Bowers & Wilkins Amp675

[Diver300]

I'm trying to repair a Bowers & Wilkins Amp675. It's an active base speaker.

The mains power comes into the small brown circuit board, which is mounted on the green circuit board. It is a low-power switch-mode supply, which powers the amplifier board.

The amplifier board is the one at right angles to the other ones.

The small brown circuit board also has a relay which switches power to the main switch-mode power supply. The main switch-mode supply is under the aluminium plate. Power from that goes to the green amplifier board. The speaker connections are on the right of the green board.

When the control switch is turned on, as signal from the amplifier board goes to the small power supply board, which turns on the relay for a few seconds.

There are two wires from the amplifier board back to the small power supply board. They are labelled "OK1" and "OK2". The relay will only stay on for more than the few seconds if neither of those wires are pulled low.

On the one that I am trying to repair, the OK2 wire is pulled low when the amplifier is turned on, so the system gets shut down after those few seconds.

I can't see any faults. The amplifier board is a multi-layer one, so hard work to trace the circuit. I can't find a circuit diagram anywhere.

Any suggestions on how to go about repairing this?

 

[Nigel Goodwin]

Over current or over voltage?, presumably it's powering up, finding a fault, and switching back off. It could be that the amplifier itself is duff, causing the unit to shut down.

I would suggest disconnecting the power feeds to the main amplifier, and seeing if it still shuts down.

While Bowers & Wilkins is a high class speaker manufacturer, they aren't an electronics manufacturer - and it's quite possible that the electronics (if not the entire unit) is bought in from China?. In switch-mode supplies high-ESR electrolytics is a common fault, due to the use of sub-standard capacitors.

 

[Diver300]

I'll try running the system without the amplifier connected.

I think that the amplifier uses PWM. The amplifier board has six 220 µF, 100 V capacitors in parallel. The capacitance is approximately correct but I don't have any way of measuring the ESR. It might be a good idea to either measure them or just replace them.

There is obviously some fault-detection circuit that signals via the "OK1" and "OK2" and I am wondering if that could be triggered by faulty capacitors.

I've only been given the amplifier part, not the case or the speaker, and I might try putting a resistor in place of the speaker in case the fault-detection circuit is detecting the lack of a speaker.

 

[Nigel Goodwin]

I'll try running the system without the amplifier connected.

I think that the amplifier uses PWM. The amplifier board has six 220 µF, 100 V capacitors in parallel. The capacitance is approximately correct but I don't have any way of measuring the ESR. It might be a good idea to either measure them or just replace them.

The amplifier is probably class-D, so used PWM with a low-pass LC filter to feed the speaker.

Those capacitors probably aren't the issue - however, everyone should have one of those Chinese kit component testers (I have four or five), they are great fun to build, and amongst many other functions will test ESR as well.

There is obviously some fault-detection circuit that signals via the "OK1" and "OK2" and I am wondering if that could be triggered by faulty capacitors.

I've only been given the amplifier part, not the case or the speaker, and I might try putting a resistor in place of the speaker in case the fault-detection circuit is detecting the lack of a speaker.

The lack of a speaker isn't a 'fault' and wouldn't be detected (or even checked for).

As I said, disconnect power to the power amplifier and see if that stops it tripping out. The most common failkure in amplifiers is the output stage going S/C, this then sticks high voltage high current DC across the speaker, which blows it - the protection circuits are there mostly to prevent that happening.

 

[Diver300]

I found that it wasn't easy to disconnect the amplifier bit, so I did some more circuit tracing. One of the fault-detection signals also goes to the main switch-mode power supply. It is shorted to ground when the power supply voltage isn't within limits, which it wasn't, and that was what stopped the small power supply from keeping the relay turned on.

Looking around on the main power supply, I found that a 4.7 Ohm in-rush limiter resistor had gone open circuit. It was a wire-wound through-hole resistor, and didn't show any visible signs of failure.

I replaced the resistor and the amplifier now stays turned on. I haven't actually tried it with a speaker or with any signals.