Philips PSU (715G7350-P01-000-002S) behaving funny

[hervou]

Hi all,

I'm becoming crazy trying to fix the Philips TV of my old man (model 43PUS6201/12 - the TV, not my dad). I've read again and again discussions about this and similar TV models, on badcaps (thanks so much !). I see no other option than asking for your help.

The symptom is simple : when I plug the power cord, the red "stand-by" LED does not light on anymore, and the TV powers-on neither with the remote control nor with the rear button. I guess at this point you immediately think at the 5V (or other voltage depending on the hardware) stand-by output...

So, I disconnect the suspicious PSU from the rest of the system, I feed the board with 220V AC and measure DC outputs on CN9101. Notice here that I DON'T shortcut the 3.5/5V standby output with PS_ON in the first place, neither with a wire nor a 1K resistor. Therefore I'm expecting to see only this standby voltage and none of the 12V. Curiously I see the exact opposite behavior :
- 12 volts on 12VA/16VA
- 12 volts on 12V/16V
- 0 volt on +3.5/+5V

Well, I guess I must have misunderstood something. Anyway, measuring 0 volts on an output that should obviously deliver 3.5, 3.7 or 5 volts (*) leads me to think that the PSU is faulty.

(*) : This detail is documented on page 34 of the Chassis service manual kindly provided by elektrotanya.com. I feel the doc is a bit unclear which behavior to expect with each of PSU variants 715G7350, 715G6973 and 715G6887. Not important actually.

Then I chase leaking caps, smoked ICs and all, nothing obvious. Finally I play it humble and buy a replacement board at AliExpress (sorry...). But... I see the exact same behavior oin the new board (looking like a worked-on piece). I ask for refund, buy another one, same thing again. At this point I think that the main board might be damaged and shortcut this standby output voltage, and each time kill the PSU on first power-on. So, I should test output voltages before connecting it to the main board. I ask for refund again and buy a PSU. This time I test it carefullly. And... Same behavior again !

I see only two possibilities :
- I'm dumb and I've misunderstood something around how these outputs are supposed to behave
- my karma is aweful, I probably was a serial killer in a previous life or something. And all these PSU I receive have a similar issue with this 3.5/5V.

Then I've also measured voltages with PSU and mainboard connected :
- I read a quite stable 0.8 volts on 3.5/5V
- and a stable 0.56 volts on PS_ON
- of course, 12V and 12VA outputs stay at 12 volts

Not sure whether these measurements are meaningful, though.

Thank you for reading, helping, advising and chiding

Hervé

[Diah]

shows us please yours equipment that you used by measure volts..
and check this ceramic cap if its broken from leg or had any sign of broken at surface

[hervou]

Thank you for your help Diah.

I've got only a low-end digital multimeter, brand "Lifedom" (photo attached). The battery is new and I use the same caliber "20V" to measure the 12V and 3.5/5V, so I guess there is no doubt on this particular caliber.

Note : although I've had electronics courses I have no skill in understanding a power supply schematic or in cleanly unsoldering and testing components, that's why I have not tried to components (for example those contained in the famous "repair kit").

Capacitor C9149 looks sane on the aliexpress.

Out of curiosity I've had a look at the original board, and this capacitor was not soldered. I've noticed a couple of differences between both boards :
- original board had no C9149, no FB9904 and no C9922
- aliexpress board has neither L9301 nor C9309. And two more components are missing on the rear side : Q9107 and U9901.
I have no clue what these differences imply !

Cheers,
Hervé

[Diah]

put the original PSU alone on table and plug it with main cord... just tell what do you have on this points black probe as yours DMM and red with V DC.

[hervou]

Breaking news.

Preamble : I remember having tried to understand the schematic to identify the possible faulty component. At some point I must have considered that the smart next step was to unsolder one leg of D9105 to test it. But as I've got neither a desoldering braid nor a powerful enough iron (and no experience in that field too) I've damaged the soldering point on the board (and I would probably have burned the diode if it wasn't already). See the piece of art attached. Now as you might see on the picture, in its "natural position" the cathode is unsoldered.

Back to the present : in these conditions I check the voltage from the rear side of the PCB as you've asked. And I read 3.7 volts (!!!). Which I'm pretty sure I've never seen before.

