Is this monitor PSU fried? (Acer XB271HU, possible lightning damage)

[exscape]

My Acer XB271HU (bmiprz) 27" 1440p IPS monitor failed to start yesterday morning, quite possibly due to the lightning storm the previous night. 
It does seem somewhat unlikely to be caused by the storm as I was using it without issue until about 11pm, then powered down and pulled everything out of the wall when the storm got closer. I saw it work as usual maybe 15 seconds before pulling the cord. On the other hand, it's a hell of a coincidence that it's worked perfectly for two years and breaks when powered down for the night during a lightning storm!

Symptoms are that the power LED blinks blue, which AFAIK it should never do -- power save mode is amber. The soft power button doesn't do anything, so I can't turn it off. There's no image of any kind whatsoever, and the backlight seems to be always off, though I'm only 90% sure of that. 
A friend with a very similar monitor (XB271HUA) says his turns on to a manufacturer logo, then says "no signal" and the LED goes amber, if nothing is plugged in. Mine simply blinks the power LED in blue and that's it. The soft power button on the front also does nothing, so I can't turn it off that way.

As I have no way to get it repaired (for a reasonable price), I figured I'd take a crack at it as I have some experience with electronics repair. There are also power supply PCBs for this exact model available from e.g. eBay, to fully replace the power supply.

Board images:
https://imgur.com/a/jT6rwA7 -- I recommend right clicking the images and opening in a new tab to see the full size.
Especially note the discoloration/bad solder joint at Q703, and the resistor next to it (R720).

The output connector (shown in the capacitor image) does produce 12.1 V between the magenta and black wired while powered; the other colored wires appear to be at 0 V relative to black.
There's also a big capacitor near what seems to be the output side (see images in the "bonus question" below), that was changed to something very high, IIRC 380 V and drained very slowly (1.5 hours or so until it was at a safe voltage of 15 V or so).

The main question is really this:
Based on the images and the fact that there *is* output voltage, do you think this PCB could be the only reason the monitor doesn't work, meaning it's relatively likely replacing it will make the monitor fully usable again? 

Bonus question:
Is the weird discoloration around the transistor (on both sides of the board) likely to be a sign that it's dead (and likely by overvoltage)? 
There is no such discoloration in the eBay images I found of this power supply: 
https://i.ebayimg.com/images/g/URYAAOSwsCliqHPM/s-l1600.jpg
https://i.ebayimg.com/images/g/18YAAOSwPIdh7Wue/s-l1600.jpg
https://i.ebayimg.com/images/g/-2sAAOSwBGRiqHPM/s-l1600.jpg

If so, I'll most likely buy a replacement power supply PCB. I figure component-level repair is probably more trouble than it's worth, especially considering the risk that many components are damaged despite looking fine.

[thm_w]

Do you have a soldering iron to resolder Q703?

Those 35V caps look a bit bulged as well. Anything you can test them with?

[exscape]

Do you have a soldering iron to resolder Q703?

Absolutely, plus a hot air station. Do you think it's still whole, though? I figured it's likely to be broken if the discoloration is from excessive heat from a voltage spike/lightning strike.

Those 35V caps look a bit bulged as well. Anything you can test them with?

Not sure; ESR meter? I don't have one of those. I have a decent multimeter and a DSO (Rigol DS1102E) though.

[thm_w]

The discoloration would unlikely to be from lightning strike, that looks like heat from long term use. Maybe when you put it away it shook enough to break the solder joint.

Does the multimeter have a capacitance test? Its not as good as ESR meter but would be a quick check.

[james_s]

It's possible it was damaged by lightning, but if it was, it probably isn't any of the visual issues you can see. Lighting damage that is severe enough to be visible usually manifests itself as traces vaporized off of the board, ICs with craters in them or parts that have completely exploded.

[cruff]

Lightning damage can also be more subtle depending on the distance from the strike. About a year ago there was a strike about 100 meters away from my house that induced quite a surge in the 7200 V distribution lines running adjacent to my house, which in turn induced surges in the Ethernet cables that parallel those lines inside my house (maybe 7-8 meters). Had various electronics damaged, many with no obvious destroyed components. Some continued to work that day but died the next morning when I cycled the power on them. One of the 8 port Ethernet switches, which mostly survived, does have evidence of the discrete protection diodes having been overloaded and there is some solder vaporized around them. Had one computer loose its Ethernet port but it remained otherwise functional, a second computer which was powered off required both the motherboard and graphics card replaced.

[mariush]

No matter if it looks good or not, do the basic tests like checking diodes with a multimeter to make sure they're not open or conducting in both directions.

