Dead AUX winding on Samsung BN44-00259 PSU ?

[Dannyx]

Good day folks. This has got to be the most perplexing project I'm working on. This might be considered double-posting because I opened up another thread on this issue a while back, but haven't solved it since, however I believe I've now identified the problem or at least narrowed it down by exhausting almost every possibility, so let's go over this one more time: a BN44-00259 PSU from a Samsung TV isn't working, that being said, there is no voltage present at the output AT ALL (not even 5Vsb). This is because I'm not getting anything out of the AUX winding of the SMPS transformer, so the ICs in the primary have no way of turning on, since they have no VCC. Have a look at the schematic and the one I marked out in paint: the red line shows the path of the current from the main cap through the SMPS transformer (just ignore PFC and other BS for now), while the blue one represents the AUX winding: to my understanding, when current flows though the red line, it creates some voltage in the blue winding as well, which is rectified by DM802 and should be around 15-16v which then goes to the ICs and optocouplers and whatnot. That's the value I got looking at the VCCs required by the various ICs which run off that AUX winding (namely UM801s UP801s, etc) as well as by comparison with a functional identical board. However what I get is half a volt when the DM802 is connected to the circuit and 1-2v when it's "floating" (the cathode pin is lifted from the board and I probe it straight with my DMM). I have checked and even replaced the diodes and caps time and again. RM890 also reads fine, plus it doesn't matter whether I solder my test diode before or after it (I call it a "test diode" since I don't have a scope and the frequency of the AUX winding would be too high to measure it straight off the transformer with my DMM so I have to place a diode between it and my meter to get SOME reading). The weird part is that I went through the trouble of desoldering the transformer from the functional board and swapped it with the one I considered faulty, but it changed absolutely nothing: the functional board continued to run properly even with the transformer I considered faulty and I also get continuity between GND and AUX (blue line). The same procedure was applied to UM801s, diodes, capacitors and just about every component on the primary that isn't SMD: replace or swap with parts from the working board which STILL ran perfectly fine...I'm at my wits' end basically...any suggestions ?

[dskall]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

Do you have a list of components you have swapped? Have you compared voltages on UM801S

[R_J]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

The ic gets its start up vcc internaly from the Drain pin via a "Startup Cell", Have you swaped the ic's around with the other board?

[Dannyx]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

The ic gets its start up vcc internaly from the Drain pin via a "Startup Cell", Have you swaped the ic's around with the other board?

Yes, I have purchased a new IC and even swapped it with the other one, thinking it somehow got busted during soldering or something. I'm familiar with its mode of operation, so I DO see around 15v on the VCC pin being sourced from the D pins. This happens even if DM803 is removed so the IC appears to work properly, but can't enter normal operation mode because that VCC is not sustaining.

[Dannyx]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

Do you have a list of components you have swapped? Have you compared voltages on UM801S

The list would be slightly too long to compose, but let's just say everything that isn't SMD on the primary got replaced or swapped (including some which are just completely silly and would in no way cause the issue, like the bridge rectifier, main cap, cap CP801s, even some of the bloody ferrite beads - I went that far !) Speaking of paranoia, one thing that's slightly different between the two boards is that as I was soldering and desoldering stuff, eventually the pads and traces starting falling off so I had no choice but to create nodes and bridges using solder and plain wire, some of these "shunting" some of those beads (imagine LB807 and LB808 are still there, but paralleled with some pieces of wire on the solder side of the board). Of course, I wouldn't be daft enough to ever shunt "active" components like resistors or diodes like this - I'd only go up to the nearest solder joint and blob up the solder and extra wire there to essentially recreate the broken trace leading to that component. I doubt it's THAT sensitive to stuff like this though :|

[dskall]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

Around 15v. Looks like minimum VCC turn-on is 17v

[Dannyx]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

Around 15v. Looks like minimum VCC turn-on is 17v

I know, but that's another strange thing: what actually IS the minimum threshold for the ICE3BR IC ? Under the "absolute maximum ratings" the datasheet points to "VVccOff" as the minimum voltage and if we go lower on the same page, we see it's 9.8v. However, the "VCC turn on threshold" immediately above that says 17v, so can I go as low as 9.8v or 17v ? Either way, I recall seeing 13-14v on the functional one with PFC up and running and all. It could be that, initially, when the board is first plugged into the wall, current flows through the transformer windings 1>3, to the D pin and then to the VCC pin as described in the datasheet and the VCC pin does reach 18-19v very briefly and is then allowed to go lower once the IC has "started", but all this happens so fast you can't capture it without a scope, so all you're seeing is that "normal operation VCC" let's call it.

[dskall]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

Well budm would be the one to consult about this.

[Dannyx]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

Well budm would be the one to consult about this.

