Trace Elliot RAH 600 SMX power amp issue - hellppp!

[Fran]

Hi.

I've had this amp for years. It's developed a fault in one of its power amps (there are 2, sharing a common power supply)

I'm getting dc voltage at the output of the faulty amp pcb. I've replace all mosfets, and the driver transistors. I've checked all diodes including the zeners in circuit, with a digital multimeter set to diode range, and can't find any faults. By swapping the new mosfets in differing combinations with the old ones I get varying dc voltages at the output. I've had had little as about 275 mV, to as much as over 60V. Spears to be negative. Ie, negative multimeter to the tip of the output jack. The other power amp is fine. The power supply gives as near as damn it +-75V.....its beaten me. The only issue I've really detected is this,t one of the psu smoothing caps gives up its voltage more rapidly than the other one, but as I said, both power amp boards share the same psu, and the other power amp is fine....
Please help, I'm running out of ideas. Thanks. fpettifer@hotmail.com

[diydidi]

This is all I have. Hope it helps.

[guitician]

I wouldn't think the 275mV would be bad to run on. Did it play went it was that low?

[Fran]

I wouldn't think the 275mV would be bad to run on. Did it play went it was that low?

Hi, thanks for responding. I was getting a very,very quiet,but highly distorted signal. At this voltage I did feel it was safe to connect my loudspeaker.. The thing is, the other power amp works perfectly. And with zero input signal, it's voltage at the output is zeroV. There's no apparent damage to the pcb, everything looks as it should. I didn't take the diodes out of circuit before testing them, as I was able to compare their readings with the ones on the other (working) power amp....Its gonna be something simple that I've overlooked, but I'm now considering having to save up and pay someone to look at it for me,.. I've just fitted yet another set of mosfets from a different supplier, and the output voltage is no about -6v. DC. well I think DC, because when you change the polarity of the multimeter leads the reading goes to +ve voltage. All tips greatly appreciated. Ta. Fran

[Fran]

This is all I have. Hope it helps.

A great help.. Thanks😉

[guitician]

Having a working side helps for comparing voltages and tracing what a working signal looks like a various stages within the amp. Good luck.

[Fran]

Gotcha. Open circuit R15. 15k resistor. I've also replaced the psu smoothing caps, as the ones fitted are 23+ years old. The amp now seems to work ok.. How do I set up the buss. Please help me with this one. I have a multimeter and a ESR meter.

[mozwell]

I have never liked amps with no source resistors on the mosfets, the source resistors help a little with current sharing, and IMHO makes the amp more reliable.
If you can, with power off, insert a 0.22 ohm 5W resistor in the source of each mosfet, then with power back on you can then measure the current through each mosfet by measuring the millivolts across each of the 0.22 ohm resistors.
a rough guide would be 25 to 40mA per mosfet but others here may have a more definitive bias value. 25mA and 0.22 ohm = 5.5mV DC

If you cant do this, with power off, you can insert a multimeter set to measure AMPS DC in SERIES with either the positive or negative supply rail (where it enters the power amp pcb) turn the power on & measure the current with no signal & no speaker. The current through each mosfet is half what you measure, as there are two mosfets in parallel. adjust P1 for approx 50 to 80mA. Power off, remove the multimeter & reconnect the wiring to the power amp. I would use the 10A range on the multimeter.

Bias really does need to be set up, but if you haven't moved P1 and you cant use either of the earlier methods, then leave the amp powered on, with no signal & no speaker. After (say) 30 minutes check the temperature of the mosfets, if they are slightly warm only, you are probably ok, plug in a speaker, put signal through the amp & if it sounds ok, you should be good to go.
If they are too hot to touch, the bias needs to be properly adjusted.
Note the cases of the mosfets will be at rail voltages, so don't touch them directly. You can power the amp off, and wait until the power supplies have discharged to less than (say)12V and then you can touch the mosfet case. Alternatively, you can feel the heatsink temperature.

[Fran]

I have never liked amps with no source resistors on the mosfets, the source resistors help a little with current sharing, and IMHO makes the amp more reliable.
If you can, with power off, insert a 0.22 ohm 5W resistor in the source of each mosfet, then with power back on you can then measure the current through each mosfet by measuring the millivolts across each of the 0.22 ohm resistors.
a rough guide would be 25 to 40mA per mosfet but others here may have a more definitive bias value. 25mA and 0.22 ohm = 5.5mV DC

If you cant do this, with power off, you can insert a multimeter set to measure AMPS DC in SERIES with either the positive or negative supply rail (where it enters the power amp pcb) turn the power on & measure the current with no signal & no speaker. The current through each mosfet is half what you measure, as there are two mosfets in parallel. adjust P1 for approx 50 to 80mA. Power off, remove the multimeter & reconnect the wiring to the power amp. I would use the 10A range on the multimeter.

Bias really does need to be set up, but if you haven't moved P1 and you cant use either of the earlier methods, then leave the amp powered on, with no signal & no speaker. After (say) 30 minutes check the temperature of the mosfets, if they are slightly warm only, you are probably ok, plug in a speaker, put signal through the amp & if it sounds ok, you should be good to go.
If they are too hot to touch, the bias needs to be properly adjusted.
Note the cases of the mosfets will be at rail voltages, so don't touch them directly. You can power the amp off, and wait until the power supplies have discharged to less than (say)12V and then you can touch the mosfet case. Alternatively, you can feel the heatsink temperature.

brilliant. Thanks for the advice. The amps pcb is getting a little ragged by now, due to removing and refitting some of the components for testing. If I hadn't of removed the defective resistor I would never have realised it was faulty. I'd previously tested it in circuit and was getting a false reading, which was why I hadn't grasped the nature of the fault in the first instance. But lesson learned, and it's been fun, fixing it too.

Now, back to the bias adjustment. I've measured it in with the meter as you said in series with the negative rail, and the bias was a little low at just under 40mA. If adjusted p1 and set the bias a little higher at 55mA, for no other reason than its a 'nice' number. I've left it on test for half an hour and it's drifted down a little to 52mA. I've reset it now back to 55 and give or take .1or .2 of a mA it seems to be stable. I've got a digital thermometer and the heat sink temperatures are both exactly the same at 20.3C. The ambient temp is 19.00. So I think it's fixed. Thanks to everyone for your help. Ink now I went the wrong way about it but we got there. Chuffed