The Q703 part ... that's long term heating of the PCB, it's not caused by lightning. You need to clean the pads with a soldering iron and apply new solder. The R720 is also producing a lot of heat, hence why it's positioned so high on its leads, to have airflow around it and not heat up the pcb as much... but it still conducts heat into the PCB.  The solder joints look bad, so I'd also remove the solder and apply new solder.
I'd replace at least the capacitor closest to Q703 even though it doesn't look swollen - it has been subjected for a long time to the heat from Q703 and R720 and also from the mosfet or transistor on that heatsink below it.  It's hurting.

And you'll also need to inspect all the solder joints of components around that part, but it doesn't take a lot of time to just resolder all components around that part.

The capacitors near that connector are degraded, they do seem swollen, those should be replaced.

You can discharge the big capacitor after tests by applying a resistor across its terminals ... try 470 ohm - 1 - 2 kOhm for 10-20 seconds across the leads. But don't use your fingers, hold the resistor by its body with some pliers or something.

[Haenk]

Q703 has desoldered itself, it seems. Judging by other solder joints, it might have been crappy soldering in the first place, or it became too hot (not meeting current demand). I would exchange it, put a little heatsink on it (to be safe).
R should be fine, the caps are also not too close to the transistor, they should be fine (thermally). However they do look bulged and have probably (very likely) failed or are about to fail. Check other caps as well.
You should check all soldering joints close up. E.g. JW15 (bottom) and C708 (bottom) don't look too great. Might be just a viewing angle issue, but I'd just resolder odd looking joints.

[exscape]

Thanks everyone!
I've gotten started. I think the transistor is dead -- my transistor knowledge is extremely rusty but these values don't make sense, right?

I'm using the multimeter in diode mode and checking various combinations.
The results are:

+ on base, - on emitter: OL
+ on base, - on collector: 0.72 V
- on base, + on emitter: OL
- on base, + on collector: 1.60 V

Assuming it's dead, any ideas how to find a replacement?
It's marked K946 / C200 / Y (three lines).

The bulging caps are probably also dead, as the multimeter jumps between 15 uF and OL 1-2 times a second. I've only tested one so far though.
The cap closest to the transistor is likely fine as it reads as almost exactly 330 uF and looks fine.

Edit: All but one of the bulging caps give crap readings. I'll replace all of them anyway of course, assuming I can find a transistor replacement as well.

[mariush]

It's probably 2SK946 ... but I'd suggest desoldering it and testing it again out of the pcb.

Go on Youtube and look for mosfet testing using multimeter, to see how to properly test mosfets.

[wasedadoc]

It is not unusual for a swiched mode power supply that has been slowly dying to keep working while continuously powered and then fail when repowered after a break. There is a good chance that the lightning had no direct effect, only that it prompted the repowering.

Replace the capacitors on the ouput side and likely all will be well again. The high voltage capacitor on the output of the mains bridge rectifier usually last much longer than those on the outputs of the low voltage rectifiers.

[exscape]

It's probably 2SK946 ... but I'd suggest desoldering it and testing it again out of the pcb.

Go on Youtube and look for mosfet testing using multimeter, to see how to properly test mosfets.

I did test it out of circuit.
The pins on the PCB are labelled as (left to right, with the flat side of the package towards me) E C B, so I assume it's not a MOSFET?

[Haenk]

The casing does not match for 2SK946; try 2SC200 instead.

[strawberry]

PCB is cheap and worn and lightning is just coincidence
Q703 turns on Main supply, so something is wrong with it. maybe cheap dry capacitors overloaded it

[exscape]

I did find a service manual (not for free, but for cheap) but the power board schematic/layout is incorrect; there are major differences (or perhaps rather: there are some similarities), so that turned out pretty useless. The transistor in question doesn't even seem to exist there.

PCB is cheap and worn and lightning is just coincidence
Q703 turns on Main supply, so something is wrong with it. maybe cheap dry capacitors overloaded it

Hmm, what do you mean by "turns on Main supply"? Would the power LED light up even if that didn't work at all?
The puzzling bit is the 12.1 V on the output. Though I have not checked with a scope, nor have I tried to load the supply at all. Nor do I know what the correct output voltage is; in the service manual it's 19 V, but as mentioned that layout is way wrong, so that doesn't really prove anything.

Any advice on how to find a replacement transistor? Even assuming 2SC200 is correct there are still several choices (like 2SC2001, 2SC2002 and others).
It would be very nice to have something available here as that would make shipping FAR cheaper and easier (like $3 vs $50) than most alternatives.
I have just about no idea what kind of max voltages, current gain or anything else I need.
Based on the previous comments I assume it's a NPN, but I don't really feel I know that either.

