I am hoping to get some clues as to why my beomaster 5000 shuts itself down after a minute or so. It used to happen occasionally, now every time.

I am not experienced in repairing electronics but expect I can learn my way into it. Appreciate any troubleshooting advice I can get. Thanks.

Welcome to Beoworld !

Typical symptom of dead idle current trimmers.
They lose contact and the idle current gets too high.
Normally not an expensive repair. I've done many.
New trimmers, soldering, adjustment, basic electronic skills and a
minimum of instrumentation is needed.
Do one channel at a time, beginning with the one with the highest idle.
There can also be other reasons but the idle current is by far the most often seen.
I replace the trimmers on all machines of this type even if actually still OK as they will
fail sooner or later.

Martin

Thanks Martin for your advice, I appreciate it.

Can you tell me where I can buy the trimmers? Is there anything else you would replace on this machine?

Cheers,

Chris

Chris,

Any good electronic parts dealer will have them.
I can also sell you two if you cannot find some locally or on the net.

The capacitors will have aged and there is a kit for fixing this but
if your Beomaster is functioning and sounding good (apart from the
shutting down), there's really no rush with replacing the caps.

The same goes for the muting relay. It's a standard fault and I can sell
you a new one but if your machine has good sound through both channels,
there's no need to replace it at this time.

Check for cracked solder joints while in there. Especially at the
output and driver transistors. Also where the vertical DIN connector
board meets the preamp and around the RCA sockets.

Martin

Thanks Martin,

Could you give me your price for all the parts: trimmers, capacitor kit and muting relay. thanks alot.

You have a PM regarding parts.
Check your inbox at the extreme top right corner of this page.

Martin

Hi Chris,

The output stages were well-designed, but the drastic shutdown protection-circuit can be improved upon as follows:

(I'm assuming that you have the Service Manual on hand)

On Page 1-8, PCB2, output stage:

1. Locate R224 2.2KOhm - replace with 3KOhm.

2. Locate R225 390 Ohm - remove and replace with 3x 1N4148 diodes in series. Cathode-side facing the 100 Ohm trimmer you just replaced.

A little technical background first - the above 2 resistors are in parallel - why not just put one resistor of the combined value of 331 Ohm, and save some money? The reason is quite simple - this circuit must be very stable to ensure a consistent shutdown at the same parameters every time. 2.5V is passing through these resistors, and if there was only 1 resistor, it would heat up with unpredictable effects on the resistance, and you would have shutdowns at any random volume/distortion level.

So 2 resistors were used, to ensure minimum heat effect on the resistance values.

What this mod does, is as follows:

Now, with 3x 1N4148's (each with a voltage drop of 0.6V = 1.8V), there's only 0.7V remaining on the unstable circuit, which generates a very small amount of heat, and a very fast response causing an immediate and instantaneous tiny drop in volume, without ever getting to the extreme condition of shutdown.

I also replaced TR208 and TR209 (BF857) with the much more responsive 2SC5171 - they should be hFe matched to within 10% of each other.

Regards

Menahem

Menahem, I have repaired so many of these machines, I've lost count, and never
ran into problems regarding any of these issues.
You will have to explain a bit for me to understand your motivations for re-designing
and I would like to comment a little on your post as well.

First, you wrote:
"... the drastic shutdown protection-circuit can be improved upon as follows".
None of the components you refer to (the BF transistors, the trimmer, R224/225) are in the protection circuit.

You then wrote:
"2.5V is passing through these resistors, and if there was only 1 resistor, it
would heat up with unpredictable effects on the resistance, and you would have
shutdowns at any random volume/distortion level."
Well, I'm not sure that's what you actually mean:
2,5V across (not through) a resistor of 330 Ohm would see a current of apprx 7mA.
That would again see apprx 0,02 W in the resistor. Hardly enough to heat anything
and surely any ordinary 1/4W resistor can handle this and it doesn't matter much
if it had been a 390 Ohm or 330 Ohm so that will not be the reason.
The trimmer is part of this circuit too, so the 2,5V will actually be divided
even further down and the power dissipation likewise.
Besides, the protection circuit merely measures output stage DC current so
the signal content, distortion and/or high level output alone will not trig it.
Only a high current will and that will be if the idle current is wrong, a
driver gone bad or one or both output stage transistors fail.
It also measures through an RC-delay to prevent it from reacting to f.e. startup
rush-in currents and other sudden spikes etc.

