Samsung Plasma TV PN64H5000

[csplinter]

My wife was home this morning watching the TV when the breaker tripped about 5 minutes after having powered it on. Upon resetting the breaker, the TV would no longer power on.

I tested the outlet, then connected the TV directly to the wall. There was no standby light, or any sign of life whatsoever. I suspected a problem with the main power supply, so I visually inspected it finding no signs of burning marks, or bulging capacitors. The only flaw that I can find is that the main fuse is blown.

What is the next step? If I replace the fuse I will have to solder a new one with this board. I really want this working again as soon as possible, and for a long time to come. I would hate to close this up only to experience a repeat soon after. This TV is a beast to move on and off the wall. Maybe there are some parts I should buy in case needed, or to proactively replace?

This board is the bn44-00690b rev 1.0

Thank you so much, in advance.

[JoeO]

Search youtube "pn64h5000 power supply'.

2 videos about this power supply.

[csplinter]

Thank you for looking, but there are actually no videos about this particular power supply that I could find. The top result is for a PN64E8000, which is similar, but the fix involved buying a board pulled from another TV!

If there is another one I'm missing please show me. I had researched this quite extensively, and found one thread on AVSforum where a user had the exact same circumstances as me, but once again solved the problem by replacing the board.

Perhaps I'm asking too many questions at once. Let my main question be, "for those who have plenty of experience fixing TVs, would you recommend just swapping the fuse first?" My goals are primarily to limit any possible damage. I would feel pretty foolish if I burned something up when I could have just purchased a working board. And only very secondarily, "what can I do to harden this thing?"

[csplinter]

Sorry for self bump, but I just noticed that the price of in-circuit ESR meters has fallen. I just got one from Amazon, so I will have that at my disposal Monday.

[keymaster]

What tools do you have ? Any exp. with electronics and mains power?
replacing the fuse only ,almost never solves the problem . fuse blows when a problem is present.
remove the power board form the tv
Start from the mains inlet and go dipper
Check the bridge, diodes and mosfeds. you are looking for a short.

build a test rig with a lamp(60 watt incandescent) , in series with the board mains input. So you dont burn the fuse all the time you check the board.

attention : discharge all the caps , they bite. ( use a low value resistor or a useless screw driver )   

[csplinter]

I have a multimeter, a good soldering station, and coming in the mail an in-circuit ESR meter and a multi-component tester/identifier.

I'm in copier repair and deal with high voltage on a regular basis, but don't usually do any component-level repairs.

I think I have identified two bad diodes. I'm getting a resistance reading in both directions, and 1.8v measuring with leads connected backwards with my meter in diode mode. I get .523v measuring with leads connected to the proper ends. Can someone please confirm my thinking that these seem bad? They can be properly tested on the board, right?

They are circled in the following images

[shakalnokturn]

Hard to tell about the diodes without lifting a leg. Probably OK though.

Start at the beggining:
The relay has a single contact, test resistance between neutral and live on either side of the relay.
One side will tell you about the line filter, the other about the  diode bridges.

If that tests good, move along to measuring between + and - at the diode bridges, that in turn will give you a clue on the PFC switching transistors.

If all is well move along to test across + and - of the three bulk capacitors.

If there is no short there either you can risk replacing the fuse and applying power.

[PKTKS]

As said above is not possible to say that wo schematics and detailed analysis

To make things even worse CAPs on such high power devices will hold the
charge for long periods.  Even discharged those caps can "rebuild" the charge
by internal means...

The proper way is removing or just lifting..  very time consuming as those
parts are almost always attached to huge heatsinks

Just don't risk your DMM on such CAPs which may have high left over charges..
the spikes are high energy

Paul

[csplinter]

Thanks all. Good to know that the diodes can't be tested in circuit.

I have continued to testing and confirmed the function of the bridge rectifiers, and have not found any shorts. Later today I will have my new MESR-100 ESR meter and I will confirm the function of the capacitors. If all goes well, I will try a new fuse tonight.

[csplinter]

I couldn't detect any fault with the electrolytics using the MESR-100 meter, so I proceeded to replace the fuse. I immediately heard a sizzle which I assumed was the fuse blowing. Upon further inspection I noticed a crack in this bit that I've yet to be able to identify. It has no visible markings.




Maybe it was there all along, or maybe it just cracked, but I didn't notice it before.

[csplinter]

It's a thermistor. I haven't tested any ICs other than the bridges yet. I have a tester coming tomorrow for those. I'm disappointed to see that this is just another safety device, because it seems like it's probably just indicative of further problems elsewhere, correct me if I'm wrong?

[Pete2]

If the fuse and the thermistor are blown then there might be a short somewhere, maybe the rectifier brigde? Try measuring with Ohms mode (make sure no power applied and caps discharged before doing that). If no short found, you could replace the thermistor with a suitable resistor or a link as a temporary solution. Of course best to replace it with the same type eventually assuming there's nothing that will cause it to blow again.

