Hi all,

I've lurked for a little while but am posting now because I could do with some help. Software and engine stuff I'm okay with - electronics not so much.

I was just in the middle of watching 'Whiplash' the other night when there was a decent crack/pop sound and the TV set went off - no sound, no picture, no standby light.

- I unplugged it and checked the 13A fuse in the plug with a meter (Fluke 233) which was fine.
- Took the rear cover off and had a look and a sniff. No obvious damage.
- Had a look online for the schematic (attached) to see where to begin. Found that the fuse (3.15A slow blow) immediately after the AC in had gone (F101 on the schematic, part number T3.15A H 250V 50CT VIOLET(1-LINE)).
- Stopped at this point as I didn't really know how to find out what caused the fuse to blow. A friend who knows more than me suggested it could have been just a surge so I replaced the fuse with a new one and (as expected) it popped again.
- This time, because the cover was off, I noticed a short at the MOSFET on the heat sink (Q101 on the schematic, part number mdf10n65b 650v 10a to-220fp). Sure enough there looks like a scorch mark across two of the pins. These might have been there before or I may have caused more damage by my above actions - I'm not sure.

I have scoured many forums and watched many YouTube videos to try and learn the steps I need to take to (1) narrow down the root cause of the problem and (2) eventually fix it but haven't been able to figure out the next step. One thread I read suggested de-soldering and testing all the components but common sense suggests there are better methods. Another mentioned looking for a repair kit specific to my board and just checking the components which (apparently) commonly fail. This sounds better but I couldn't find a repair kit for my board.

I know I could just buy a new power board and replace the whole thing and (hopefully) this would resolve my issue - unless the issue was somewhere else - but I thought it might be educational and fun to have a go at fixing it. Also, while we don't have a TV, the wife and I are actually having conversations! I've bought a cheap component tester (https://www.ebay.co.uk/itm/113195528307) which will come in a few days so I think I have most of the kit I need (soldering iron, basic tools, meter, etc., I already own) but I honestly lack the know-how.

If anyone is interested in helping or guiding me along to which components are likely to have failed or what to test next (and how to test it) I would be grateful.

Cheers,
Arram.

Other info - TV model LG 42LM3450, power board LGP42P-12LPB, other details in attached spec.

 

It is very difficult to fault find a circuit where the fault causes the input fuse to operate on power-up; since you cannot probe the circuit to measure voltages are various circuit points.

That said, you know that the input fuse is operating due to excessive current and based on your observations with the cover off – favourite would be the fet Q1 having failed short circuit (drain to source). Since you have a meter, measure the resistance value at Q1 (drain to source) – you might need to trace out which pin is connected to other components in order to identify the gate/drain/source pins (note that a zener diode and resistor are connected gate to source).

If the resistance drain to source (in both polarities) is low, the transistor is toast. Of course it is possible that the reason it is toast is that IC101 is faulty, permanently switching on the transistor – but you won’t know this until you replace Q1.

I have one of the ebay component testers – I paid a bit more for a boxed version, it is very impressive for the price. It has the ability to identify almost any 2 or 3 pin component and give component parameters that you would expect to pay much more for an instrument to determine.

 

As you say that you have no electronics knowledge a switch mode power supply is not the place to start learning. I think the fault is most likely to be on the part of the circuit before the high frequency transformer. The whole of that area is at mains potential so You can't connect an oscilloscope to that part of the circuit unless you power it from the floating output of an isolation transformer. There is also the problem that may component failures cause other components to fail and unless you replace all the failed components it will fail again. I think you should start learning by working on battery powered devices at first.

Les.

 

If transistor Q1 is short circuit (drain to source) the primary switch mode transformer winding will provide a low impedance path for the fault current to flow. If Q1 is faulty it is possible that the fault current has taken out the bridge rectifier BD101.

If Q1 measures low resistance (D-S), remove it from circuit and power the board – the input fuse should not operate.

 

Thanks for taking the time to get back to me, both of you.

Les, you might be right about me not having enough experience to be working on a switch mode power supply. I think I know enough not to injure myself and it wouldn't be the end of the world if I had to buy a new TV. I'd like to try some of the more simple options available (if any) and if it becomes too complicated for me to get my head around I'll just give up and buy a replacement board from China.

Hymie, thanks for your input. I've been playing football this evening and I'm at two weddings over the next two days but I'll try and spend some time on it over the weekend! I imagine I'm not going to be able to get replacement parts straight away but there's an RS shop down the road from me so I might be lucky. At the least I should be able to measure the transistor and confirm if it's toast. It may well be that there are multiple components that are dead but I won't know until I check.

I'll post an update at the end of the weekend.

Thanks again.

