Update on Whirlpool refrigerator powering down after a power glitch (really long)

Page 1 of 2  
This is an update on my new (July 2015) Whirlpool '4 door' French door fridge. I previously reported that while on a 1 week vacation in April 2016, the unit powered down and spoiled all the food. The ice also melted and warped the laminate flooring in front of the unit. I explained to WP that there was some kind of a power glitch because the WP oven flashed the PF (power fail) icon, although it didn't fully loose the time as it was just off by about 5 minutes. The WP microwave totally lost the time and was blank. There was NO indication on the fridge itself. If there was, my friend that came into the house while we were away, probably would have noticed it. Pushing and holding the 2 buttons that turn on and off the cooling, eventually flashed the “cooling is off” icon. Duh, it was already off, just not showing. But now it was showing. After doing it again, the panel briefly read “cooling is on” and the compressor and fans all started. Everything was normal with the exception of the water dispenser. The fridge cooled down and worked normally. After calling WP they took the “it worked as it should have” attitude. I asked WP if it was in the sales information that it was supposed to shut down and spoil all the contents? All they could repeat is that it “worked as it should have”. The water dispenser was stuck in measure fill mode where it provides 16 ounces and then shuts off, nice if you put an 8oz glass under it. I couldn't get it back to the normal mode of providing water only while the paddle is being pushed. I did NOT try a full power down (unplugging). The WP tech said there's nothing wrong, however, he reset the controller board by unplugging it and then re-plugging it. After that, the water dispenser worked normal. The controller was obviously stuck in some bad states and required a “reboot”.
I was going to send an email to the WP CEO, however, on their website, it gave an email for the corporate problem solver. So I did that at 2AM and got a phone call from WP at 9AM … pretty good, I thought. After I complained that a fridge should recover after a power outage and should run 24/7/365 WP offered me $150 (I facetiously call it hush money). Plus they wanted to send out a 'senior tech' to check it out. The senior tech turned out to be the same tech that came out the first time (WP's authorized repair company). I explained to him that Corporate had summoned him, not me. He called the WP tech line (only for technicians) and after talking to them, he said that I probably found something that WP didn't or couldn't find. We tried turning the circuit breaker off and on both fast and slow and couldn't reproduce the problem. WP seems to think that there is some kind of icon or light that tells of a power failure. Attention WHIRLPOOL … THERE ISN'T!
Later I called the guy at Corporate and told him nothing was done by the 'senior tech'. He now jumped on the “it worked at it should have” bandwagon. I told him their product was crap and it should be fixed. I told him there might actually be something wrong with the controller, so they should replace it. He said he can't tell the tech what to do. He eventually offered to buy back the unit for what I paid for it.
Yesterday, I went to the appliance store (not a big box store) and the guy there told me that he's seen this problem in several different brands, even some of the ultra high end $5K stuff. He also informed me that my French door fridge with an extra refrigerator drawer (so called 4 door refrigerator), was not actually made in the USA but started life in Korea. Apparently, Samsung owns the patent on the 4 door fridge, so WP buy them and puts their own badge on it. Nice, I wanted to buy US and WP secretly sold me something from Korea. To get around the patent thing, one company now builds a 5 door unit that basically breaks the fridge drawer into 2 drawers.
After working in the industry with high end fault tolerant systems for some 25 years, I realize now that WP and the rest of the industry doesn't know the meaning of fault tolerance and recovery. Apparently, this industry is just now learning stuff about electronic micro controllers. Stuff that I and my former company learned many years ago. So taking the WP buyback wouldn't accomplish anything.
My first though was, when away from home for several days, I would build a unit to a timed power failure to the fridge each day for say, 5 minutes, thus allowing the controllers to reset. However, I discovered that if the unit was in the “cooling is off” state, after the reset, it remembered that and came back in the “cooling is off” state. So, I would actually have to design something to press buttons and interpret the results … way to complicated. Bummer! I found data loggers on the market that will log periodic temperature readings to the cloud and then notify me via email or text when it falls outside my preset limits. They seem to fall within the $150 'hush money' amount. At this point I think that is the best option.
Sorry for being so verbose but I just thought people would like to know.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 7/9/2016 11:46 AM, Art Todesco wrote:

I wanted to buy a US made fridge too, but we only have a 33" wide space. To get what we wanted for features, Samsung and LG had the only model that fit.