[csplinter]

Thanks, Pete. I have already confirmed the function of the rectifier bridge. There is no short present there.

I will try again tomorrow.

[PKTKS]

To make things even worse CAPs on such high power devices will hold the
charge for long periods.  Even discharged those caps can "rebuild" the charge
by internal means...

Free energy confirmed!

(edit: this was a joke, dielectric absorption, etc)

No  no no this is no joke..

Depending of the chemistry high quality caps may rebuild part
of the energy drained entirely by internal chemical re-arrange

Discharging just before you test some area or device powered by such
high quality ELCOs (every time you do so..) is a clever procedure

On the other hand BAD QUALITY CAPs with poor chemistry and
high parasitics (ESR and EPR) will never be able to recompose their plates...

Odd.. but sometimes happens

If you use a tester powered by 3.3 LDO and not prepared
to handle spikes of say 35 or may be 50 V you will have trouble...

a 1000V or 800V HIGH quality CAP sitting for some time (after
periods of full charge rate) may damage your fragile 3V tester

Paul

[shakalnokturn]

Still check PFC and main PSU switching transistors for a possible short, if they appear good here's what may have happened (a little guessing so I may not be spot-on!):

The NTC thermistor is there to limit inrush current when power is applied (gives the fuse and diode bridges an easier time).
The standby supply probably runs on the NTC all the time. Once you switch the TV on the relay kicks in and bypasses the NTC.

Maybe the relay contacts are bad and the TV has been running the main power supply through it, causing it to overheat and crack, then on the last power cycle it arced over no longer limiting inrush current and blowing the main fuse.

[drussell]

Let my main question be, "for those who have plenty of experience fixing TVs, would you recommend just swapping the fuse first?" My goals are primarily to limit any possible damage.

Absolutely not.  Your TV has a severe short somewhere.  Just replacing the fuse is likely to make things worse.  The short is direct enough to have blown up the inrush limiter (though they usually still pass current even with a chunk missing like that, just less than they're supposed to.)

I think I have identified two bad diodes. I'm getting a resistance reading in both directions, and 1.8v measuring with leads connected backwards with my meter in diode mode. I get .523v measuring with leads connected to the proper ends. Can someone please confirm my thinking that these seem bad? They can be properly tested on the board, right?
...
Thanks all. Good to know that the diodes can't be tested in circuit.

Uhhh, no... 

Diodes can often be tested in-circuit and what you have measured indicates properly working diodes, not defective ones.  In one direction, you should get the diode drop for that type of diode (about .5 volt for standard silicon,) which you are seeing, and a higher voltage in the opposite direction, which you are also seeing.  Those diodes are almost certainly fine.

Maybe the relay contacts are bad and the TV has been running the main power supply through it, causing it to overheat and crack, then on the last power cycle it arced over no longer limiting inrush current and blowing the main fuse.

That doesn't make sense because the OP stated that the TV was on for about 5 minutes before it blew up, tripping the upstream circuit breaker in the process.

It is almost certainly a blown transistor (or multiple transistors.)  They're usually easy to spot just by checking all the IGBTs with a multimeter since when they fail they tend to fail to a permanent hard short.  The OP should be able to find the short, either the PFC transistor(s) or main PSU switching transistor(s), as you stated.

Once the shorted components have been found and replaced, power it up ON A DIM BULB to be sure that they're not just going to blow again due to blown drivers, lack of oscillation in PWM controller, etc.  If the OP can't seem to find anything shorted, they could power up with a very low wattage bulb and see where the voltage stops but I can pretty much guarantee there is a shorted transistor (or perhaps diode but unlikely in this case with such a hard short) that can be found first, without attempting a power-up.

As for the inrush limiter, it is originally a WMEC WTR15D080, so 8 ohm cold, 5A nominal max current.  A Cantherm MF72-008D15 has the same specs but anything really close, like a GE CL-50 or CL-60 (7 ohm /10 ohm, 5A) or one of the Ametherm SLxx  series would work fine also if you can't easily find exactly 8 ohm cold in your locale.  They're typically only 20% tolerance anyway so not that critical as long as you pass enough current through it to warm it up properly in normal operation (so, for example, don't replace it with a 15A nominal model, etc.  It needs to hit thermal equilibrium at around the right point, though it is very "springy".)

WMEC ICL guide
Cantherm ICL guide
GE ICL guide
Ametherm ICL product datasheet matrix

[shakalnokturn]

Maybe the relay contacts are bad and the TV has been running the main power supply through it, causing it to overheat and crack, then on the last power cycle it arced over no longer limiting inrush current and blowing the main fuse.

That doesn't make sense because the OP stated that the TV was on for about 5 minutes before it blew up, tripping the upstream circuit breaker in the process.

