Kogan KGNFHDLED40C PSU crapped out ...

[pedro]

Called out of retirement to look at a Kogan KGNFHDLED40C (rebadged chinese crap with Samsung panel). Died with a loud "crack" sound and smell of escaping magic smoke.

Power board shows a failed diode and a burst ceramic-body main fuse, the apparent source of the noise. Have posted in the schematics area with pic hoping someone can ID the board and maybe lead to a schematic.

Board labelled variously as "TE22 REV1.0 DX" and "39-TE22-PCA". Also one tag as ZD-95(G)F. Controller is Q0270R.

Panel FTW is Samsung LTA400HM08.

[R_J]

Re: Kogan KGNFHDLED40C PSU crapped out ...

Have you checked the other fuse? do you have any standby voltage? where is the failed diode located?

[pedro]

Re: Kogan KGNFHDLED40C PSU crapped out ...

The main fuse - T5AH/250V - is immediately after the input header. The other fuse - T2AL/250V - (looks to be between the rectified DC area and the main PSU transformer driver pair) is OK.

The diode actually measures OK. It has a small crater on the base between the legs which looks v.suss but meters good. FSU10B6, presumably in PFC duty.

As an aside, cap brands present are a mix - Chang/Chong?, CapXon and xunda. The HV ones are CapXon 47/450.

Have been hoping someone may have a light-bulb moment with the pic in the schematics section and say "that looks like the ..." and I may be able to track down a schematic. Always the optimist.

[pedro]

Re: Kogan KGNFHDLED40C PSU crapped out ...

Just powered the board up (no output connections) with a meter across the input fuse location. As the variac was wound up the current hit 2A at 160V input. 300VA into a board with no external load???

(Can't get the TV onto the space on the workbench and too hard to move the variac etc to the set.)

Didn't check for standby yet. Just to help, there are NO pinout markings at the output connectors (or at the other boards).

Link to pic: here

[R_J]

Re: Kogan KGNFHDLED40C PSU crapped out ...

You can use a incandescent (60-100w) lamp either is series or in place os the fuse to limit the current.
If F102 is ok, I suspect the standby circuit (small 8 pin dip near j11) is ok. you should have standby voltage across the 2 capacitors close to the diode mounted on HS4
Have you check all for diodes in the bridge rectifier, also check q102 fet and D101 diode.
There is what looks like a low value resistor (R105) in the primary, is it ok? if it is open it likely supplies voltage to a circuit that has a problem.

You could also post a clear high quality picture of the other side of the board (the picture you supplied is not that clear)

[pedro]

Re: Kogan KGNFHDLED40C PSU crapped out ...

You can use a incandescent (60-100w) lamp either is series or in place os the fuse to limit the current.

Normally do that, it was really a test to verify there was something faulty on that board - hence using the variac.

the picture you supplied is not that clear

Well it was only intended for recognition purposes, not diagnostic work ...

I'll hopefully get a bit of time later on to get back on it.

Tnx for replies, appreciated.

[R_J]

Re: Kogan KGNFHDLED40C PSU crapped out ...

I found some pictures of that board but no schematics, there are other models like TE32-21 which are close but I could not find a schematic.
I did find this one for a Leke that looks similar and has a schematic, it might help if some of the circuits are similar

[pedro]

Re: Kogan KGNFHDLED40C PSU crapped out ...

it might help if some of the circuits are similar

Quite similar in a lot of areas.

Update: Had a meter across the failed input fuse (F101, see attached partial schematic) to monitor current draw, but it was playing up. As previously mentioned, indicated 300W into the PSU only @160VAC with nothing getting that warm. Turned out to have an internal shunt failure - go figure, a $A400 DMM.

Anyway, jumpered the F101 and cranked up the variac, saw J2 go luminous and vaporise within 5 secs with still less than 50V in. OK, there's a short somewhere which explains the original fuse failure!

Replaced J2 (and the bridge in case it was faulty), short still evident. Lifted F102, short still evident. Lifted J1 and the short disappears. Note there is no other load connection to the VDC caps except via F102.

Now there are a zillion other parts of the circuit hanging off that link (i.e. VDC negative), too many to try tracing more. So it will prob be a shotgun approach to the extent of lifting various semis first to see what changes.

Meanwhile SWMBO is saying dump the whole thing and buy another. Not just yet.

[R_J]

Re: Kogan KGNFHDLED40C PSU crapped out ...

There are really 3 power supplies here, the small 8 pin dip near j11 should be the standby supply. I suspect it is protected by f102 so that won't be the problem.
The first supply after the bridge is the PFC supply, I suspect Q102 is for the pfc circuit, you should be able to remove it and check if the rest of the power supply works. without the pfc circuit working you should have about 340 volts on the main cap c104.
If it is still showing a short I would check the main supply which would be Q302
circuit. another way would be to remove a connection between the main capacitor and the Q302 circuit so it is not connected and that way you can check if the pfc circuit is working or causing the short, If its working you should get around 400 volts on the main caps.
If you can provide a clear high res picture of the bottom and top I may be able to help better.

[pedro]

Re: Kogan KGNFHDLED40C PSU crapped out ...

I'm a bit ahead of you already on that same track. Having SS pcb's and jumpers at least makes separation easier than DS pcb's.

Lifted J4 and the short disappeared. Pulled Q102 and reinstated J4, no short. Testing/replacing Q102 is next step, don't have a spare 2SK3455B or equiv lying around and it's Sunday. The question in my mind is whether having the PFC disabled thusly will allow otherwise normal function of the whole PSU (for testing ...).

