Need Help with a Yamaha DXR10 Smoking SMPS

[Briesmith]

I am trying to repair a Yamaha DXR10 powered amplifier but finding it difficult because I have a smoking resistor which starts to burn as soon as I switch power on. (So I have it turn it off straight away.) The fuses don't blow and I connect a dim bulb which goes dim so no straight short to find. I have attached a marked up circuit diagram and I am looking for help in identifying the best place to look for what's causing the excessive power draw. Needless to say the amp isn't working. The mosfet circled in red was the start of my problem when it exploded. I replaced it and the resistor network also circled in red. The circuit in question and the components I am suspicious of are circled in yellow and numbered in blue.

No 1 marks the resistor that smokes.
No 2 marks a mosfet that seems to test OK.
No 3 marks a resistor that tests OK.
No 4 shows where the circuit joins the PWM.
No 5 ids the last capacitor in the circuit before it goes to ground.
No 6 shows where the mains AC seems to provide a GND for the circuit. (Don't understand why the mains AC joins the rectified AC coming from the bridge rectifier at No 8.)
No 7 circles a structure I just don't understand.

If anybody can advise where I should start to look for the source of excess current I will be grateful.

Best wishes

Brian

[stj]

check c108 and d101 on the far left, and c118 - the big cap on the right.
it's typical japanese design - overcomplex to use less chips and keep the resistor and transistor companies afloat!
i cant really tell what the hell it's doing!

[stj]

looking at it more, the circuit feeds back on itself so much you may find it quicker to check none of the small transistors are shorted than to try to understand what it tries to do!

[Briesmith]

Thank you Great Sage for taking the trouble to look at my problem. There are several structures, clusters of components, that I have no idea what their purpose is. And I've never encountered a design where the rectified output from from the Bridge rectifier is joined to the unrectified mains in as in this case. But thanks again for trying to get your head round things for me, much appreciated.

[stj]

another aproach may be to get a datasheet for the pwm chip and start looking at the signals on it to see if any are missing.

[redwire]

Mosfet Q102 and the smoking resistor R116 are the start-up power for the primary PWM section.

Normally, if mains voltage is enough and C108 charges up there is enough gate voltage to turn on Q102 which supplies power to the PWM control IC 12VDC rail (4).
The SMPS is then supposed to start and run. Then power is from IC102 the 12V reg. from the aux. winding. Which also causes start-up mosfet Q102 to get switched off by the aux. winding also turning on Q101, once the SMPS is running.

I would guess the SMPS is not running and it ends up spending too much time in start-up mode and the resistor runs hot and smokes, as it should. It's only supposed to be on Q102/R116 for a few seconds.

So I would check the SMPS for faults. Power mosfets Q106, Q107 and the sense resistors etc. The array of six resistors R148-R151, R211, R212 is just to save money instead of using one resistor there. They might have fused open-circuit?


I have heard of some people lifting a leg on the smoking startup resistor to disconnect it, and then powering the SMPS IC from a couple 9V batteries (for 18V) at point say TL107 and seeing what the PWM controller is doing. But this is hard, a last resort, powering it from mains and poking around with a scope etc. No fun.

Weird the UCC2893 datasheet show the IC has a start-up power circuit built in at pin 16. So that is supposed to power the IC as well. I would say it's not enough to power that nutbar eco-green pulse transformer snubber thing, so Yamaha added all that shit.
Yamaha does have bizarre electronics designs that end up no cheaper than if they did it in a straightforward manner.

[stj]

actually the line of resistors is not to save money,
it's to increase the length from end to end to minimise the chances of it arcing over if anything goes wrong.
it's a standard thing in high voltage circuits to increase the gap a spark would have to jump across.

[Briesmith]

Grateful thanks to you both. The main buggeration is there's no time to test or measure anything before I have to switch the amp off to stop the resistor bursting into flames. At the moment I am removing items of interest and testing them on the bench to see if they are OK or not. But that's very destructive and a risky approach as it can introduce errors and faults that weren't there previously. Fault finding on equipment you can't turn on is very difficult!

[redwire]

I was talking about the array of six sense-resistors to make up a 39.6mohm 6W which ends up well over 18A really so Yamaha must only populate one or two of them. They are all so small ~0.051-0.068 ohms 1W each they must leave some out.

To power this thing for a long time, I'd disconnect the smoking resistor and use a 12V battery (that can be hazardous live it will be "hot") to power the PWM controller IC. It's dangerous to poke around a live primary section either way.

[R_J]

If Q107 shorted, causing the source resistors to go open, The UCC2893 is very likely bad, When the source resistors went open, and Q107 was shorted, around 380vdc was sent to the CS pin of the ic, if the mosfet was shorted drain to gate it likely damaged the ic's gate drive as well

[Briesmith]

Thanks again everyone. RJ don't you know it's Christmas? Supposed to be a time of goodwill! (I'm secretly dreading the rational conclusion that I will have to replace the PWM IC if only to eliminate it as failing but I don't want to do that until I'm sure there's not still a circuit problem that will just blow it up again.)