Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

SMPS repair

Status
Not open for further replies.

whiz115

Member
Hi guys

i just need a confirmation on some things so can repair a SMPS from a TV.

the components marked in red in the photo are burned. I guess first the mosfet got shorted
and then these components failed altogether. Ι can't find anything else, can i assume
only these are failed? or maybe the SMPS IC is also burned?

also i want to ask about the fuse... it's a "buss sr-5 t3.15a 250v", what can i do about it?
how can i substitude if i can't find that type of time delay fuse?

same about the thermistor it's blown, how can i subsitude if i can't find the 7S100L type?

thank you.
 

Attachments

  • sch1.jpg
    sch1.jpg
    160.7 KB · Views: 275
You just need fuse 3,15A and NTC termistor with 10R for at least 3,15A, change mosfet and current sense resistors (5x 10R atleast 3W ones) ?
 
3.15A A/S is a standard fuse value, and VERY common in TV's. The thermistor could be replaced with a similar device from a different TV, or even a 4.7 ohm 7W resistor, whivh is commonly used for reducing the surge.

It's likely the chip has been killed as well, I would change that, the FET's, the other bits you've identified, and check EVERYTHING else. Semiconductors in particular, diodes etc. often die when the FET's fail - in my (considerable) experience FET's are far more unreliable in TV PSU's than bi-polar transistors.

I used to assemble repair kits for the most common TV's, and change the entire kit every time one came in with a duff PSU.

One unusual failure is the reservoir capacitor, these sometimes go O/C, and while some sets work quite happily, others destroy the PSU.

Even more unusual is transforrmer failure, I've seen a few failures over the decades - presumably it's a poor design, as it's usually the same model where transformers fail.

If you don't replace all the faulty parts it will self destruct the instant you turn it on.
 
Ok with the fuse i have found same type..

Nigel you are right!! there are also two shorted diodes, seems such PSUs become domino if something gets burned, i guess if two diodes failed none is trustworthy and each one of them should be replaced, unfortunately its not possible i replace every component.
 
Ok with the fuse i have found same type..

Nigel you are right!! there are also two shorted diodes, seems such PSUs become domino if something gets burned, i guess if two diodes failed none is trustworthy and each one of them should be replaced, unfortunately its not possible i replace every component.

They do act like chains of dominoes, I've repaired thousands of them - before the advent of LCD sets almost all TV and set-top box faults were SMPSU failure.

As I also said, FET ones were much worse than bi-polar - which is presumably why they were much less common.
 
other than the SMPS IC, i've managed to get all of the components i was thinking to replace, but i worry to turn it on...
the resistors blown were SMDs and literally exploded, some of them destroying also the pads, but it's fixable.

capacitors are in good shape, can i hope it was the shorted FET causing all this, so i can go on and turn it on now?
i don't mind repairing it again, but it's a waste of components.


it's a pretty neat looking power supply, not the usual ones somebody would think..
 
Last edited:
I agree that the IC is probably dead too. If you want, you should be able to check it by feeding a DC voltage to its power pins and seeing if it produces an output - check the datasheet to see what conditions it needs, exactly, in order to oscillate. Even simpler would be to check continuity from the output pin to the supplies - if it looks like a near-short then it's definitely dead.
Here in the UK the FAN6300 is in stock at RS components, so it's still available - although obviously this depends on your location.
 
The IC connects pretty well directly to the FET's, it's unlikely not to have been killed.
I agree that the IC is probably dead too.

...and the FET gate is shorted to drain and source!! they fail in a very destructive manner... :(

this is going to take days to be repaired, if i manage...

If you want, you should be able to check it by feeding a DC voltage to its power pins and seeing if it produces an output - check the datasheet to see what conditions it needs, exactly, in order to oscillate. Even simpler would be to check continuity from the output pin to the supplies - if it looks like a near-short then it's definitely dead.

i did a basic pin check before posting here and i didn't noticed something wrong,
though trying to power it up and checking the gate drive pin it's PWM and it's hard without an oscilloscope.
 
Last edited:
Change the chip - as RS Components have it in stock, they are available pretty well most places in the world.

