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

[Dannyx]

Good day folks. I've had this Samsung TV lying around for almost a year now and not having anything else to do today, I decided to dust it off and see what is actually wrong with it. Obviously my first observation was that it's not getting any power AT ALL, because the LED is not coming on and there's no 5Vsb present at the output of the power supply. I took the power board off and taking some measurements revealed no blown or shorted parts, however I'm not getting anything on the secondary. I get around 317V on the main cap (CP815) and 17v on the VCC pin of UM801S, but nothing else. Here's the weird part: if you look at the schematic (to my mind at least) the VCC for UM801S can come from nowhere else than through DM803, from DM802 and RM890 from what seems to be the AUX winding of the transformer. That makes perfect sense and having said that, you'd expect to see 17v after DM802 as well, but I get nothing, despite getting 17v after the other diode (DM803) just fine, so how does this work ? Voltage just magically appearing like that ? :| I've already replaced CM804 and DM802, but nothing changed. Despite inexplicably getting the 17v at VCC of UM801S, nothing comes out of the secondary...any ideas ? Cheers. I've uploaded the schematic for your convenience and I know the second picture is a bit poor...

[Dannyx]

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

The pics didn't upload for some reason....there you go

[Dannyx]

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

Oh, they DID upload.....sorry Messed up this post big time.

[dick_barton]

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

That makes perfect sense and having said that, you'd expect to see 17v after DM802 as well, but I get nothing, despite getting 17v after the other diode (DM803) just fine, so how does this work ?

One side of DM802 would be the rectified voltage, 17V DC in this case but the other side will be an AC voltage.

[Dannyx]

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

One side of DM802 would be the rectified voltage, 17V DC in this case but the other side will be an AC voltage.

The cathode side should show 17v, but it doesn't - that's what bugs me :|

[dick_barton]

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

The cathode side should show 17v, but it doesn't - that's what bugs me :|

What AC voltage does it show. Is the diode s/c or gone resistive.

[Dannyx]

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

What AC voltage does it show. Is the diode s/c or gone resistive.

The frequency of the output of the transformer is too high to be able to read with my DMM. Nothing appears to be shorted on the board...

[dick_barton]

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

I would replace the diode just to satisfy myself that there's no issue with it.
Two diodes in series. First diode has no DC voltage & second diode is showing 17V DC???
I agree with you that does not make sense.


Edit
The PFC isn't working since there's only 317V on CP815, so I am puzzled as to how the transformer could provide any secondary voltages.

[srhofmann]

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

Is it possible that UM801S isn't turning on? A quick look at the datasheet shows it has undervoltage lockout below 18V on Vcc. From the datasheet......

The Undervoltage Lockout monitors the external
supply voltage VVCC. When the SMPS is plugged to the
main line the internal Startup Cell is biased and starts
to charge the external capacitor CVCC which is
connected to the VCC pin. This VCC charge current is
controlled to 0.9mA by the Startup Cell. When the VVCC
exceeds the on-threshold VCCon=18V the bias circuit
are switched on.

[Dannyx]

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

Is it possible that UM801S isn't turning on? A quick look at the datasheet shows it has undervoltage lockout below 18V on Vcc. From the datasheet......

The Undervoltage Lockout monitors the external
supply voltage VVCC. When the SMPS is plugged to the
main line the internal Startup Cell is biased and starts
to charge the external capacitor CVCC which is
connected to the VCC pin. This VCC charge current is
controlled to 0.9mA by the Startup Cell. When the VVCC
exceeds the on-threshold VCCon=18V the bias circuit
are switched on.

Ok, but what would be the cause ? I too agree that 17v is a bit too low, since 19v is the norm for such ICs, at least usually...

[Dannyx]

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

I would replace the diode just to satisfy myself that there's no issue with it.
Two diodes in series. First diode has no DC voltage & second diode is showing 17V DC???
I agree with you that does not make sense.


Edit
The PFC isn't working since there's only 317V on CP815, so I am puzzled as to how the transformer could provide any secondary voltages.

I removed the second diode (DM803) and I still get 17v on the VCC pin, so it's definitely coming from somewhere else.

[dick_barton]

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

It would be worthwhile checking CM808 if VCC is not initially charging to the min 18V via the soft-start as suggested by srhofmann.

