Troubleshooting PFC with SPC1012T not raising output

[CoppellTVRepair]

It has been a long while since I posted here to ask anything....hope you can help.

I have this Samsung power supply that uses SPC1012T as driver for the PFC and it's giving me a problem that I can't narrow down though I've troubleshooted and fixed hundreds of boards similar circuits.

I am attaching a schematic from a different Samsung board, but the topography is very very similar if not 100% the same and it uses SPC1012T, which BTW is a pinout match to UCC28063 for which there's plenty of information. I haven't confirmed they are functional equivalents, but I strongly suspect they are.

While I haven't verified how close this diagram is to main board, as I said I've serviced tens - or more likely hundreds of them - and I can literally recognize most all major circuits and components here. If there are differences they will be very small.


The way this circuit usually fails is a results of aging power MOSFETS QP801S and QP802S. It is likely due to natural aging resulting for thermal cycling, but the reason is currently irrelevant.

When they fail they'd sometimes short out the gate and burn the gate circuit driver group. In some rare cases the surge would make it all the way to the IC itself and it would need to be replaced as well; most of the time it does not though.

At least a few times I've seen the driver IC fail on its own, i.e. it would be shorted (usually measured over a filter capacitor on the VCC_PFC line, not shown here), but the rest would be intact; of course sometimes when it fails it may also cause cascade failure up...but that is not the focus right now.

My current problem is that none of the above was wrong on that board: the MOSFETs were good (still replaced with new, same thing), everything in the gate circuit was also good (did not replace) and the driver IC was good (replaced, no change).

The circuit gets activated when VCC_PFC is passed to it of about 15V. A good circuit responds with a very short drop in that voltage to about 13.5V, still above the IC activation level, and within that very short time the output voltage goes from 160V to 390V, then the VCC_PFC goes back to 15V.

My board starts the PFC and drops down to 13.5V, however it stays there for a few seconds. The output voltage goes to about 250V and does not want to go higher. One simple way to test a "lazy" PFC is to manually activate it while the output is still higher - if it works fine it would be able to build on top of what it has already accumulated and keep raising.
This one does not - it caps the limit to about 260V.

In 1-2 seconds some of the watchdogs (not shown here) detect that the output voltage is low and turn off the relay over the NTC for the power section of the board. Then it starts again and the whole thing repeats resulting in clicking power.

I am positive the problem is limited to the PFC; it is not in the standby power, not in the VCC_PFC and not in excessive load on VCC_PFC from the stage consuming the PFC output.

Looking for ideas on what to do and why.

I would also very much appreciate it if someone points me to a document explaining how this particular topology works....without the heavy math, I just need to understand the principle so that I can isolate the problem by means of signal observation and comparison to a working board.

Thank you!

[momaka]

I would also very much appreciate it if someone points me to a document explaining how this particular topology works....without the heavy math, I just need to understand the principle so that I can isolate the problem by means of signal observation and comparison to a working board.

Well, there's not much math one needs to do when repairing these APFC circuits anyways.
The topology / method of operation is relatively simple and pretty much the same for every APFC circuit: the "boosting" MOSFET(s) (QP801S and QP802S in this case) turn ON, causing a certain current draw through the APFC inductor. When they turn off, the APFC inductor wants to continue "supplying" this same amount of current through it (since inductors don't like instantaneous changes in current), and it does this by pushing current through diode DP802 and into the primary bulk electrolytic capacitor(s). The result is an increase in voltage across the primary bulk capacitor(s). The APFC controller continues to switch the APFC MOSFET(s) ON and OFF until the desired voltage across the primary caps is reached. Sensing is done through the resistor dividers formed by RP819-RP823 and RP824 as well as RP828-RP832 and RP814+RP815, which both feed into pins 2 (VSENSE) and 8 (HVSEN) of the APFC controller.

So what can cause the APFC voltage to stay low?
- bad input film cap before the APFC inductor (CP801s in the above diagram - 1 uF 450V looks like)
- bad resistor in the voltage sensing circuit (typically it would be the "lower" resistors in the resistor divider - i.e. RP824 and/or RP814 / RP815)
- bad input bulk electrolytic capacitor(s) (i.e. CP805s and CP806s in the above diagram)
- bad driver circuit for one of the APFC MOSFETS (check that the small driver transistors don't have an open-circuit between any of the terminals out of circuit.)

With that said, check all of the components mentioned above, preferably out of circuit. In particular, APFC circuits are known to put a lot of stress on the input electrolytic filter cap(s), so always remove and check those caps when troubleshooting a troublesome APFC. I personally also like to put a PP film cap across the input electrolytic cap to help with off-loading some of the high-frequency ripple current from the electrolytic cap. PP film cap rated for 450V minimum (630V preferred) and at least 100 nF of capacitance is a good start. Speaking of film caps, check the one before the APFC inductor (CP801). If this one has lost its capacitance due to high stress / ripple current over time, it won't be able to supply the APFC circuit with high enough current to make it work.
Other things worth checking are the current sense resistor of the APFC (RP802, which is rated 0.02 Ohms and 2W) if you have a meter capable of measuring such low resistances. It's unlikely that it would have drifted / gone higher resistance, but you never know if the MOSFETs have failed before. Speaking of resistors drifting, high-resistance resistors have the tendency to go open circuit or drift high, so that's why I suggested to check the resistor dividers for the voltage sensing at pins 2 and 8. While at it, check the APFC inductor coil current sensing resistors RP803 and RP826... though I imagine these would cause the APFC controller to shut down if it can't detect any current in the coils of the APFC inductor (due to the resistors drifting to higher values or going open-circuit.)

And that's about all I can see that could go wrong.
If all of the above components check out fine, try measuring the voltage at pins 2 & 8 of the APFC controller as the APFC tries to come up to normal operating voltage and compare to what the datasheet says it should be. Make sure to use a meter that has a high impedance (i.e. most quality multimeters) and not a 830 "junker" meter with not-so-high impedance.

[CoppellTVRepair]

Thank you very much...I needed to hear that from experienced pro.

I'll sleep over it (some of it I can actually remember from my voltage multipliers course in college) and will get to test when other work clears up, then report here. I firmly believe one has to give if they take.

Very rarely seen bad film capacitors in TV power board BTW. Can't say never, but definitely rarely.

Fluke 87v and Tektronix scope are my measuring devices and I'd put my life on them if I have to....I am not rich enough to use cheap tools.
20 years into it I still say a silent thanks to Fluke every time I catch myself trying to measure 200 VDC in resistance mode.