Dead Samsung LE28b450 (no 5vSB) BN44-00259 power supply

[Dannyx]

Re: Dead Samsung LE28b450 (no 5vSB) BN44-00259 power supply

I have another update: I believe the transformer may have gone open circuit. Have a look at the schematic: pins 1 and 3 of the transformer should have a very low resistance and produce a beep when I place my meter across them, however this doesn't happen. Neither do pins 1 and 2, however pin 2 (the unconnected center tap) and pin 3 do produce a beep, so either the windings between 1 and 2 snapped somewhere or the resistance is indeed supposed to be high, which I kinda doubt, because most SMPS transformers don't behave like this.

[Dannyx]

Re: Dead Samsung LE28b450 (no 5vSB) BN44-00259 power supply

EDIT: Christ, now I feel like a dumbass....just forget what I previously wrote: I miscounted the pins >_> One of the pins of DM802 is perfectly in line with the transformer, so I count that as pin 1, hence the error...I'm sowy

[Dannyx]

Re: Dead Samsung LE28b450 (no 5vSB) BN44-00259 power supply

UPDATE (hope someone's still around): I came back to this power supply since I had a little extra time on my hands and also scored an identical working board, so it should now be easier to pinpoint the issue by comparing the two. The first thing I was going to do even before getting the working board, though never had time, was replace UM801S ICE3BR0665J - I did that today but nothing changed unfortunately: I still get no voltage on DM802 and the supply doesn't turn on. Then I started swapping components from the faulty board with ones from the working board one by one, thinking that it eventually HAS to turn on this way, at which point I'd know I have found the culprit part. I did this nearly all day long and swapped just about every through-hole component on the primary side but to no avail: the functional board continues to operate properly even after populating it with parts from the faulty one, which of course doesn't work still....I even replaced the transformer and LB808 and LB807 which is silly as all hell when you think about it, since those are just bloody pieces of wire, but I thought that maybe somehow they would make a difference. However, something interesting happened which brings me just a few steps closer to figuring out this prick: as I was taking some measurements with the supply plugged in, I tried measuring between primary ground and the "D" drain (pins 4 and 5) of ICE3BR0665J. When I touched my positive probe to the D pins, I heard a faint click and the voltage I got was around 397v, which is what it should be, since the PFC should switch on immediately after plugging the board in. In other words this board was now up and running, with all the voltages present and stable. As you might have guessed, after unplugging it from the wall and discharging the main cap, the earlier problem reappeared, though I was still able to trigger it to turn on using the meter trick again...what's up with this ? I'm guessing the meter allows just enough current to flow to trigger the internal charge pump of ICE3BR0665J (from what I've read in the datasheet, current initially flows from the D pins to the VCC pin). Not being able to figure out why the IC doesn't switch on by itself, I'm thinking about placing a small capacitor between GND and the D pins, hopefully replicating the meter trick...worst case scenario it's gonna go bang though I doubt it would be anything too violent, since that cap would be nothing more than a so-called "snubber" which is often seen between the Drain pin and the Source pin of FETs in SMPSs. I'm thinking of dropping it right after LB808.