True, I didn't read back far enough.

A dead power transistor makes more sense I guess, but then again if the TV had been on for 5 minutes when it crapped out, the relay should be bypassing the NTC in that state.
The current spike when the supposed transistor fails is enough to fuse the relay contact, then the NTC, then the fuse?

Another thought is that if the PSU design isn't failsafe enough and the relay contacts were bad, the voltage drop caused by the NTC when running the main PS could cause the PFC to duty-cycle itself high enough for self destruction?

In any case it is wise to check the relay...

[drussell]

the relay should be bypassing the NTC in that state.
The current spike when the supposed transistor fails is enough to fuse the relay contact, then the NTC, then the fuse?

Generally power supplies that use this kind of NTC rather than just a resistor for inrush limiting don't bypass the NTC in operation, they just let it self heat to the ~77 degrees C where the resistance is down its operating resistance (at about .2 ohms on this part's ratings) though I suppose you could make it a tiny fraction of a % more energy efficient by bypassing it and only running a relay coil but most manufacturers don't bother with the expense of a relay if they're splurging for an NTC IRC.  Does that cost more than a resistor and a relay?  Who knows.  Design choices. 

This TV might be bypassed but may well not be unless the IRC blew on a subsequent powerup attempt rather than the initial short.  I suspect it normally runs unbypassed, though a closer look at the circuit boards could probably tell us for sure. 

I don't have time right now to try to zoom the photos and trace it.

[shakalnokturn]

Solder side for relay and NTC outlined.

[csplinter]

Diodes can often be tested in-circuit and what you have measured indicates properly working diodes, not defective ones.  In one direction, you should get the diode drop for that type of diode (about .5 volt for standard silicon,) which you are seeing, and a higher voltage in the opposite direction, which you are also seeing.  Those diodes are almost certainly fine.

I'm getting more confused now. I had expected to see a reading from about .4v to .7v on the meter in diode mode, with negative to cathode and positive to anode. I also expected to see no reading with the negative to anode and positive to cathode. This is the expected reading that I learned from several youtube videos, and it's fits my intuition that a diode is supposed to work like a check valve allowing current one way but not the other. I'm a complete novice at this, but all these expectations were met when testing each and every diode except those two highlighted in my board images. I don't doubt what you're saying but I feel like I'm missing a big piece of the picture now.

[james_s]

The problem with testing diodes (or any part) in circuit is that the reading you get will be influenced by any other paths the current may take. Often you can test a part in circuit but you have to know how to interpret the result. A good diode will usually give you somewhere in the area of 0.3 to 0.6V in one direction and open circuit in the other. When they fail, they usually fail short circuit in which case you'll measure 0V drop in both directions, but if for example there's something else in parallel with the diode you could get a false reading. For example if the diode is across a transistor and the transistor fails shorted, testing the diode will test shorted even though it's actually ok.

Now if the fuse is blown and it also tripped the breaker that means there's obviously a short somewhere in the primary side of the power supply. I would start at the power input jack and work your way from there. Primary suspects are the bridge rectifier on the input and the chopper transistor(s) in the primary of the power supply. Note that if any of those parts are bad, there is a good chance the fault has taken out other parts.

[csplinter]

Thanks James_s. You are confirming my thinking and what I've been reading so far.

I was confused by drussel's statement that "In one direction, you should get the diode drop for that type of diode (about .5 volt for standard silicon,) which you are seeing, and a higher voltage in the opposite direction, which you are also seeing." This seems counter to everything I've been reading so far. However I've ruled out these questionable readings due to parallel components such as those you described. After all, these diodes were found at the far reaches of the board in what you're calling the secondary. I'm starting to better understand that this is basically two power supplies on one board, and I need to be looking further upstream.

As I said previously, I have checked two bridge rectifier chips and the tests (done in-circuit) came back negative for any fault. If parallelized components can cause a false positive when checking diodes on the board, can they also cause a false negative? I can't think of how they could at the moment.

My next task is testing the transistors, which I've put off for holidays and having been under the weather, but will hopefully get to desoldering them soon. 

[Paul Moir]

Parallel components *almost* always give only false positives,  falsely identified as bad.   A rare example would be two 100 ohm resistors in parallel to distribute heat; if one is open both will measure 100 ohms in circuit.

As James_s points out, you can quickly check semiconductors in circuit (without removing them) if you simply check to see if they're completely shorted.  While this doesn't prove a component is good, it can quickly identify ones that are bad.  And this is frequently how power semiconductor's die.  Even then you've got to trace out to see if they're influenced by a low DC resistance component like a transformer or inductor.
Note that a shorted transistor is often a symptom and not the cause of the failure though.

For what it's worth; I would also not be looking at the downstream components much at all yet.  There's not much that can happen past the isolation gaps that could cause the fuse to go and the thermistor to detonate.