Yes, now that I know my fridge may be the same as yours I'll certainly watch for power failures.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 7/9/2016 12:26 PM, Ed Pawlowski wrote:

Actually, it is 36" wide, but still may be made of the same bad stuff.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
Unfortunately, testing for all the possible failure modes for electronic eq uipment/appliances is an art that few companies seem to understand. Whenev er I buy anything, I look for the simplest possible unit. Plain old motori zed controllers are clunky and not in fashion, buy they are a whole lot mor e reliable if you don't actually have your own electric power distribution system. Until batteries and photocells and wind-powered generation are com mon, we will just have to do the best we can. That may include deliberatel y trying various failure modes when we first get new stuff, so that if we f ind a problem it is covered by the manufacturers warranty.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 7/9/2016 10:30 AM, snipped-for-privacy@att.net wrote:

That may not always be possible. Certain capabilities are just impractical to implement WITHOUT a microcontroller. Imagine a "set back" thermostat with HALF the capabilities of an inexpensive digital unit -- implemented with bimetal springs, motors, gears, etc.

When shopping for SWMBO's new vehicle, we dismissed Nissan and Kia products as we encountered problems with the electronics in vehicles "sitting under the watchful gaze of the dealers" -- i.e., if THEY can't keep BRAND NEW vehicles operating perfectly, *we* don't want to be wasting trips to the dealership to repair/replace those things, even if "at no cost" (our TIME and INCONVENIENCE is worth something to us, even if not to the dealer!)
Granted, its a small sample size -- and, could be "flukes". But, lots of car manufacturers and lots of different models to choose from; why pick ones that have demonstrated themselves to have faults?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On Saturday, July 9, 2016 at 11:46:15 AM UTC-4, Art Todesco wrote:
Given that you can't reproduce it by cycling power on and off, I would think that should give you some confidence that it may have been a rare occurrence that lead to it being in some weird state, like the power sequenced through some intermediate state, where it had lower voltage, etc. If it's a typical power fail, it sounds like it would be OK, but it could get into that locked up state again. So I guess you either live with it or let them buy it back, which is generous. Another aspect is how often there would be a power fail where it could wind up in that state and nobody would be around to spot it. Here given the reliable power and the fact that even if it got into that locked up mode, it would be very low probability of that happening without someone around to notice it before it could defrost.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 7/9/2016 8:46 AM, Art Todesco wrote:

?? Through-the-door ice dispenser? (i.e., gravity fed with ice so melting ice feeds *water*?

It depends on what causes the "PF" icon to illuminate. E.g., one way of "detecting" a power outage is to simply report an outage every time you power up! (think about it :> ). I.e., require some form of acknowledgement with each power-up event on the assumption that the only "NON-power fail" instances would be those where the user is initially INSTALLING the refrigerator (once it is plugged in, it remains plugged in "forever").
Do the installation instructions tell you to "reset the PF indicator" after plugging in the unit?
If this is the case, then the actual outage could have happened in an unceremonious manner. I.e., no telling how piss-poor the software (and hardware) handled that event! The unit may not have NOTICED the outage as it was happening but, rather, reported it as "reapplication of power" when the power was restored.
There is a subtle but important difference: if the unit was not designed to handle the outage as it was happening, then there is no telling HOW it would misbehave as a consequence. If it was designed to NOTICE the outage and take measures to ensure "what it was doing" was preserved (i.e., was cooling on? or off?) so that it could be RESTORED when the outage was over, then failing to do so (in your case) would be evidence of a bug in that process.
OTOH, if it just uses the PF icon as a means of saying "I've just been turned on and I don't know what I should be doing...", then the software (and hardware) may, in fact, be performing exactly as intended!
[Note that a more prudent set of assumptions -- assuming you can preserve the desired temperature setting -- is to idle for 2-5 minutes on the reapplication of power (resetting this timer on EACH reapplication of power in case the power is flickering on and off, repeatedly) for the pressure in the compressor to equalize. Then, resume the temperature control algorithm]

I've never quite understood why microwaves loose time so quickly! It seems like even the most momentary glitch causes them to lose their minds!
I know there is a lot of cost pressure on their design (35 years ago the electronics for a microwave oven cost < $15; so even a $0.10 component gets a lot of scrutiny!). But, I can't see why their "hang time" would be so bad...