 

If transistor Q1 is short circuit (drain to source) the primary switch mode transformer winding will provide a low impedance path for the fault current to flow. If Q1 is faulty it is possible that the fault current has taken out the bridge rectifier BD101.

If Q1 measures low resistance (D-S), remove it from circuit and power the board – the input fuse should not operate.

I took out the transistor, put in a new fuse and connected the 240V. You were right, the fuse didn't blow.

When I measured the transistor D-S I got open loop. But when I switched my leads I got 2.76 mega ohms.

I didn't really get time to do any further research/learning this weekend but I will one evening next week when I'm not working.

 

Clearly the fault current path is through the fet Q101 (once removed the input fuse does not operate), but when removed measures high impedance in both polarities suggests that IC101 is at fault (not switching the Q101 on/off as required).

Once you get your component tester, check the fet with the device. It might be that when replacing the fuse for the first time, the fault current blew the fuse and at the same time blew transistor Q101 open circuit.

If the fet is OK, you could replace it and open circuit the path of resistor R103/zener diode DZ105 (which supplies the initial power for the switcher IC101).

If the fuse does not operate, remake the path of the resistor/zener diode – if the fuse now operates, IC101 is probably at fault.

You should minimise the number of times your input fuse operates, which could destroy the bridge rectifier (and other components). When testing the board, it might be worth using a 1.6A fuse which should not operate.

Of course it is possible that a secondary circuit fault is causing the fuse to operate, if it is possible to disconnect the output of the PSU from the TV, that might be worth a try – with an output overload causing the fuse to operate.

 

Got one of these LG PSU boards in front of me (LGP42P-12LBP). Failure mode same as your's Arram - F101 blown, but there's also a 330pF 1kV capacitor (C102) across Q101 FET that's got a big black hole (soot) in one side. Suggests to me that your capacitor might also be of concern. I'll be replacing this one with a 2kV component.

 

Got one of these LG PSU boards in front of me (LGP42P-12LBP). Failure mode same as your's Arram - F101 blown, but there's also a 330pF 1kV capacitor (C102) across Q101 FET that's got a big black hole (soot) in one side. Suggests to me that your capacitor might also be of concern. I'll be replacing this one with a 2kV component.

I've no idea how I've missed this but now I've looked closer there is a small scorched hole in C102 on my board also. I think I'm going to get myself a decent lamp and some new glasses...

 

Attached are pictures of the damaged cap, fet and component tester result. The threshold voltage seems normal against the datasheet.

Do you think it's worth replacing the two damaged components? Or continuing to fault find?

 

Failure of the capacitor short circuit would have caused the input fuse to operate.
I would take Jathome’s advice and replace this with one rated at 2kV.

The fet may be OK – but if the capacitor was in circuit with the fet removed (which resulted in the fuse being OK), then the fet is likely faulty.

I would replace these two components and power up with a 1.6A fuse.

 

This is the closest I could find. I think it should be okay... https://uk.farnell.com/stmicroelectronics/stf10n65k3/mosfet-n-ch-650v-10a-to-220fp/dp/2098194

And for the cap... https://uk.farnell.com/vishay/f331k29y5rp6uk5r/cap-cer-y5r-330pf-2kv-rad/dp/1141800

Yay or nay?

 

This is the closest I could find. I think it should be okay... https://uk.farnell.com/stmicroelectronics/stf10n65k3/mosfet-n-ch-650v-10a-to-220fp/dp/2098194

And for the cap... https://uk.farnell.com/vishay/f331k29y5rp6uk5r/cap-cer-y5r-330pf-2kv-rad/dp/1141800

Yay or nay?

Quick update on the LG PSU I was looking at - changed the fuse and the 330pF (C102, but now with 2kV type) and put it back into the TV this afternoon (i.e. didn't change the Q101 FET) and it's all working perfectly once again. Your capacitor looks good Arram - sorry haven't checked your FET yet (only just seen your recent post).

 

This is the closest I could find. I think it should be okay... https://uk.farnell.com/stmicroelectronics/stf10n65k3/mosfet-n-ch-650v-10a-to-220fp/dp/2098194

And for the cap... https://uk.farnell.com/vishay/f331k29y5rp6uk5r/cap-cer-y5r-330pf-2kv-rad/dp/1141800

Yay or nay?

N channel MOSFET looks from ST looks similar (voltage, continuous current, pinout), but the peak power spec is higher for the MagnaChip part (47.7W compared with 35W, even though pulse current specs are same at 40A). My guess is that it'll work fine, but would be tempted to try the original if it isn't fried.

 

All sorted. Thanks chaps. Seems like it was the same issue as Jathome's as replacing those components did the trick. I did change the FET since it was visually damaged (and I just didn't trust it!).

Appreciate the help!