You wrote:
"Now, with 3x 1N4148's (each with a voltage drop of 0.6V = 1.8V), there's
only 0.7V remaining on the unstable circuit, which generates a very
small amount of heat, and a very fast response causing an immediate and
instantaneous tiny drop in volume, without ever getting to the
extreme condition of shutdown."
No, no, no.
The idle current has no influence on the volume at all.
The idle current has the job of driving the output stage transistors to a just-open
point to prevent crossover distortion. That is distortion from the switching
across the zero-line where one transistor takes over the signal from the other.
So both are held slightly open so they will never "lose the grip of the signal"
and the "idle transistor will be ready when the signal comes
back from the active one" (to put it reader-friendly). It has something to
do with the transistor characteristics and the way a Class B complementary
output stage like this works.
Anyway, given a specific current through the component (like here) and a specific
voltage drop across it (like here) the power dissipated from the component doesn't
depend on it being a resistor or diode(s).
The sum of the power dissipation will be the same. That's Ohms law.

What do you mean by "react" ? - Resistors "react" just as fast as diodes.
Where should the "tiny drop in volume" come from ? Not from the idle current or protection circuits surely ?
The protection circuit issues a signal to the processor to initiate a shutdown of the Beomaster, it does
not alter the volume setting.

I would always see a resistor as a more stabile component than a diode, let alone three diodes
and it will not fail from an eventual sudden spike of voltage or current.

The idle current circuit with the two resistors in parallel, I think I would
leave well alone.
Introducing diodes is the classic way of solving it but if you are not completely
sure what you do and choose your components and physical placing of
them carefully, you could end up with unstable idle current.
1N4148 diodes are quite temperature dependant and the temperature
inside the BM5000 can vary a lot due to its design, more than most amplifiers.
The diode solution is good for comparatively low power amplifiers but I wouldn't
use it for high power amps like this.
Amplifiers like the BM4400/6000/8000/Penta etc. all uses the resistor solution.
I believe the dual resistor solution and the somewhat odd ohmic values was
chosen deliberately because of this and two resistors of different values
were used to have one resistor to "grab and hold" the idle current from
running off if something goes badly wrong in the driver stage and the
other resistor (which practically represents the resulting ohmic value alone) burns.
(Not all of the above mentioned amplifiers uses two resistors in parallel, some
uses only one. My guess is that the trimmer would be the most vulnerable component in
this circuit and experience has it that the trimmer is exactly the component that burns in case
of faults to the driver stage whereas its series resistor usually does fine, even if alone).
I would need a very good reason to change any of it.
Besides, the paralleled resistors are in series with the idle current trimmer
so I don't see the reason for exact ohm matching at all. Under normal
circumstances the idle current setting is correct at about 50% of the
trimmer range (center position ~ 50 Ohms) so that is more or less what
the tolerance should be within and the trimmer will then automatically
compensate when setting the idle current.
The original trimmer is not a precision component in any way either.

Why would you go for a "more responsive" transistor for TR208/209
and what do you mean by more responsive ?
The BF857 is in fact a high frequency driver with excellent specs,
inexpensive, reliable, still available and goes way up to 100MHz.
In this amplifier we only deal with frequencies up to apprx 20KHz or maybe
a little more and the power handling is not the big issue here so
slewrate etc. surely cannot be a big issue ?

I would also like to hear why you would like to have them HFE matched.
The two BF857 (in each channel) are practically working in series.
One of them acting together with a constant current generator (providing an
always stabile idle current BTW) and there is a 180 ohm resistor at
its emitter lead to help set it's working point.
Apart from the signal (AC current) that feeds the output stage transistors, they
practically see the same current flow no matter the HFE match.
At least that's how I see it. If anything can set the current flow off, it
would have to be a differing HFE in the output stage, where it is a bit
more important to get a good HFE match and still the output stage transistors
in most of these amps seems to be non-matched anyway, with their respective
emitter resistors compensating nicely.

