Help, SMPS Question

[Alembic]

Hi everyone

I am attempting to repair a PSU from a
Soundcraft eSeries mixer.
The issue is this: It appears to have gone
massively overvoltage. According to the schematic
the voltage across the reservoir cap should be 170Vdc
I am reading just over double that. Visual inspection
of the board clearly shows excessive heat around some
resistors. I replaced a few suspect components and
managed to achieve +and- 15v on the output but
that didn't last for long. Back to square 1.. struggling.
Q: What part of the design defines the 170 vdc ?
Thanks for any insight or assistance.
Schematic is available.

[domas]

Re: Help, SMPS Question

post pictures and schematic so. else don't expect too much help as it is all very vague.

[Alembic]

Re: Help, SMPS Question

Sure thing. just added them. should be
there by now.

[SteveNielsen]

Re: Help, SMPS Question

What components did you replace?

[everell]

Re: Help, SMPS Question

Is your line voltage (incoming power) 230 volts 50 hz or 120 volts 60 hz? The schematic you posted looks like something for 120 volt application rather than 230 volts.

[SteveNielsen]

Re: Help, SMPS Question

Are you sure you did not use the AC scale on your meter to read DC? That can approximately double the reading.

[eccerr0r]

Re: Help, SMPS Question

Looks like a typical UC3842 design. A lot of UC3842 designs are autosense 120/220VAC. The 170V will be 300+V or so on 220VAC and is normal.

R13 which is common to most UC3842 designs does get warm when used at 220VAC. A lot of these designs also fail to come up when C1 or C8 fails due to leakage. I've seen D2 die too because someone was cheap and used a 1N400x there instead of a high speed or Schottky diode.

[Alembic]

Re: Help, SMPS Question

What components did you replace? R13,D4 D7 D6 Fet1,ic1
R13 Value read was ok but replaced anyway due to visible
appearance it does get WARM.
Line voltage: 240AC 50Hz
Proper Multimeter settings used re: AC/DC

eccerOr: D2 ok but I replaced it with same component make.
Will replace C1,C8 - unable to check capacitors properly as
I do not have cap tester.
D6 gets hot very quickly. i.e. in the time it takes me to
take a few quick measurements and then, if left
long enough,becomes u/s.

Thanks to all.

[CapLeaker]

Re: Help, SMPS Question

what about D5, C20 and R15?

[budm]

Re: Help, SMPS Question

Scotland AC Voltage is 220~240Vac. The diagram is shown for 120V operation.
What is the exact model of this SOUNDCRAFT?

[Alembic]

Re: Help, SMPS Question

what about D5, C20 and R15? D5,R15 reading ok.
What is the exact model of this SOUNDCRAFT? E8 series mixer.

Thanks

Incidentally I came across the following text file.
Seems I`m not alone.
===============
When I saw the first couple of these mixers with catastrophic PSU failure, I thought of them as random events. Now I've seen at least half a dozen I'm beginning to think there's a design flaw. The PSU is a bog ordinary switchmode, and the failures are all the same - high ESR caps around the 3842 PWM controller, blown 100 volt clamp zener, blown +48 volt rail rectifier. After that there are a few variations on the theme - blown chopper fet, source resistor and 3842, blown resevoir caps on +/- 15 volt rails, and most sinister of all,almost always vast slews of blown 5532 and TLO74 opamps in the mixer itself. What is obvious is that the PSU goes massively overvoltage when it fails. What i'm trying to figure out is WHY?. These sorts of PSUs are common to masses of rack gear etc and are failing all the time (usually high ESR caps) but they manage to fail without killing so many passengers (so to speak). What is it about the EPM PSUs that makes the failures so destructive. I notice that Soundcraft use 85 degree caps in areas which (due to board browning) obviously get plenty warm.

[rievax_60]

Re: Help, SMPS Question

A faulty C8 will cause high output voltages and shorted clamp zenner, R16 also if it goes high. Because the regulation is sensed on the live side, notice no optocoupler.
This design should always include a 10K bleeder resistor from Gate to Source of the MOSFET to prevent the MOSFET turning on due to leakage current in case the UC3842 doesnt start for some reason.
Faulty output rail capacitors could cause the clamp zener to overheat and fail also. The capacitors directly after the rectifier diodes are the most important.

[eccerr0r]

Re: Help, SMPS Question

D6 (the 100V zener) should not really heat up, seems kind of weird - if it's not oscillating it seems a bit difficult to warm up that diode? Is any outputs working?

From what I've seen, these UC3842 based PSUs all are cheapcheapcheap. They don't last forever. Bad capacitors hurt. They were made to do 120V-240V making it more stressful when using it at one extreme. They are not terribly efficient either (though not horrible - 70%-ish).

It looks like the +48V is "slow start" which I suppose is normal for audio equipment.

[Alembic]

Re: Help, SMPS Question

Hello everyone.

Responded to your suggestions re
components to check.
As it stands, I replaced the UC3842N as I felt
it could be defective.
I then decided to attach a dummy
load of two 12v auto bulbs to the secondary
+and- outputs and powered up.
The bulbs are pulsating and light dimmly this
is also accompanied by a light ticking sound (approx
5 ticks a second) Re-checked board for open/dry joints
but seems ok.
R13 and D6 do get warm but nowhere near as warm as
before. Is this condition indicative of a particular failed
component?

Thanks everyone.

[CapLeaker]

Re: Help, SMPS Question

So if IC1 was bad and FET1 was fried, then R2 should have been fubar too or at least it should be high in value.

[budm]

Re: Help, SMPS Question

"load of two 12v auto bulbs to the secondary" What are the Wattage of these two lamps? What are the cold resistance of these two lamps?

[Alembic]

Re: Help, SMPS Question

Hi.
In reply.

R2 should have been fubar: R2 reading ok =22 ohm.
FET was ok. However made a standard re-check of FET
i.e. G,D,S check, appears to be ok.

What are the Wattage: One is 55w the other is 21w (It`s what I had
available to hand)
What are the cold resistance of these two lamps? approx. 0.4

Thanks.

[budm]

Re: Help, SMPS Question

"What are the cold resistance of these two lamps? approx. 0.4" 0.4 Ohms, that will cause the over current protection to kick in due to high inrush current.
"What are the Wattage: One is 55w the other is 21w (It`s what I had
available to hand)" 15V with 0.4 Ohms = 37.5A

[Alembic]

Re: Help, SMPS Question

Hi.

Point taken. What would be a good dummv load
to test this type of supply?

Thank you.

[budm]

Re: Help, SMPS Question

I do not know what the current capability of the power supply is, so I would use 30 Ohms 10W loads to get 0.50A.