LG Washer/Dryer Power Fail

[AdrianM]

Re: LG Washer/Dryer Power Fail

Oh yes redwire, you're right - I misidentified the ES3 part as a Shottky when it is in fact a fast Silicon switching diode as you say. The App note in the datasheet for a Multiple Output, 60 W, 185-265 VAC Input Power Supply uses fast Silicon diodes in all the higher voltage outputs so I figured it would be OK.


Unfortunately the LG PSU circuit only gets feedback from the 16.5V rail so the loading from the dryer fan motor leaves it with a poorly regulated voltage. With no load it puts out 23.5V dropping to 19.5V when the fan is running. The max current I measured was around 0.75A so the diode is dissipating around 0.675W so a Shottky would have been better. Oh well, next time maybe.

Indeed I replaced the TOP246Y with a TOP248Y which runs so much cooler!

[redwire]

Re: LG Washer/Dryer Power Fail

Thanks for the update.
The diode D243 is for 22V fan motor it seems. An ES3 is a normal diode 0.9Vf and will run ~20% hotter than Schottky SS36 0.75Vf for example but hopefully the upsizing is tough enough.
Hard to know how much current the fan motor uses, and it might go up with age.

Which TOPSwitch did you go with, the original TOP246YN?

[AdrianM]

Re: LG Washer/Dryer Power Fail

Topic Update: Well, after 2+ years of continued use, the machine died in the middle of a drying cycle on Monday. My immediate thought was that at least I had bought a decent amount of extra time - it's now 8 years old and it wouldn't be so bad if it was finally time to buy a new one. But...

Curiosity got me to have a look to see what failed this time. The fact that it went completely dead (no panel lights or response whatsoever) hinted at PSU again. Fuses all OK? Check. 240V on PSU input? Check. 12V/22V on the output? Nope.

A quick diode test revealed that D243 had failed short-circuit. This was the diode I that I replaced with a 2A 100V PIV Schottky that I had to hand at the time. I couldn't identify the original but I thought this ought to be OK. Obviously not. Anyway, I grabbed a 3A 200V Schottky and replaced the 1000u cap that follows it (after having been subjected to AC I thought this might be a good idea) and to my delight it all sprang back into life. Very happy with that. I've updated this topic in case anyone attempting a similar repair finds it helpful. D243 is now a Diodes Inc SMC case designated: ES3D-13-F (It just about squeezes into the gap between the diodes on either side.)

[CapLeaker]

Re: LG Washer/Dryer Power Fail

Don't get me wrong, I am also in favor of the larger TOP248 and I would have done the same, given that same situation.

I just wanted to figure out if a new TOP246 would have done the same, as I mentioned before in my post #4, since they do fail like this for some reason. Like redwire stated in his last post "thermal runaway".

[redwire]

Re: LG Washer/Dryer Power Fail

The best repair is when you fix it and up the power.

TOPSwitch gets hotter, on-resistance goes up, TOPSwitch gets hotter... and repeat. From 25°C to 100°C it goes up 150%
This is also why I am in favour of the larger part, less thermal runaway.

Does it seem strange to price-point these TOPSwitches at silly levels every few watts?
Is 0.5mm more silicon for the mosfet really going to cause price to skyrocket?

[momaka]

Re: LG Washer/Dryer Power Fail

Good to hear it's fixed and all running good now! Let's hope it stays that way too. The dryer fan will have to try much harder now to kill the PSU. But if it uses crappy bearings, maybe not. On that note, it probably uses sleeve bearings, so should have dropped a bit of oil on them (if that's even possible, as some fans are pretty good at making their bearings inaccessible.) But hey, if the washer makes it another 3 or more years, I'd say you still did a good job. At least that's how much I am expecting to get out of my tub bearing job. Some of the reviews of the really cheap bearings online say they only got a year, which is pretty terrible of course. I think 3 years is just about enough to say the repair was worthwhile (and of course, if more, then that's all fine).
Yeah, doing a bearing job is pretty much a full disassembly process. But I didn't have to touch any electronics, though. Just some blood, sweat, and tears to get those stupid bearings out.