Menahem, don't get me wrong. I know that we haven't always agreed on
everything and I am not sure about this one either but I really would
like to understand your motivations.

Martin

Hi Martin,

Firstly, thanks for your comments. I definitely take your criticism in a positive light, and certainly took no offense here!

My motivations were to get this unit fixed, as described in this post, which came to a dead-end, with no help.

http://forum.beoworld.org/forums/p/28475/236340.aspx

Here in Israel, there is a dearth of real technicians, with the normal response being - Not worth repairing - throw it away. I get a fair number of older stereos in all the time for restoration, from a handful of Stereo Shops whose owners are always older, and like me, don't buy into the current way of the world - "older is junk, and newer is better".

So I really pull out all the stops to keep my customers happy.

And, when I get stuck, there's not many from whom I can ask for help.

Especially B&O - most techs don't want to hear about it.

So I had this BM5000, to which I did my normal full restoration, and still had the problem of shutdown in the Left channel, at about 4.8 volume. The customer had reported shutdown at 3.0, so there was an improvement after the restoration.

I did change a couple of transistors in my restoration, to improve circuit stability - the parallel drivers with 2SC1845's - I got much closer to 0mV on the "circuit-floor" with them - and as mentioned above the BF857's. The BF857/2SC5171 is a matter of sound only. There is an audible improvement in the sweetness of sound using the Japanese components, and I just like them. Same thing like the craziness among audiophiles to find the "best-sounding" op-amp. Transistors do sound different. As a matter of perfection, I use matched transistors wherever possible, even if technically unnecessary - it's just the way I do things!

Another favorite is the MJ15003 instead of the 2N3055.

Anyway, I was wasting a lot of time on it - I had removed every semiconductor in the left output stage, and according to my semiconductor analyzer, every one checked AOK. I wasn't going to rebuild the entire output stage with new components - that would have been ridiculous. I checked every resistor, and of course, all the caps were new, including the film caps. The right channel had exactly the same components.

Just for trial, I replaced the BDV64 and 65 with brand new ones - still shutdown at 4.8.

I was watching the problem happen - monitoring the Left idle voltage, it would remain stable at low/medium volumes, and at 4.8 volume, it would accelerate up to 280mV, at which point the shutdown was triggered.

Anyway, I had reached a dead-end, and I wasn't quitting - I had to think outside of the common solutions to this problem. So I tried the diodes, and that solved the problem. The machine has been at the customer for over 3 weeks now, and running just fine. To this day, I don't know where the problem was.......

Your description of the technicals is so much better than mine - thanks! Sometimes I tend to oversimplify the technicals, and miss the real issues - for that I'm not known for my talkativeness!

Regards

Menahem

The load on the output stage is not related to the specific volume setting
but rather the signal content and it's input amplitude so what shuts down at
3.0 one place can easily shutdown at a higher or lower setting somewhere else
or if playing something different.
Anyway, I had a similar case to what you describe which proved to be
nothing more than one of the output stage transistors causing a thermal runaway.
A transistors amplification rises with its temperature so if nothing stops
it, it will run away and eventually destroy the transistor.
Furthermore, the transistors will have to sink the heat they generate as they
regulate current so the cooling is vital, also to provide for a little delay
in heating/cooling for the thermal runaway protection.
In my case, two IR thermometers monitoring the output transistors showed that one
transistor obviously had no particular cooling, it's temperature drifting
up and down very fast with the load put on it.
It had only very little cooling paste and what it had was dry.
Cleaned and fresh paste added and the thing played on happily - and still does.
Transistors can also lose thermal contact inside. It happens but it is rare to see
and these cases often result in the transistor blowing anyway so won't be easily noted.
In your case, I think I would have focused a bit more on the constant
current generator (where a zener diode is often found) but that may not have
helped anything if the problem was indeed with one of the resistors in
the idle current trimmer series.

Owners of a problematic BM5000 reading this should also check for cracked solder joints in
the amplifiers, especially at the BF857 transistors.
A thing that has killed many amps by now.

Martin