LG Washer/Dryer Power Fail

[AdrianM]

We have a 6-yr old LG combo washer dryer model 1480YDS that has started to cut out part way through the drying cycle. This has now become a highly repeatable failure giving me a chance to diagnose the problem. On restarting the machine (after an imposed wait of a minute or two) the display would show PF (power fail error) indicating an unexpected removal of power part-way through a programme.



Searching around the internet this fault seems to be fairly common to a wide range of similar LG washer dryers and because the only solution people were finding was to replace the entire control board for around £130, I decided to work out what sub component was failing.

I wondered initially if it might me a glitchy relay or something else relating to the drying phase causing the microcontroller to crash out because the machine would happily do a full wash lasting over an hour, but would always stop after 5 minutes or so of tumble drying. Equally it might have been the increased heat of clothes drying - but disconnecting the heater element (both cooling things down and removing a big switching current) made no difference to the time it took to cut out.

I figured that if programmed machine logic was involved it would report a more specific error so the PF error is indicative of an unexpected power down. This got me thinking about the power supply itself. Sure enough, a thermal camera image of the control board in action revealed one very hot TOP242 IC. The shot here shows it still working at 110C but it will go past 130C whereabouts the built-in thermal shutdown trips. Bingo!



The TOPSwitch datasheet confirms the trip-point to be 130C to 150C and a ghetto-rigged fan proves the point finally keeping the machine going all the way through a 1 hour + drying cycle with the TOPSwitch staying under 50C.



Tempted as I am to bore a hole in the control box lid and mount the fan there full-time, it's obviously not fixing the problem at source. I think the most likely reason for the TOPSwitch to overheat would be the 330uF 400V input reservoir capacitor ageing. All the capacitors are suspect of course so the next step is to try and remove/test/replace as necessary. Unfortunately the PCB is flooded in the case with a clear potting compound that I'm not looking forward to tearing out. Will try and keep this posted.

[stj]

Re: LG Washer/Dryer Power Fail

try melting it with white spirit, it works on Honda ECU's

[momaka]

Re: LG Washer/Dryer Power Fail

Check the small startup cap and also the output caps of the SMPS. I think those would be the primary suspect in causing the TOPswitch to overheat. In addition to that, maybe add an aluminum ribbed heatsink on top of the TOPswitch with some thermal adhesive (Arctic Cooling has a two-part thermal adhesive suitable for that).

[CapLeaker]

Re: LG Washer/Dryer Power Fail

As I recall, these TOP242 or similar IC's are known to fail. I'd recap the board and put a new TOP242 and be done with it.
Makes me wonder... I have a 2 year old LG washer (no washer dryer combo) and it looks similar to yours.

[stj]

Re: LG Washer/Dryer Power Fail

topswitches are fine if the startup cap is good.
if it fails the top can explode.

[AdrianM]

Re: LG Washer/Dryer Power Fail

Thanks everyone. I ended up prising the board out after running a knife around the edges.



The potting compound is very soft and not too bad to scrape off to get to the component pads. I started with the 330uF input cap but this actually metered-up OK @ 322uF 0.28 Ohm.



I've since exposed all the parts in the PSU area and none of the caps are out of spec. Sure I can re-cap them all but some have lower ESR than the new ones I'd be replacing them with so I think its a red herring. All the bigger O/P filter caps are good low ESR types. The only exception is the 47uF start-up cap that momaka and stj mention. This one reads 40uF 1.2 Ohm which isn't exceptionally bad I don't think, but I will see if I can find a better one.

Notice though that there are some browned areas notably above the SM diode that's rectifying the 22V feed that ultimately powers the dryer fan. The TOPswitch output transformer also has two 12V outs, one of which feeds the logic stuff via a 7805 and another isolated tap that seems to be referenced to the HT for something or other. I think the 16.5V may be for the IGBT driver for the drum motor.

These supplies are all fine when loaded by the washer cycle alone. But as soon as the drying cycle is engaged the 22V supply comes into play and this is what ramps up the heat in the TOPswitch. The dryer fan is by far the biggest load supplied by the regulator. I guess it could be a problem with the fan? It runs OK and, as I said before, if I blow air over the TOPswitch it settles down to a sensible temperature and will run all day like that. I'm mighty puzzled now.

[stj]

Re: LG Washer/Dryer Power Fail

check the blower spins o.k.
a friend had to replace one because the bearings had corroded.

[redwire]

Re: LG Washer/Dryer Power Fail

The blower fan may have aged and is using more current, and the power suppy is undersized.
I would run the blower from a DC power supply and see if the motor is ok and no lint and crap blocking it, doesn't need oil, brushes OK etc.

