Behringer 1222FX power supply UB-SPSU2 - giving me double voltages ?

[chazpope]

I am doing it on its own - no load - mixer works with other psu. Caps were Xunda, replaced C6,C9, C12, C16,C4,C5 also on the HT side C10,C13 (very very tight) - D7 was ok - found a 10ohm unshown resistor in series with it.

no change, if anything the pulsing became faster.

All voltages are about half what they should be, there is no great consumption (after switch off it takes a while to stop pulsing). Before I messed about with it yesterday at least it pulsed between correct voltages and 0. I am measuring the pulsing on the 12v (only it has a load) - the rest pulse less because they have caps but no load.

Tried disconnecting UR+ no change. Before that I measured about 1.5 at the gate of IC3, not sure what it should be.

Not sure about pulling out big diodes again, I pulled out D6 twice and that destroyed the pads (it worked after it went in last) I guess I can cut the AC tracks on all but the +15V to test it next.

Kind of running out of ideas with this crap - if I fail is it possible to buy one...

[diydidi]

How did you check those diodes? DMM are crap at testing diodes. Get something/ multimeter that can put 12v or so over the diode (reverse) while it's testing. Then you will know if they're leaky or not.

[chazpope]

? tested with my meter diode thing - 1.3v one way, .7v the other way - is that not enough ? I have not seen this kind of leakyness. I guess I can swap D6 and D4 - the 12v is not used by this mixer, not sure if pcb will take it, and I can't rebuild the top of it cause it is too busy.

[J M Fahey]

? tested with my meter diode thing - 1.3v one way, .7v the other way

Well, that's a *BAD* reading for any diode !!!

the 12v is not used by this mixer,

That's not the point, diodes must stand way over twice the peak voltage supplied by the PSU.
So D4/5/6/8 must stand at least 40V PIV (I'd specify at least 50V) and D9, say, over 150 PIV.
You are measuring 1.3V reverse across one of them. BAD.

[chazpope]

well my cheap meters just do that when set to diode test - show 1.3V on open circuit and it drops to .7 in forward direction, I wlll re-check tonight - also with higher voltage

I meant I would reuse the diode from the 12V circuit because it is not connected to mixer. not sure it I have a fast diode like these.

[diydidi]

Just make sure the new diodes are fast recovery. A normal rectifier diode will burn up.!!

[Enzo]

Do you have some diodes in your parts drawers? Get out a 1N4007 or a 1N4148 - any basic rectifier. Now set it loose on the tabletop, and measure with your meter diode test function both ways. You SHOULD see something like half a volt one way and open the other. Does your meter show that? I have to imagine even a cheap meter does not indicate 1.3v with the probes not on something. Now compare this experiment with your readings in the SMPS.

[chazpope]

I tried on a FR302 and a UF3003 - this meter shows 1.376V with its probes in the air, it shows .747V on the diode forward direction, also .511V with probes shorted. It has always been like this since I bought it 20 years ago and I have learnt to use this reliably on diodes and transistors not knowing what other meters show. admittedly it is a pocket crap meter Altai DM 5 - somehow I find it more reliable than my other crap meters

Using another meter I get the results that you describe on any diodes incl D6 (FR303) from the circuit - I get about .4V forward and nothing backward. On a SB360 I got .138V but there my 'strange meter' shows .6v so - could it be just a different design of meter? So I think nothing wrong with the diodes - I also did the 12V test - no current flows.

just not to be embarrassed on this forum today I ordered a posher meter

I pulled D4 out of the circuit and put in in place of D6 in the circuit - no change.

I put a blue LED in the 5 V output and it is flashing at about 2hz - this goes on for some time after turn off. 1000uf caps also stay charged. voltage after D6 is about 12v but erratic.

[Enzo]

It sure sounds like something is wrong on the old meter diode test function. For go/no-go tests, even the cheap little Harbor Freight meters will work OK. Those even come up as free with a coupon often enough.

[chazpope]

ok so diodes aside, I cut the AC tracks to D4, D8, D9 to eliminate any short there, I checked D6 umpteen times incl with 12V backwards.