How do you know that this "pushing and holding" didn't turn it off? (i.e., does the first "push and hold" illuminate the indicator while the second actually toggles the setting?)

years ago. It has the ability to play the "back" side of the tape without flipping it (*or* flipping the playback head -- it uses a 6 channel playback head so the "active" part of the head is present over the front AND back tracks simultaneously). But, there is a set of events that can confuse the mechanism into running backwards but *thinking* it is going forwards (!).
I.e., normally, the tape counter starts at 0000, increases to some maximum value reached at the end of "side 1", then starts counting DOWN as it plays "side 2" effectively rewinding the tape in slow motion. But, in certain cases, the tape will play "side 2" (i.e., moving BACKWARDS) while continuing to count *upwards*.
[This makes absolutely no sense in any set of operating conditions!]
I have *two* of these decks so I was able to verify that it is a "feature" that both have.
When I reported this to the manufacturer, they claimed it was the way it was *designed* to operate. I was speechless: "Yes, that is very obviously the way the actual design was *implemented* (as that is how it actually WORKS); but that can't possibly be the way it was INTENTIONALLY designed to operate?!"

Very likely. Computers have memory. How that memory is used is up to the developer. There is an art to designing "reliable" software that doesn't allow memory failures/glitches to alter the operation of the "program". Some "thing" in memory obviously effectively was acting as a gatekeeper to the "change water dispense mode" setting. Perhaps one part of the code *thought* it was in ICE dispensing mode and, as a result, ignored any of the choices available in water dispensing mode...

That's a dodge. You stumbled on something that is "less likely" to be encountered. But, that doesn't make it impossible to verify! Give me a few hours with the schematic and source code listings and I'll tell you how your "bug" can happen!
What typically happens is the testing happens in typical environments. Open door: does light go on? Check! Close door: does light go off? Check! Set temperature. Wait. Is interior cool? Check! Lower temperature. Does compressor start within 20 seconds? Check! Wait. Is interior COOLER? Check!
I.e., does the unit operate the way we THINK folks will be using it? This is very different from "does the unit operate the way it SHOULD, regardless of the combination of unforeseen events that might occur?"
When I test other folks' implementations, I *always* can find something that doesn't work right. Usually, because they never anticipated anyone doing what I've just done (e.g., pressing ON and OFF at the same time).
The conversation ALWAYS follows the form: <gasp> "What did you just do???" "I did ...." "But you're not supposed to *do* that!" "Then DON'T LET ME!!" d'uh...

That's a foolish approach to the problem.
Imagine driving down the road and having a piece of luggage fall out of an aircraft flying overhead, landing on the hood of your car.
Would you keep driving down the road, repeatedly, hoping to REPRODUCE the event (even when there are no aircraft in the sky)?
You have to, instead, ASSUME that what you saw (or heard reported) actually *did* happen -- and then figure out how it COULD HAVE! Typically, the attitude is taken that "it was a fluke" and it is dismissed (because its too hard to chase down!)

ThankYouVeryMuch -- WhereDoISign??
The replacement controller would most probably have the exact same flaw. It may NEVER bother you -- but that would be because the airplane never flew overhead, again! :>

Sort of like two and three blade "razors"...

This is actually a valid way around patents! Identify the "independant claims", then AVOID them by any technicality possible. I.e., if the claim references a screw, use a nail.
The goal of the patent examiner is to restrict the scope of your claims; the goal of your *lawyer* is to broaden them!

Sure it will -- it gets rid of the WP problem.
Of course, it then leaves you with a "refrigerator (lack thereof) problem!"
The problem comes from introducing a new technology to an industry that isn't accustomed to thinking in that context. And, the folks that have "earned" the ability to make the decisions ("experience"), are the most clueless of that technology -- as all of their "valuable" experience was gained in a different world!
AND, they are the LEAST willing to admit how ignorant they are! "The God has clay feet..."

