LG TV wipeout

[Skepticist]

Nice soldering iron mod there
I used thinner copper for mine which probably explains why the heat gets sucked out of it quickly. Got a couple of dirt cheap 40W soldering irons with tips that have a 1/8" rod inside the heating element, silver soldered copper tool to the end of 1/8" brass rod to make some SMD tools - mainly for removing resistors, caps and leds. Tried with 8pin and larger ICs with mixed success so might go for thicker copper next time.

The pin 11 voltage variation I'd still put down to the variable frequency operation of the IC and multimeter characteristics as the Vgs max in the specs is 25V with a threshold of 3V and typical operating for switching around 10V or so (Vcc?). You'd need a CRO to see how big the gate pulses actually are.

[loopyloo]

The pin 11 voltage variation I'd still put down to the variable frequency operation of the IC and multimeter characteristics as the Vgs max in the specs is 25V with a threshold of 3V and typical operating for switching around 10V or so (Vcc?). You'd need a CRO to see how big the gate pulses actually are.

Pin11 measurement.
Sorry about the bad photo. It was a little tricky holding a loaded probe on a tiny SMD pin with one hand and operating a camera with the other.

[Skepticist]

That's what I was expecting, short pulses a long way apart giving a very low average DC voltage reading. The IC might be crippled but it's still trying to get running.

[loopyloo]

I wonder what frequency this thing is supposed to run at.
I will have to be at some minimum rate for the caps to do their job.

[loopyloo]

Just made a startling discovery, well to me it was.
Just looking over the circuit and noticed R621 is also a 2.2ohm resistor. Could it be that 2 low ohm resistors (that can allow a larger current through) be destroyed from that lightning hit ?
Yes they can.
I'll replace that one now too but I'm not sure I've got any left.

PS Found one !

OK so what's the tally now ... ?
4 FETs
3 Caps
2 Fuses
2 Resistors
1 Zener diode
and a partridge in a pear tree.

For those partaking in the almost 9,000 views, stick around, more to come !

[Skepticist]

It regulates by varying the frequency instead of the usual PWM operation
There are some formulas in the datasheet that determine min and max frequency limits. Light load results in max frequency and Rfmin & CF are critical to that control and the current out of RFmin is what determines actual switching frequency (highest current for high frequency and vice versa). The variable current is courtesy of the opto IC102 so high secondary voltage raises the switching frequency decreasing the efficiency of the transformer T101 to stabilise the voltage. RFmin voltage must always stay above 1.3V or normal operation can't happen is way I read it.

[Skepticist]

Would be a hell of an inrush to cause those resistors go open cct
There's another one in the MICOM power cct R702

[loopyloo]

Would be a hell of an inrush to cause those resistors go open cct
There's another one in the MICOM power cct R702

Yes a hell of an inrush but incredibly brief.
R702 is ok.
I have a feeling this set was hit when it was powered off.
I think I might connect it back into the TV and try it out again. (looking for a miracle)

[Skepticist]

I'd be hesitant about connecting the TV until there are some sensible looking figures from the power supply
Miracles have been known to happen but this needs some major divine intervention for instant recovery atm

[loopyloo]

Yeah it didn't change anything.
I guess all that's left is to try the new chip.
Oh well, I'll have to put it in my prayers tonight.

[loopyloo]

Look what I found under that chip !

[Skepticist]

A pair of shorts no less :O
what was shorted to where?
and that pad 8 isn't looking too healthy or is that just the reflection?

This could be the turning point

The tradition is the last bug doesn't show itself until 4am (usually)

[loopyloo]

A pair of shorts no less :O
what was shorted to where?
and that pad 8 isn't looking too healthy or is that just the reflection?

This could be the turning point

The tradition is the last bug doesn't show itself until 4am (usually)

Lol no that's not the board, just an image from googly. No shorts underneath.

Just looking around at the moment. I put the new chip in but still the same.
Maybe I should start cutting tracks. There is a link between R114 and ZD102. I could lift that to see if the thief is below.

[Skepticist]

Had me celebrating too :|
If the problem was south of R114 you'd see most of the RFmin voltage dropped across the resistor, if it's 8.2k the max current through there would be 1/4 mA
The problem has to be closer to the chip (lightning delivery too if that's the second replacement already)
They're cheaper from China but the wait is a bit longer

[Skepticist]

Maybe try replacing C101 (0.47uF 16V) and C104 (470pF 50V) - do C104 first
There's Q103 that has an RC trigger (R107+C102) to discharge C101 if the frequency goes too low so check them closely, make certain D102 is ok with junction test to make sure it's not broken down
Pretty much have to try everything that could have suffered from a spike on the primary side at this stage
The voltage on LINE pin7 measured ok earlier but I'd also try a replacement for C105 (garden variety 100nF 50V), even have a go at C106 (0.22uF 16V)

Even if it's only tacking a temporary substitute

[loopyloo]

About all I can do I think is tack over the top cos this is all smd stuff. All I've got in caps is ceramic of old boards, should have some new 1n4148's around somewhere.
Trouble is it's pretty hard to remove smd caps and resistors the first time without wrecking them cos of that glue they use for flow soldering.

[Skepticist]

Been through that too - made a 'U' shaped soldering iron tips that do both ends of the components at once and a hook to pull them off the pads but that glue never makes it easy.
Diodes and resistors can be tested in-circuit but a leaky cap is awkward and any of those caps directly connected to the chip could be the culprit (never any guarantees though), soldering a cap in parallel is unlikely to help.

[loopyloo]

Been through that too - made a 'U' shaped soldering iron tips that do both ends of the components at once and a hook to pull them off the pads but that glue never makes it easy.
Diodes and resistors can be tested in-circuit but a leaky cap is awkward and any of those caps directly connected to the chip could be the culprit (never any guarantees though), soldering a cap in parallel is unlikely to help.

Yeah I know I'll have to remove the caps. Funny thing is smd stuff looks tiny on the board, but when they're removed and not having that solder blob on each end they become almost invisible.
Now I'll have to put 3 pairs of glasses on.

Buying them is a pain too. Can't usually buy just a few, it's gotta be like 200 or 500 when you only want one so it gets expensive.

[Skepticist]

Yes been through all that too - drop one and you'll never find again.
I use the cheap reading glasses from the discount store for this, highest power they have like 3x or better - the main danger with them is you get so close to the board you could end up with the soldering iron on your nose
Need to know what physical size you need (0603,0805 etc) and Element14 can supply cut tape and most are around $1 for 50 of 1 value which isn't too bad - the shipping will cost more than that

[tw2005]

Yes been through all that too - drop one and you'll never find again.
I use the cheap reading glasses from the discount store for this, highest power they have like 3x or better - the main danger with them is you get so close to the board you could end up with the soldering iron on your nose
Need to know what physical size you need (0603,0805 etc) and Element14 can supply cut tape and most are around $1 for 50 of 1 value which isn't too bad - the shipping will cost more than that

That's where RS is good. Just a lot of that stuff comes from the UK but if it's local, next day and free