Used this forum quite a bit to help diagnose a problem with a GSS20IE.. I've never worked on a fridge before.. Know the basics, but with the main board and all that.. Figured i'd search around for some tips.. Found a whole lot here, and wanted to say thanks and put a warning about dealing with some repair shops.

First day it's noticed, I go look at it.. Now, this is a relative of mine, and for some reason they think if you know how to fix a computer, a refrigerator should be no problem. Along with a central AC unit, and a dishwasher.. I fixed the AC unit they hit with a weed whacker and the dishwasher that had a pump go bad, so.. Maybe they're right.

Got digging into it, found the heater resistance to be 28 ohms. Tells me that's good.. Everything else on the fridge is working, so I really doubt it is the main board.. There's no good way to test that, other than "Well, i've replaced everything else", and since it's about $100, I really would prefer not to deal with that as a "maybe" fix.. Rather know for certain. I figure, if everything else is working, the odds of that board being bad are very low. This is where I come here.. Start reading and...

What's left? Thermistor and thermostat.. Pull it the rest of the way apart, and wedge my shoulders in and set my meter to continuity test so that I don't have to stare at the dang thing while i'm stuck in the freezer section. Seriously.. You techs that do this all the time.. I mean, what's up with that? I'm not a huge guy.. 5'10" and 200lbs, and I was acting like a contortionist to get in there. Yes, it would have been smarter to measure from the back, but I like to be certain i'm measuring the right thing. Finally get the thermostat unplugged and find the wires on the plug.. Stick the meter in there.. No continuity.. Aha.. I've found the problem. Head off to my local parts dealer. Tell them I need a thermostat for the model fridge I have..

Here's where the problem comes in... They don't have it in stock. Guy looks at me and says "That's not the problem. It's the heater.." I tell him my troubleshooting. He shakes his head, says "No.. We order the heaters in lots of 30, we order the thermostats one at a time.. It's the heater".. Tell him I ohmed it out.. 28 ohms is pretty standard for a coil.. I mean, it's just a shorted piece of wire basically.. I even start thinking that MAYBE that's a little high for this type of coil, so he's planted the seed of doubt in my mind. There was problem number 3.. (Problem 1 was the defrost didn't work, problem 2 was I went to this asshat for the part).. He tells me that the themostat cuts out at 140 degrees. Now.. I hear this.. I think he's on crack. What kind of idiot puts a thermostat with a 140 degree cutout in a friggin freezer? (The answer to that is GE Engineers).. Then he pulls out the "I've got 25 years experience.. I know what i'm talking about"... Ok.. THAT'S when I should have walked away.. But... Ok.. He's got the heater in stock.. I figure, worst case, I replace a $20 part that doesn't need to be replaced but is problematic.. Best case is I get to go back and call him an asshat.

Well.. Best case scenario.. I went back.. Had words with him.. He didn't even ORDER the part I needed, after saying he would, so said it would be another 2 days... No. Arrogant butthead. I went to another shop, which I should have done long before (They weren't open on Labor day the first time) and picked up a thermostat. Got back, broke out the hairdryer because the thermostat was just encased in frost.. I couldn't even find it without getting frostbite.

Blowdry the whole deal, clean up the water.. Find the thermostat, and... Something looks odd... Pull it off and get a good look at it.. And I see the attached pic.. (Hopefully it attaches) If not.. You can see it at http://tinyurl.com/2lg4dm

I'm not positive, but I don't think the back of that thing is supposed to be swelled out like that.. I cut it out, hook my meter up to it.. Measures out at 56k ohms. Or, about 56k ohms higher than it should. Look at the new part.. Hey.. No swelling on it.. Ohms out to 0.1 ohms. Pop it in.. Now, just waiting to verify it goes through a defrost cycle properly.. Did pick up a thermistor just in case. I shorted out the heater on the back, and it does come on.. So, we know without question the thermostat and heater are good and installed properly.

Now.. I'll update whether that was the only problem or not.. But I do have a few questions..

1) How do you pros replace that if it goes bad? I just cut it as close to the thermostat as I could, soldered the new onto the existing wiring, and electrical taped the snot out of it. I saw no other way to really do it, without breaking out molex kits to put new pins on the end and put it in the connector.

2) Is the guy who sold me the part right? I mean, he was convincing enough to sucker me into a heater, but is that the main point of failure on these units? With a 30-to-1 ratio?

3) How long would it take this thing to go through a defrost cycle? Obviously, I'm not a TOTAL idiot, because I unplugged it to replace the thermostat.. From a power-on standpoint.. How long would it take before it went into a defrost cycle? Any ideas?

4) Anyone actually seen this problem with a thermostat before? My best guess is that at some point water got into the casing, and then froze.. Thus, pushing the 'seal' out from the case.

Last.. For the pros around here.. Don't use the "I've got xx years of experience" line... Be polite about it. Every one of you reading this should know that I did a pretty decent job of diagnosing a bad thermostat. Pushing someone to buy a 2.5x higher cost part that they tested to show good (Which with xx years of experience you should know 28 ohms is good) is not very nice, and will likely have someone very angry coming back to throw a good heater at you. And those things are glass.. You don't want that. At the very least, don't do what this guy did. My only guess is that he took me to be a total moron who couldn't operate a multimeter if my life depended on it.

Good you found the problem, that thermostat is just a safety in case the heater gets stuck on if the board goes bad, to keep the plastic in the frig from melting...I have never heard of one of these going bad, usually it is the heater(s), the thermister or the motherboard....it will go into defrost when the board is satisfied a certain number of door openings etc, have been met, it will be hard to know for sure how soon this will be...

I'm not convinced the thermistor is good.. That's why I got one.. But any ballpark figure on when it would defrost? Is it just based on door openings? Because if so.. Wouldn't that mean it could take months between defrost cycles? I mean, if it takes 100 door openings to go into defrost.. And it was my grandma's freezer.. That could literally take 2 months.

Luckily.. This is my cousins, and with a 6 year old and a 3 year old.. It's pretty well guaranteed that it'll hit a cycle in 5 or 10 minutes if it's based solely on door openings.

Not sure, seems I have read where the first time the frig is powered up it goes into defrost in 4 hours, then it seems like it is around every 20 hours or so, but again I cannot remember, if I find that post again I will find out for sure...

Cheked it yesterday evening, after it had been running for 24 hours, and there was a significant reduction in the amount of frost on the evaporator.. Wasn't 'frost free', but it was about half of what it was.. Considering I didn't 'manually' (Meaning with a hair dryer) defrost the thing, that's good.. And when I put it all back together tonight, i'll blow dry it to be sure it starts out with no frost.

So.. All told, looks like just the thermostat was the problem. Some folks calls it a bi-metal, I calls it a thermostat.