How often do you *really* expect to encounter these types of failures? If "often", then invite the WP folks to come out and see it the NEXT time it happens (i.e., if they can witness it twice, then it's obviously not an "isolated incidence").
If faced with your problem (with MY skillset), I'd look at the actual controller design and make educated guesses as to whether/how it was detecting power failures. Many microcontrollers have brown-out detection hardware built-in. I'd then *assume* it was being used but that the power conditioning circuitry was presenting problems for it (in whatever sort of mains problems you were encountering -- borrowing a power line monitor could help you see if you were encountering frequent, brief outages; long, sustained outages; frequent transients; etc.).
You might be able to play games with the power feeding the controller to make it's (faulty?) algorithm more resilient (e.g., adding bulk filtering to give the controller more "up time" as power fails -- but doing it in a way that doesn't interfere with the power failure detection circuitry!).

*WE* appreciate the heads up as we'll be in that market, soon. Always nice to learn from other folks' experiences -- instead of being on the bleeding edge, yourself!
Good luck!
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 7/9/2016 6:28 PM, Don Y wrote:

As I stated to them, there is no pf icon in this thing. There is one in my WP wall oven, but not the fridge.

Well, I'm sure it turned off something in the controller. The fridge compressor and fans were obviously not on as it was warm in both the fridge a freezer. When I pushed and held the 2 buttons for the required 3 seconds, the icon lit that said "cooling is off".

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 07/09/2016 09:46 AM, Art Todesco wrote:

I have a 24-year-old WP side-by-side that will likely pass away any day now. Hope they still make a model with a simple reliable old-fashioned thermostat.
Why does a refrigerator need an on-board computer to control the temperature? Seems silly.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
"George in SC" wrote in message
On 07/09/2016 09:46 AM, Art Todesco wrote:

I have a 24-year-old WP side-by-side that will likely pass away any day now. Hope they still make a model with a simple reliable old-fashioned thermostat.
Why does a refrigerator need an on-board computer to control the temperature? Seems silly.
Cheap-er to manufacture and looks better
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 7/10/2016 8:52 PM, George in SC wrote:

Why does a phone need a camera?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 07/10/2016 11:34 PM, Don Y wrote:

OMG! U mean they're putting cameras on refrigerators now? So I can take a selfie @ 3am in my pajamas? Will it automatically upload to FaceBuck?
Or does it just document my gluttonous appetite and notify the corporate health coach?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 7/11/2016 2:59 AM, George in SC wrote:

Reread my comment. :>

I think there are actually some that photograph the *contents* of the frig and make that available for display. I.e. peer inside the frig without opening the doors (I guess insulated GLASS hasn't crossed their minds...)
<http://www.digitaltrends.com/home/smarter-introduces-a-fridge-cam-at-ces-2016/
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 7/11/2016 5:59 AM, George in SC wrote:

Wireless connection to NSA, with copy to the health care admin. Reports with time stamp who ate what out of the refrig. You see, Comrade, it's for your own good.
--
.
Christopher A. Young
learn more about Jesus
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 07/11/2016 09:37 AM, Stormin Mormon wrote:

Is it y3k ready?
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 7/11/16 1:16 PM, Corp Ulence wrote:

No, it has been "1984", since, well, 1984 ;-)
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
George in SC wrote :

Yes! It's on 'The Internet of Things' so that you can see how many eggs you have left in the fridge while you are out shopping.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 7/11/2016 1:34 AM, Don Y wrote:

Why does a fish need a bicycle?
--
.
Christopher A. Young
learn more about Jesus
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload
On 7/11/2016 8:35 AM, Stormin Mormon wrote:

It can't jog for exercise.
--
Maggie

Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

It's jealous of the bowtie.
--
_____________________________________________________
Knowledge may be power, but communications is the key
  Click to see the full signature.
Add pictures here
<% if( /^image/.test(type) ){ %>
<% } %>
<%-name%>
Add image file
Upload

Related Threads

    HomeOwnersHub.com is a website for homeowners and building and maintenance pros. It is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.