Sure hope I see him around...he's helped me a lot since I've been on here, though every tip is welcome

[budm]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

With the board by itself, hook up DCV meter to the two legs of the filter cap on the output side of this power supply, then apply the power to see if you see any dcV spike showing on the meter. If you do, it means the circuit did try to start up. The rectifiers on the cold side are all good, same for the Zener in the cold side around the IC are good?

[Dannyx]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

With the board by itself, hook up DCV meter to the two legs of the filter cap on the output side of this power supply, then apply the power to see if you see any dcV spike showing on the meter. If you do, it means the circuit did try to start up. The rectifiers on the cold side are all good, same for the Zener in the cold side around the IC are good?

I actually did try this on my own but didn't post the results because the post would've been astronomically long, so I only post details when asked for, so: no shorts anywhere on the board as far as I can tell (even checked most of the SMD stuff, primary and secondary). When the board is first plugged in from a cold start, interestingly, I do get some voltage after DM957 (TO220 diode in the secondary) and up to the drain of QM801. The other secondary winding (pins 11-12) doesn't put out anything on its diode, where those 5v should be coming from. Can't remember the value I get exactly, however it's definitely not withing spec - like 2-3v or something. However when I unplug the PSU I am somehow able to create some sparks if I short the legs of diode DM857, so even though there doesn't appear to be any significant output, there IS something there. It's no doubt created by those 320 volts flowing through the primary (1>3) of the transformer, despite pin 3 not getting switched by UM801S ICE3BR.

[dick_barton]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

Grasping at straws here but have you checked the DC resistances between pins 10 & 12 on the primary and 8 & 9 on the secondary of TM801s?

[budm]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

Since you did get some Voltage spike on the output side that mean the current MUST have flown through the primary winding and complete the path all the way back to the power return ground (the negative leg of the main filter cap), so that means the MOSFET inside the IC was turned on for seconds and for what ever reason it stops switching due to fault some where.

[gabiz_ro]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

Remove QB801 ant test without QB801.

Also,from a quick look,you could remove DM857 and DM852 (to eliminate any anomalies that may appear if something wrong on these rails) power source still can power up and deliver only 5V (stand-by)

[Dannyx]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

Grasping at straws here but have you checked the DC resistances between pins 10 & 12 on the primary and 8 & 9 on the secondary of TM801s?

The pins you mentioned are ALL on the secondary, though I did check those and my meter beeps, indicating a very low resistance. Without a proper transformer tester I can't tell what it's doing internally though, since a DMM is not the proper way of testing transformers at all. If you mean pins 5-6 on the primary, those have continuity as well.....just the idea of removing the transformer again makes me sick :|

[Dannyx]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

Remove QB801 ant test without QB801.

I did that, thinking there's a short somewhere which pulls down the AUX winding, but it didn't change anything: I still get squat on its collector pad (since the collector goes straight to the AUX winding and should have voltage present at all times). I removed it on the functional board as well and sure enough, there is 15-16v where its collector goes. The transistors are both fine as they both work on the board which is functional.

[gabiz_ro]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

Earlier say that changed already UM801S and swapped TM801S and no result
After test without DM857 and DM852 and still nothing replace PC801S and check or replace CM801 CM802 CM807 RM869 ZM801

[Dannyx]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

Earlier say that changed already UM801S and swapped TM801S and no result
After test without DM857 and DM852 and still nothing replace PC801S and check or replace CM801 CM802 CM807 RM869 ZM801

I had already swapped all those components with ones from the working board, but the problem's the same: that one continues to operate very happily, while the faulty one is still dead.
There's also another rather interesting detail I left out, which is probably the most annoying part of this whole repair: if I put the negative probe of the DMM on the primary ground and the positive one on either of the Drain pins of UM801s, the supply suddenly comes on, PFC and all....To my mind, what happens is the DMM allows just enough current to flow through the primary of the transformer (from pin 1 through the meter and to ground) to excite the AUX winding which sustains the VCC of UM801s. This could also mean that UM801s could STILL be faulty, despite initially installing a new one. I say "initially" because after putting it in and still not getting anything, I swapped it with the one from the working PSU, then also found a third IC lying around, installed that one, then one of them actually popped after somehow shorting that BA pin to ground and it's just......aaaaarrghghh...I honestly lost track at this point. The point is that despite pointing my finger at the transformer, it could be UM801s that's still to blame.

[redwire]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

One trick is to make the aux. voltage using batteries, a couple 9V with series diodes from the junkbox to get down to say 16V or above the UVLO. Sometimes 18V is fine.
This allows the (standby) SMPS IC to stay running.
The batteries are floating at hazardous live potential, but at least voltage measurements can be made.

Because the SMPS IC has (battery) power before mains is applied, soft-start sequence is expired but nothing has blown up yet when I do this.

I wonder if your snubber is working, since the multimeter on the drain adds capacitance to that.

[dick_barton]

Re: Dead AUX winding on Samsung BN44-00259 PSU ?

Access to an oscilloscope and isolated input probes would come in really useful at this time.