[strawberry]

KTC200
Y:hfe 120..240

Hmm, what do you mean by "turns on Main supply"? Would the power LED light up even if that didn't work at all?

Mistake with PFC inductor

[exscape]

KTC200
Y:hfe 120..240

Hm, can't really find that one to buy either.

I took a few measurements at the transistor pads when it's out of circuit and found these voltages:
Base-emitter 16 V
Base-collector -25 V
Collector-emitter 41 V
(To be clear, these have the positive probe on the first pad listed, and the negative on the second.)

Not sure how to reason about those as I can't visualize the circuit whatsoever without a circuit diagram. But I assume the transistor needs to tolerate those voltages and a bit higher, right?

[strawberry]

schematic

[Haenk]

https://www.el-component.com/bipolar-transistors/2sc200

The KTC200 is the same as 2SC200, just in case it's confusing. The hfe comes in different classes, "Y" is required.

Available e.g. on Aliexpress:

https://de.aliexpress.com/item/4000337343353.html

or ebay USA:

https://www.ebay.de/itm/290680296806

I think there is likely only very little risk of getting fake ones, so those Aliexpress parts are probably fine.

[exscape]

https://www.el-component.com/bipolar-transistors/2sc200

The KTC200 is the same as 2SC200, just in case it's confusing. The hfe comes in different classes, "Y" is required.

Available e.g. on Aliexpress:

https://de.aliexpress.com/item/4000337343353.html

or ebay USA:

https://www.ebay.de/itm/290680296806

I think there is likely only very little risk of getting fake ones, so those Aliexpress parts are probably fine.

Ohhh, I didn't at all realize they were the same! Thanks for pointing that out, I focused more on the prefix than the common suffix of C200.

Unfortunately AliExpress is too slow (they estimate a bit into October) and eBay also slow (3+ weeks) and expensive at $25 or so shipping. I realize I'm being a pain here though and I apologize for that.
I'm trying to find this particular transistor OR one similar enough (using your first link as a base), but can't find ANY of those listed as compatible -- and even then, I expect that not all of them will have all the parameters required...

From one source, the only ones that seem plausible (after a filtering using the page, then checking the hFE in the datasheets) are (links are to datasheets): MPSA06, MPSA05RA, ZTX455.
All of them have Ic >= 500 mA and higher max voltages than the C200 transistors.
hFE is listed as 100+, 100+ and 100-300 respectively.
Could one of those work instead? It would save both weeks and/or a decent bit of money.

From the other source most are only sold in bulk, but one promising is BC63916-D74Z. hFE = 100-250 at 150 mA, not sure which Ic I should really look at in the datasheets though. The hFE is only listed as 25+ for other test currents.

I should point out that I haven't checked the pinouts of all of these, but presumably that's solvable by some twisting if necessary?

[exscape]

Uhhhh.... I checked the output with the oscilloscope now as I figured it might look bad -- and it looks... far worse than I expected.

In case the image doesn't work: there's a TON of ringing with Vmax = 52.8 V, Vmin = -39.2 V (Vpp 92 V). The average and RMS are around 11-13 V, which is what the multimeter shows...

Is there even any point in trying to repair this? Could the capacitors and everything else in the supply have survived this?
To be clear, this is in the current state, without the transistor. Which might have been a very bad idea, but the average output voltage was the same WITH the transistor as it is now, so I would guess that the waveform was the same before I touched the supply PCB at all.

There's also the question if the output looked like this while it was connected to the rest of the monitor, in which case that too might have taken damage.

[strawberry]

what output.. sec.winding.. linreg.. rectifier
broken rectifier would lead to MOSFET failure due to capacitance loading

[exscape]

what output.. sec.winding.. linreg.. rectifier
broken rectifier would lead to MOSFET failure due to capacitance loading

I measured at the connector that goes to the rest of the monitor, so the output of the power supply. 12 V, 19 V or whatever it's actually supposed to be. I figured 12 since I measured 12.1 V with the multimeter, but I now realize that was probably entirely meaningless.
The ground lead was on a heatsink that had <0.1 ohms to the connector's ground pins, because it was so much easier and safer to place it there. It doesn't stick to the ground pins of the connector.

[strawberry]

R709 38,3k
R710 10,0k
19V

[Rasz]

Uhhhh.... I checked the output with the oscilloscope now as I figured it might look bad -- and it looks... far worse than I expected.

In case the image doesn't work: there's a TON of ringing with Vmax = 52.8 V, Vmin = -39.2 V (Vpp 92 V). The average and RMS are around 11-13 V, which is what the multimeter shows...

dried output caps = this ringing