[AdrianM]

Re: LG Washer/Dryer Power Fail

would have been nice to see if a new TOP246 would have done the same, but at least it is working now with the new TOP248.

These parts seem to have rather high on-resistances in general so it feels right to take this opportunity to waste a little less power.

But yes, I must admit to being curious myself about a like-for-like replacement however the repeated tear-downs and re-building were beginning to wear both me and the board connectors down. I strongly suspect a direct swap would yield the same result but I did actually buy a new TOP246 as well as the TOP248. When I get a chance I therefore plan to breadboard a small test circuit following the procedure outlined in the datasheet and compare their on-resistances.

[petehall347]

Re: LG Washer/Dryer Power Fail

if its running cooler it is better . surely the rating is maximum and not normal running . higher rating fuses might well work for a while . a new motor would clinch the deal

[CapLeaker]

Re: LG Washer/Dryer Power Fail

would have been nice to see if a new TOP246 would have done the same, but at least it is working now with the new TOP248.

[petehall347]

Re: LG Washer/Dryer Power Fail

i thought d243 is marked as a zener on the diagram then i made it bigger and now its a schottky .

[AdrianM]

Re: LG Washer/Dryer Power Fail

D243 was a Silicon but is now a Schottky. I finally extracted the TOP246Y:


And replaced it with a TOP248Y plus 12K resistor in series with X pin to maintain the current limit (plus a little bit):


So now with the Fan running, after 30 minutes it's only getting to 47C and not tripping the thermal shutdown anymore:


I'm coming around to the notion that the dryer fan bearings are worn and putting extra load on the PSU. When I monitor the supply rails with and without the fan running I see the following:

[Ideal] Idle -> Fan running
[12.0] 13.1 -> 13.2
[16.5] 16.0 -> 15.9 (Feedback tap)
[22.0] 23.7 -> 19.5 (Fan supply)

(Ideal is what is printed on the silkscreen against each rail)

Feedback is from the "16.5V" rail so that rail stays closest to the design ideal. The Control input circuit is pretty slack though - using 9.1V Zener (ZD242) plus a diode drop (SD241) plus Control pin input current of 6mA * R243 to maintain 5.8V at the C pin resulting in a theoretical stabilised output of 16.2V

Anyway, this TOP248Y upgrade has indeed remedied the "Power Fail" problem for now, and as everything else is working correctly I can keep an eye on the Fan and see if it continues to draw more and more current (unless it falls apart first).

[petehall347]

Re: LG Washer/Dryer Power Fail

Thanks for the vector to the AN. I noticed that the output diode for the fan supply also runs very hot. I can't find the part's spec. it's case code is 3L2 125 It measures 0.46V forward drop on the diode test range. I guess it's a fast recovery Silicon diode not a Schottky.

is d243 not a zener ?

[AdrianM]

Re: LG Washer/Dryer Power Fail

I was just thinking slightly differently about the overheating... the T0220 heatsink looks to be about 12C/W and it's getting to 120C over ambient therefore the TOPswitch is dissipating something in the order of 10 Watts. The converter efficiency should be something like 80% which implies an active load of 50 Watts. But I know the AC power going in is only half this when the switcher overheats - so it has to be that the TOPswitch is only running at 40% efficiency for whatever reason.

I know this is just more "spinning the wheels" until I get to try a new IC but I felt like writing out the figures here so I can keep track of the process.

[AdrianM]

Re: LG Washer/Dryer Power Fail

With the case of my LG washer, I replaced the bearings last year (after I could no longer stand the freight-train noise the washer made with those shot bearings for nearly 2 years ). Works okay now. But when I looked at it while it was disassembled, indeed there are quite a few engineering design goofs (e.g. drum cover has non-smooth back, which tends to collect water and then grow mold over time - absolutely silly!).

Oh, perhaps I have that to look forward too :-( Pretty major disassembly you took on there!