The TOP242 is the lowest power part in the family.
You could drop in a higher power part like TOP243 or TOP244 which would (only) run cooler due to the much lower on-resistance.
Have to check if this would work - the transformer current must be the same, the circuit must have the resistor on the X pin. If so, you can drop in a beefier part.

TOP243 which has 1/2 the on-resistance and more output power (25W vs 15W), or TOP244 which has 1/3 the on-resistance (28W vs 15W). Lower on-resistance = less heat. Potentially 1/2 to 1/3 the heat.

[petehall347]

Re: LG Washer/Dryer Power Fail

its 6 years old and fine until now .. even if original spec parts are replaced it in theory will do another 6 years . only if the other parts will last that long .
i say run the fan off an external supply . add a 1 ohm resistor and voltmeter and calculator or simply an ammeter of an accurate clamp on type if you like .

[momaka]

Re: LG Washer/Dryer Power Fail

The only exception is the 47uF start-up cap that momaka and stj mention. This one reads 40uF 1.2 Ohm which isn't exceptionally bad I don't think, but I will see if I can find a better one.

1.2 Ohms is indeed not bad for a small cap like that. Nevertheless, knowing that this appliance is made by LG, the cap is probably a SamYoung, SamWha, or some other similar Korean brand (or worse - Chinese) that aren't known for their reliability - in which case, it wouldn't be a bad idea to replace it anyways. Same can be said about the rest of the PSU caps, if they are of the same inferior brand, but since this is not the problem here, I agree that you don't need to replace them.

Since you've identified that the issue happens when the dryer fan is running, it may be a good idea to measure its current draws, as redwire and petehall347 suggested. Should be easy to do with a multimeter.

Also, if bad comes to worse, you can always disable/disconnect the washer/dryer power supply's 22V rail and instead wire it to an isolated 20V laptop power adapter. That should provide more than enough current for the fan.

its 6 years old and fine until now .. even if original spec parts are replaced it in theory will do another 6 years .

Not necessarily.
If the dryer fan bearings are starting to go bad indeed (or even not bad but just aged and showing slightly more friction), it's possible for this to make the fan draw more current. And if the washer power supply was marginal or not specced properly to begin with, then a new replacement TOPswitch very likely won't last the same amount of time.

[AdrianM]

Re: LG Washer/Dryer Power Fail

Unfortunately the dryer fan is actually a brushless motor so there's a three-phase inverter driver. Also the machine logic is monitoring the fan speed via a voltage summer/comparator and will halt with an error if the fan isn't detected. I was trying to test the PSU with the fan unplugged when I discovered this.



(image from servlib.com/lg)

The fan seems to spin freely and measures 4 Ohms across each winding so appears to be in spec. I just ran the dryer while measuring the 22V rail. With the fan off it's 24V then drops to 19V with the fan on. I guess I could cut a track between the 22V reservoir cap and the fan driver and measure the DC current - although I'm not sure what it will tell me without knowing what a "good fan" should draw.

[stj]

Re: LG Washer/Dryer Power Fail

maybe the cap is going and cant handle the high frequency ripple caused by the motor driver.

[momaka]

Re: LG Washer/Dryer Power Fail

The fan seems to spin freely and measures 4 Ohms across each winding so appears to be in spec.

That doesn't matter.
As the fan is loaded, the current draw will change. On that note, check if there are any dust filters behind/in front of the fan. If they are clogged, they would cause it to draw higher than normal current. (Hint: always clean fluff from your dryers ).

I just ran the dryer while measuring the 22V rail. With the fan off it's 24V then drops to 19V with the fan on. I guess I could cut a track between the 22V reservoir cap and the fan driver and measure the DC current - although I'm not sure what it will tell me without knowing what a "good fan" should draw.

Going from 24V and sagging down to 19V is pretty sad (though not necessarily bad).
It doesn't matter if you don't know what a "good fan" should draw. If you know how much current the fan is drawing and you know the voltage it is running on (in this case, 19V under load), then you can tell how much power it is drawing (Volts x Amps). From this, you can figure out if perhaps the fan is drawing near maximum design power of the TOPSwitch IC, which is 15W as mentioned by rewire in post #8 above).

That said, YES, you can cut a trace (or remove the rectifying diode) for the 22V rail and insert 20V DC from a laptop power adapter there. Then you can measure the current draw on that rail much easier... and then determine if it's getting close to the maximum of what the TOPSwitch can handle.