Just checked RS - they even have a website in your local language - no idea what it says though :D
 
hi again

as tomizett proposed i tried to power up the SMPS IC hoping somehow it has survived...but it didn't so1
I ordered it and replaced it but the problem is i'm trying to repair a tv which has no indicator LED.

although nothing is blown after replacing the components and i can measure 303VDC at the high voltage capacitors
(with 215VAC mains), that's all i can measure.. since i'm no expert i need some guidance what else to check, because i
don't get any meaningful voltage readings anywhere, maybe also the transformer winding is not ok?? the continuety
tester shows short circuit on most pins, also most of the components are SMD in the bottom side which i don't have
access when the PSU is powered.

It has a side connector to the logic board and maybe it's grounded to the chassis from the screws. Some
PSUs don't work or blow up if you turn it on without load or not grounded properly, is that right?

in the following photo you can see the voltages i was trying to measure, also the diode in the red square was
shorted! it's a 5A fast schottky diode, if such a big diode got damaged, does that mean the logic board is
also damaged?


[edit]

with a power meter i've noticed the TV initially draws 5 watts for around 3-4 seconds and then it draws
around 300mW and constantly cycling from zero to 300mW.
 
Last edited:
When diodes are paralled like that they VERY commonly go S/C, you should replace all three with diodes from the same batch/packet (they usually come 5 in a packet). The SB5100 in particularseems to be an EXTREMELY common failure. Basically it's aserious design flaw, there should be small value resistors in series with each diode to spread the load equally.

Is this a Vestel (made in Turkey) board, they are notorious for failure of the SB5100's due to their design flaw - but usually it's only the diodes that need replacing.
 
When diodes are paralled like that they VERY commonly go S/C, you should replace all three with diodes from the same batch/packet (they usually come 5 in a packet). The SB5100 in particularseems to be an EXTREMELY common failure. Basically it's aserious design flaw, there should be small value resistors in series with each diode to spread the load equally.

Is this a Vestel (made in Turkey) board, they are notorious for failure of the SB5100's due to their design flaw - but usually it's only the diodes that need replacing.

i can replace the other two but they don't seem damaged, they measure ok, you think this is the problem
and not the transformer? what about the UF5402 do you think they should replaced too, although they measure fine?

that zero-300mW cycle i see in the power meter it's a "smps hiccup" ?
 
The thermistor is a std ICL from Digikey, you can match specs.
The FETS are ganged and if only 1 failed then that is a sign of thermal runaway where tempco or an avalanche failure. for current sharing They must also be thermally matched
 
Last edited:
The thermistor is a std ICL from Digikey, you can match specs.
The FETS are ganged and if only 1 failed then that is a sign of thermal runaway where tempco for Ron is negative. Look at HEXFETS which have PTC Ron for current sharing and others. They must also be thermally matched if not PTC.

the NTC has been replaced, bridge rectifier diodes also replaced (two were shorted) the FAN6300 replaced with an onsemi FAN6300A
also the 5 resistors shown in the photo of the first post, the TV board seems it's slightly different revision and has only one mosfet in the high
voltage area which was shorted and also replaced and in the secondary in the low voltage area only one SB5100 seemed shorted and replaced.
 
Last edited:
i can replace the other two but they don't seem damaged, they measure ok, you think this is the problem
and not the transformer? what about the UF5402 do you think they should replaced too, although they measure fine?

that zero-300mW cycle i see in the power meter it's a "smps hiccup" ?

As I mentioned above, the serious design flaw of no series resistors means the diodes need to be well matched to stand any chance at all, so replace all three with ones out of the same batch/packet. As I asked before, is this a Vestel set?.
 
As I mentioned above, the serious design flaw of no series resistors means the diodes need to be well matched to stand any chance at all, so replace all three with ones out of the same batch/packet. As I asked before, is this a Vestel set?.

yes, now i'm planing to replace all SB5100 found on the board, the UF5402 also the three BA159 and two SMD RS1008FL
and the C15 zener found in the optocoupler.

is that right replacing all these components? what about my measurement with the power meter does it seem like
a "smps hiccup" ?
 
yes, now i'm planing to replace all SB5100 found on the board, the UF5402 also the three BA159 and two SMD RS1008FL
and the C15 zener found in the optocoupler.

is that right replacing all these components? what about my measurement with the power meter does it seem like
a "smps hiccup" ?

It sounds like the PSU is either tripping (excessive load on the secondary) or failing to start up (so keeps trying to start, but failing).
 
It sounds like the PSU is either tripping (excessive load on the secondary) or failing to start up (so keeps trying to start, but failing).

[update]

i tried to power it up with a 100ohm load at the 12v output pins and the resistors in the snubber circuit burned again.

so its over, thanks for the advices.
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top