[AdrianM]

Re: LG Washer/Dryer Power Fail

I think the IC can get too hot from the SMPS having low efficiency or heavy load. The thermal IR camera can show if another part is stealing output current.

Thanks for the vector to the AN. I noticed that the output diode for the fan supply also runs very hot. I can't find the part's spec. it's case code is 3L2 125 It measures 0.46V forward drop on the diode test range. I guess it's a fast recovery Silicon diode not a Schottky.

I did briefly try a 1N5822 (40V PIV Schottky) in place but the actual PIV must have been too high because it went into break-down. Good thing the current limit(s) kept things sensible. I've swapped it for a 2A 100V PIV Schottky in the same SMD package size and that's running cooler already.

I went over the PSU circuit once again and found another CR snubber that I'd left off my first schematic (R241, C241). I've also corrected some part ID's.



I've now desoldered most of, and tested all of the individual parts around the TOPswitch and nothing has been out of spec. DZ241 is a 180V Littelfuse TVS so I took it out and tested it with a scope using the flyback from a relay coil. Sure enough it clamps at about 180V. I'm pretty certain there's nothing wrong with any of the supporting parts although I have replaced the big DC cap CE72 with a EPCOS / TDK B43545A5337M000 becuase the original has probably clocked-up enough hours already.

I've got to wait until next week for the replacement TOPswitches to show up.

[momaka]

Re: LG Washer/Dryer Power Fail

The S.Korean joke is the big sticker on the front boasting a "10 Year Guarantee" on the DC Motor.

Yup. I know that all too well, as I have a slightly older LG washer myself. 10 years warranty on the DC motor (the part that is absolutely NOT going to break anyways). Mine shot its drum bearings at 7.5 years old - I guess that's not too terrible. Still, considering how little load we've put through it, I think it's laughable. My sister's Candy washer (overseas) is about to turn 25 years old and never needed anything serviced. Still works like a champ and goes through at least 2-3 full loads daily (it's a smaller washer made for European standards). The Candy washer we had before that one made it to 20 years before its coated steel drum enclosure rusted away. And my grandmother had a GDR (East Germany) -made washer that gave trouble-free operation for 30 years. So I think that puts a good perspective on how low the quality has gone with today's appliances.

Still, I agree that it's not worth discarding an appliance over a tiny failed part. With the case of my LG washer, I replaced the bearings last year (after I could no longer stand the freight-train noise the washer made with those shot bearings for nearly 2 years ). Works okay now. But when I looked at it while it was disassembled, indeed there are quite a few engineering design goofs (e.g. drum cover has non-smooth back, which tends to collect water and then grow mold over time - absolutely silly!).

[redwire]

Re: LG Washer/Dryer Power Fail

I think the IC can get too hot from the SMPS having low efficiency or heavy load. The thermal IR camera can show if another part is stealing output current.
But running hot with low loads on the PSU?

Power Integrations AN-14 lists reasons for a TOPSwitch to overheat. As budm mentioned, the snubber zener can apparently cause trouble too. So there are some supporting parts to check.

[CapLeaker]

Re: LG Washer/Dryer Power Fail

The IC shouldn't be at 60C at idle. That is a very hard pill for me to swallow on a TO-220 that has a heat sink on it. Id say the TOP244 would be running at full tilt at 20W.

[AdrianM]

Re: LG Washer/Dryer Power Fail

Oh! I really done goofed. I can't begin to explain what led me to read that wrongly. So now it's even more baffling...

Going back to my plug-in power meter test, and seeing 8 Watts from the AC input then @ 80% eff., the baseline load supplied by the TOPswitch when the machine is idling is just 6.4 Watts. I can then easily and repeatably get a 80C+ rise on the part by adding 13.5 Watts of load by running the dryer fan for 5 minutes. This takes the IC from around 60C when idle up to the ~140C trip-point. Surely even the TOP244YN should manage 20 Watts without this much drama?

[CapLeaker]

Re: LG Washer/Dryer Power Fail

I see what redwire sees. That isn't a TOP242, it's a TOP246! So you need a new switcher first!!!