(But again, make sure you use an isolated laptop adapter... or just one with the ground pin floating, so that you can tie the adapters ground to the washer's ground. I imagine the washer's PSU ground is already connected to earth ground of the house wiring. But nevertheless, I always like to have only one ground entry point for avoiding ground loops and noise.)

[AdrianM]

Re: LG Washer/Dryer Power Fail

So this thing turns out to be "hot" everywhere. I traced out the front-end and AC Line voltage is but a diode-drop away from the so-called GND bus. GND isn't actually tied to chassis but it makes it hard/risky to work on.



My line isolating xformer probably won't handle the power of this thing so I'm going to have a tough time powering the 22v circuit independently or checking the ripple with a scope.

The TOPswitch is configured for half-frequency operation (F connected to C) and Line-sense and External Current Limit Features Disabled ( L & X grounded). It's also a bit weird around the flyback snubber and output control sensing areas. Do we still think a TOP243/244 would drop in here?

[redwire]

Re: LG Washer/Dryer Power Fail

Assuming the very hot IC is the problem, and not a side effect due to low line voltage or a bad electrolytic capacitor CE72, CE244, CE242 etc. I would say the IC needs cooling from extra PCB copper, or a heat sink/bigger package. The conformal coating over it acts as an insulator. A stressed/aged IC can have higher on-resistance and shifted current-limit so it cooks itself.

You can try add a heat sink or replace the TOP242. I would put in a TOP243 and extra resistor, to get 1/2 the heat.

With the "X" pin is connected to the source, then the part operates at max. current limit.
TOP242 is 0.45A and a TOP243 is 0.75A so going to the TOP243 it's best to add a resistor at X-pin to prevent the transformer from saturating. From Fig. 54b a 13k resistor from "X" to "S" will keep the same old 0.45A current-limit as a minimum. But the bigger part runs cooler with 1/2 the on-resistance.

Wait - the "X" pin is not brought out on the DIP-8 or SMT-8 package part, so this idea wouldn't work for that. Unless this the TO-220-7 packaged part TOP242Y ?

[budm]

Re: LG Washer/Dryer Power Fail

Also verify the RCD snubber for that SMPS IC are still OK.

[AdrianM]

Re: LG Washer/Dryer Power Fail

Here's my test of the fan motor power. I pulled D423 off the board and jumpered it onto a breadboard with a separate 1000U cap to GND. From here I measured the current at 0.73A and volts at 18.3V so the fan motor alone takes 13.5 Watts. Resistance in all the dangling wires probably means more like 15 Watts when the diode is back on the PCB.



Plugging the machine into a power meter adapter shows 25W going in during this test and at idle it's 8W all of which can only be supplied via the TOPswitch. So a delta load of 13.5W shows up as 17W at the input so the convertor efficiency is as expected at around 80%

This makes the output power around 20 Watts if the input power is 25 Watts so I think however you look at it the TOP242 Y is being pushed too hard. It looks like redwire's suggestion of substituting a TOP244 is going to be the way forward, and as it's a T0-220 I can series-in a resistor with the X connection to stay with the original transformer design parameters.

[CapLeaker]

Re: LG Washer/Dryer Power Fail

Sure looks like the PSU isn't able to supply power as it should. Forced obsoleteness, I say. If it were mine, I'd recap the thing and get a top242 and a top243 and see what happens with a new top242 in place. I also think that the current top242 may have cooked itself so many times, that its characteristics changed a bit. Probably the top243 is the answer at the end, but I'd be curious if a new top242 would run this thing again like it was before.

[AdrianM]

Re: LG Washer/Dryer Power Fail

I can get a TOP244 here by tomorrow and the laundry is piling up ;-)

@redwire, looking at Figure 54b the scaling factor is 0.45 for TOP242 so with X at zero Ohms the transformer primary is current limited to 0.45A.

To get the same limit for a TOP244 with a scaling factor of 1.35, am I right in looking along the Y axis for 0.45A/1.35 = 0.333A which seems to be around 39K?

[momaka]

Re: LG Washer/Dryer Power Fail

Wow. 13.5W on a part designed for 15W.... that's really pushing things too close. No surprise it's failed after a few years. Perhaps if the PCB was designed with more heatsinking, the TOP242 might have lasted a little longer (or a lot more).

I agree with the others that a TOP243/244 is probably the best route to go. But nevertheless, I still think it would be a good idea to attach a large aluminum heatsink/heatsink on top of the new chip. PCB cooling could still be marginal, even if the new TOP243/244 chip has lower dissipation.