Q102 is shorted. Now to dig up a replacement.

[R_J]

Re: Kogan KGNFHDLED40C PSU crapped out ...

The power supply should work but at a reduced output current, I would try and test it without the full load on the output. I suspect the output voltages should be close but might be reduced slightly.
Basically the pfc boosts the voltage from 340v to around 400v which is not that much of a difference so the full power supply should operate, but like I said likely not at full load.

Check the fet for shorts, if its shorted from drain to gate, the drive ic might have been damaged and installing a new fet would cause it to short again.

[pedro]

Re: Kogan KGNFHDLED40C PSU crapped out ...

The FET's short every combination

Just been looking at the internals of the driver (ST L6562D) but I'm battling to see how a shorted G-S(-D) FET could take out the output (unless the VDC pulled the source up that high, source resistor looks like 1.8 ohms). And I can't see a valid test I could perform with no FET present.

(My profile says ProTech but I retired from this stuff a few years back, and there's not much selection available. I'm a bit rusty, so bear with me).

[R_J]

Re: Kogan KGNFHDLED40C PSU crapped out ...

The switching fet Q102 will have 340 volts DC on its drain, If the drain shorts to the gate it will supply 340 volts dc to the ic pin 7 (GD) That poor little ic usually can't handle it. If the fet shorted drain to source before shorting to the gate, that may have protected the ic's output. But who can be sure.

In this case you could at least check the resistance of the ic's internals by checking for a short between pin 7 and pin 6 then check between 7 and pin 8. If there are no shorts, the ic might be ok.
You would need to scope the ic's output to see if there is a drive signal. other than that, leave the shorted Q102 out for now and try the power supply.
If you check the power supply without the mainboard connected, the power supply will likely not turn on as it will require power supply ON signal. You would only have the standby voltage (5 volts?) at this point

[pedro]

Re: Kogan KGNFHDLED40C PSU crapped out ...

OK had a chance to get back to this for a bit. Resistance readings on the PFC driver chip (L6562D):

+ to - to ohms diode test

. 8 7 H H
. 7 8 H L
. 7 6 H L
. 6 7 H L

(Oh for a font that respected white space.)

Vcc to gate drive OK showing the junction as expected. But those GD vs ground are worrying. Yes, 6->7 should show the protective diode, but 7->6?

Anyway plugged it back in and did a smoke test (no PFC FET fitted). No smoke, no (new) smells, nothing was even warm to touch.

VDC was 340-ish as expected. The pinout on CN102 seems to align with that schematic you found. 5VSB ok and +5V also. +12 came up on START but died after ~10 secs. That was repeatable.

I'm missing something.

[R_J]

Re: Kogan KGNFHDLED40C PSU crapped out ...

There is likely a power supply ON pin that would need to be turned on to have the +12 volts stay on, without the main board connected this signal will be mising, If you knew the right pin, you can turn on the power supply by connecting that pin with a 1K resitor to standby +5 volts, in some cases the power ON pin needs to be grounded to turn the power supply on.

If the pinout from the diagram is close, then use a 1K resistor between pin 5 (ON/OFF) and pin 6 (standby +5)
Does your power supply have +24 volts also?

It looks like the only problem might be that the PFC fet is shorted. Your tests of the ic don't show that IT is shorted, but it could still be bad. At this point I would try replacing the pfc fet and see what happens

[pedro]

Re: Kogan KGNFHDLED40C PSU crapped out ...

That smoke test was with everything connected up so load was there and started with the POWER ON/OFF button, only the PFC FET missing. Was looking for some aberration to give me more clues. The only thing unexpected was the +5V holding up but the +12 going off.

As for scoping the driver, it's a PITA for several reasons. The driver is on the copper side (obviously) so hard to get access while the board is fully connected. If only the GD pin required that can be done from front OK. Also my scopes don't like having their "cold side" much above ground, so VDC- would be asking too much of them.

Sourcing the FET ... None of the usual outlets have a match. (I've tried parametric searches at RS/e14/D-K/Mouser etc). Best is ex ebay from Italy or US, happy to do that except the delay. Driver chip I can source here in Oz (ten-pack! or singles+freight).

[R_J]

Re: Kogan KGNFHDLED40C PSU crapped out ...

The +5 volts standby is a seperate circuit only for standby, the 12 volts and (24 volts?) are from the other circuit, because the boosted voltage from the pfc is missing, the +12 etc might shut down under load.

[pedro]

Re: Kogan KGNFHDLED40C PSU crapped out ...

No have both +5 and +5SB holding up, only the +12 drops off. VDC holds at ~340/350

[R_J]

Re: Kogan KGNFHDLED40C PSU crapped out ...

I looked again at the schematic and +5 volts comes from the standby +5 volts. The +5 volts is switched on and off by QS1 (in the leke schematic) I suspect there is also a transistor to switch the +5 volts on your power supply also.
Remember that the +12 volt section of the power supply is seperate.

Are you sure there is only +12 volts and not also a +24 volts? The reason I ask about the 24 volts is if it has a problem it may be shutting down the 12 volts also

Can you provide a clear picture of the bottom of your power supply?

[pedro]

Re: Kogan KGNFHDLED40C PSU crapped out ...

What I have matches both CON4 and CON3 on that leke schematic. The top CON in that row on mine is not in use.

So from that it seems I do (should have) +24V on CON6. Will check that later on, if so then 12 & 24 off T1 (leke) would make sense, and so the IC driving T1 is shutting down, prob due to PFC link. Hmmmm.