What I get is not even an Autorestart Cycle (TOP245Y pdf says it should be 1Hz). I get a half arsed duty cycle that charges C12 to about 12V and that's that, also it pulses at about 3Hz . This happens even with D6 taken out (no feedback)

Not sure what to do next other than recrete the secondary circuit ex pcb or feed 16V into UR+. I wonder if the C pin voltage can tell me something

[Enzo]

SO, D6 feeds the +UR signal back to the 431 reference, which controls the opto. I mentioned in post #3 that the 431 may be involved. If the 431 cannot control the opto, then the opto cannot control the TOP chip. COnsider, what if the 431 is shorted? Or R24 is open? The D6 voltage ramps up, but the opto never sees it, and reacts trying to "correct" the voltage. The control circuit is trying to either bring down the D6 voltage or bring it up, but nothing it does provides the feedback it needs.

[chazpope]

I will recheck this. I measured about 13V on pin 1 of the 431, same as +UR and this I guess is not enough to make the opto shine. I will check if the 431 it's blown though not sure how.

Also I have on C pin of IC2 a strange cycle of 5.65 to 5.73v every 2 secs. pin 5 of IC4 is showing 15.5V .

Also not clear why IC2 is not cycling at 1Hz when I cut all secondary AC tracks (as it did originally when I pulled out D6) , reading the pdf I figured it should be cycling

[Enzo]

Didn't you say you cut the traces on the secondary side EXCEPT that of D6? I could be wrong about the 431, but that +UR through the 431 and the opto is your control feedback, so if that remains, those other traces won;t matter.

Pin 5 of the opto. OK, I don't know what voltage it should be, but 15v sounds reasonable, it is a simple supply at C13. The 431 controls the LED in the opto, which in turn controls the phototransistor in it. That transistor then controls the curent into pin 1 of the TOP chip.

You have 13v at pin 1 of the 431, so what voltage is at D6, your +UR? Or what voltage is dropped across R24? Just womdering what current is through the opto LED. I don;t read the opto as on or off, I read it as a steady analog signal. If +UR rises, it then tells the TOP to back off some to adjust. So it may be shining, but if the 431 cannot control how much shining from instrant to instant, then the loop doesn't work.

I do not recall, do you no longer have another of these units that works OK?


Is your 13v on the 431 steady? Or pulsing up to that and falling back? Look at pin 3 of the 431, what is that doing? What I am looking for is real evidence the feedback loop functions as a loop. Of course it cdoul dbe other thiungs, like a shorted turn in the tranny, though I find that darn rare. Did we already swap the TOP? I forget.

[chazpope]

no I did not replace the TOP - it is not cheap

My main concern is that if I cut +UR (no control loop) the thing should go full wack duty and into 1Hz Auto restart mode - however it doesn't. This changed as I was replacing caps in the secondary circuit - I was careful but may have blown the top chip in the process.

I checked between IC4 pin 1 and +UR there is 554ohm (R24). The 13V is not steady it sort of floats about 12-14V whether I measure it on D6 UR or IC4. I guess basically UR is lower than the 431 target voltage so maybe there is not enough left to drive the opto so its primary side is shut and there is no control loop.

I guess I can try measuring the opto current next.

Yes I have another unit - it is in my other mixer that I play with every week. I already broke into that to take it out and try with this mixer - it worked - so I put it back. The pcb is not easy to get to the psu unit has its own casing, insulation, HT dividers etc. I guess I could break it out again - and compare voltages but can't risk breaking that too.

Pulling D6 three times totally destroyed the board there but I can still rebuild it - not sure how they rammed these diodes in the holes are just too tight for their wires to go in - unsoldering them is like pulling teeth.

Behringer service told me : don't try repairing these - they break again faster than you can fix It might be they will sell me one for 40 quid - but I don't wan't to admit defeat just yet

[Jazz P Bass]

no I did not replace the TOP - it is not cheap

I guess the word 'cheap' is relative.
$2.00 plus freight.
Find TOP245Y Stock and Compare Prices Across the Most Reputable Distributors in the Industry.