[srhofmann]

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

With DM803 removed what do you see on the output of DM802? Should see more than 22v. ZD803+ZD804 are 22v plus a little extra for the current limiting resistors RM807/8 so they should be fed with more voltage than that.

With DM803 removed have you tried measuring the voltage on the Vcc pin with a lower impedance voltmeter, think old fashioned analog? You might be measuring the voltage from a coupled line that can produce uA level currents that you meter can see. 17V is only 1.7uA if you have a 10 MO meter impedance.

If you don't have an analog meter put a resistor, 470k/1M, across your meter leads and then remeasure the voltage.

[Dannyx]

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

With DM803 removed what do you see on the output of DM802? Should see more than 22v. ZD803+ZD804 are 22v plus a little extra for the current limiting resistors RM807/8 so they should be fed with more voltage than that.

With DM803 removed have you tried measuring the voltage on the Vcc pin with a lower impedance voltmeter, think old fashioned analog? You might be measuring the voltage from a coupled line that can produce uA level currents that you meter can see. 17V is only 1.7uA if you have a 10 MO meter impedance.

If you don't have an analog meter put a resistor, 470k/1M, across your meter leads and then remeasure the voltage.

With DM803 removed I get the same nothing on the output (cathode side) of DM802...starting to question how that transformer even works. From what I can see, pin 6 is GND and 5 is the AUX winding, but what about the others ? Pin 3 appears to be pulsed by the Drain of the ICEBR IC, which at the same time is also rectified through DM801 and goes back into pin 1 alongside the voltage coming from the main capacitor....damn this is getting confusing :| Either way, starting to think perhaps the IC is not working correctly...don't have an oscilloscope to probe any pulses on the drain sadly

[dick_barton]

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

Without reading the whole of the datasheet it looks like the capacitor connected to VCC of UM801S is charged via an internal current pump circuit within ICE3BR0665J and gets it's initial voltage via LB808, LB807, TM801S pin 3, TM801S pin1 to supply.
Once VCC reaches 18V the circuit starts to oscillate and provides its own supply to remain running via RM890, DM802, DM803 and maintains the voltage on CM808.
Because UM801S is now self oscillating it also provides the PFC voltage across CP815

[Dannyx]

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

Without reading the whole of the datasheet it looks like the capacitor connected to VCC of UM801S is charged via an internal current pump circuit within ICE3BR0665J and gets it's initial voltage via LB808, LB807, TM801S pin 3, TM801S pin1 to supply.
Once VCC reaches 18V the circuit starts to oscillate and provides its own supply to remain running via RM890, DM802, DM803 and maintains the voltage on CM808.
Because UM801S is now self oscillating it also provides the PFC voltage across CP815

Had a look at that datasheet myself and came to the same conclusion, however I can't tell why VCC never reaches that 18v sweet spot...

[dick_barton]

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

There could be a number of reasons.
1. UM801S is faulty
2. The capacitor CM808 is leaky or something around it is, for example the ZD803
3. When UM801S bursts into life then the internal current pump switches off but if CM804 is leaky it cannot hold the voltage created by the secondary winding pins 5&6 so the internal charge pump switches back into life and the cycle repeats.

[srhofmann]

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

Here's the datasheet for UM801S it's got a pretty good explanation of how things are supposed to work.

And I just saw your post that you looked at the datasheet......

Did you happen to check the 2A fuse FM802B?

[Dannyx]

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

Here's the datasheet for UM801S it's got a pretty good explanation of how things are supposed to work.

And I just saw your post that you looked at the datasheet......

Did you happen to check the 2A fuse FM802B?

Actually that would be FM802S (B is the clip itself ), but yes, of course. The fuses are the first things I check before even removing the board from the Tv...if it's not terribly expensive, I'll try replacing the whole IC just to rule it out. Might actually have some vestel boards around which use this same IC, as I distinctively remember that "ICEBRxxx" deal.

[srhofmann]

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

Actually that would be FM802S (B is the clip itself ), but yes, of course. The fuses are the first things I check before even removing the board from the Tv...if it's not terribly expensive, I'll try replacing the whole IC just to rule it out. Might actually have some vestel boards around which use this same IC, as I distinctively remember that "ICEBRxxx" deal.

Well I had to ask about the fuse, sometimes it's the little things that bite us in the ass.......