So, I plug back the main board to the PSU to see if by chance the stand-by LED goes to life. And... It does. Incredible. I quickly unplug the system (this diode obviously has a function, so having it removed can do some damage...).

New test : I unplug the main board from the PSU and test again the voltage you asked while pressing on D9105 with some gear to have the cathode in contact. And when in contact the 3.7 volts disappears, I see 0 volt. I remove pressure on the diode, voltage comes back.

Does this make sense ?

Is it possible that the D9105 was a good guess and although I wasn't able to unsolder/test/solder by chance we are now certain it is the faulty component ?

By the way this still doesn't make clear to me why the 3 chinese boards (!) have no stand-by voltage...

[Diah]

take out the regulator next to diode and check it in diode mode out side the circuit.

[hervou]

You mean Q9106, which is a K22A10N1 mosfet (photo attached) ?

I've soldered out (easier than the CMS diode) and tested as requested.
I see no short between any of the gate, the source and the drain. Infinite resistance everywhere, even with the highest caliber on the multimeter.
If I should order some units of this shottky, feel free to advise a brand / model. Then I'll just have to find a friend with the appropriate equiment (or try with my 30W iron).

Thank you Diah !

[Diah]

measure it again to be sure i diode mode between source & drain.. it should read diode in one direction

[hervou]

[post edited, I've just got that a multimeter in diode mode displays volts and not ohms haha]

Ho yeah, right. I was in diode mode, but I didn't test both directions. Testing again I find infinite resistance everywhere except :
- red probe on source, black on drain : 0.9 volts drop voltage
- red probe on source, black on gate : 0.6 volts drop voltage

I've tested the nearby half-unsoldered shottky D9105 too, it shows 0 volts / 0 ohms in both directions. I guess there is no doubt that this one is dead.

[Diah]

[post edited, I've just got that a multimeter in diode mode displays volts and not ohms haha]

Ho yeah, right. I was in diode mode, but I didn't test both directions. Testing again I find infinite resistance everywhere except :
- red probe on source, black on drain : 0.9 volts drop voltage
- red probe on source, black on gate : 0.6 volts drop voltage

I've tested the nearby half-unsoldered shottky D9105 too, it shows 0 volts / 0 ohms in both directions. I guess there is no doubt that this one is dead.

yes replace it, its open... also be sure D9105 tested correct fine.

[lotas]

- red probe on source, black on drain : 0.9 volts drop voltage
- red probe on source, black on gate : 0.6 volts drop voltage

If the probes are swapped, what will it show?

[hervou]

Ho maybe I wasn't clear enough. Probes swapped gives infinite resistance / OL on both.

[lotas]

This means the Mosfet is working and there is no need to change it!
Replace the Schottky diode.

[Diah]

[post edited, I've just got that a multimeter in diode mode displays volts and not ohms haha]

Ho yeah, right. I was in diode mode, but I didn't test both directions. Testing again I find infinite resistance everywhere except :
- red probe on source, black on drain : 0.9 volts drop voltage
- red probe on source, black on gate : 0.6 volts drop voltage

I've tested the nearby half-unsoldered shottky D9105 too, it shows 0 volts / 0 ohms in both directions. I guess there is no doubt that this one is dead.

then from where this result you stated LOL

[lotas]

There is a diode installed inside the Mosfet, which shows the parameters of the diode.

[Diah]

There is a diode installed inside the Mosfet, which shows the parameters of the diode.

but he wrote that he took the other one out and he did compare the test ...LOL

[hervou]

Well maybe I wasn't clear enough. I have tested both components (and did not compare results).
- You advised to test the Mosfet, seemingly it is okay.
- I've also tested the Shottky diode and it looks like this one is damaged - but actually this is obvious since my message #5.

I just thought you were wondering if the damaged Shottky had also destroyed some neighouring component, which was a good thing to do.

I replace the shottky diode and keep you updated. Thank you guys !

[Diah]

I replace the shottky diode and keep you updated. Thank you guys !

if the diode shorted then you must change the Mosfet too, doesn't matter what reading you have by test.

[lotas]

Well then, replace U9105 too for the collection, it controls this transistor.

[Diah]

U9105 are innocent ... the crime with tester. LOL