spec upstream of that resistor, so hopefully not. It *is* 24 years old though, so perhaps its normal to expect a few parts to start blowing as the years go by?

Anyway, I'll post an update when I've replaced the part and we'll see what happens.

Sorry, I forgot to upload a copy of the relevant circuit diagram in the manual with the highlighted component, in case you're interested: https://drive.google.com/file/d/0BzDUei7aobABZUl0eEhOM2VzZGc/edit?usp=sharing

For some stupid reason Google Drive displays the pdf preview in black and white though, so if you want to see where I've highlighted the culprit component with a red circle you'll have to download it first and then open it in Adobe Reader.