On August 1st 2010, my $900 seven-year-old home refrigerator died. I spent a few days fussing with it, and repaired it myself for $60 in parts, plus another $30 for some kitchen repairs and some thermometers to monitor it. The other tools I used were not expensive, and otherwise useful for the home repairperson. This is my story. Last updated Apr 7 2011. - Herb Johnson
Aug 1st, Sunday evening: That evening, I smelled something burning, saw some hazy smoke. I thought it was a used computer I just bought and was testing. But my wife noticed the smell and haze was strongest in the kitchen, not the living room where the computer was. Half a century of repair work taught me to identify such odors. The smell was the acrid smell of burnt metal and plastic.
As we tried to find the source, I noticed that the refrigerator internal fan was running, but I could not hear the noise of the compressor (a motor which pumps refrigerant that does the cooling). the 'fridge is in a recessed cabinet and not easily moved out to inspect the back and bottom, where the compressor and coils are.
My wife monitored the fridge through the evening, and sure enough it was slowing getting warmer inside. The fans run and blow air, but all I hear from the compressor area is a "click". This is a 7-year old Sears' Kenmore ColdSpot side-by-side refrigerator/freezer, Model # 106.41212101
Long story short, my wife called Sears and got their 24/7 service center. They confirmed we purchased the 'fridge, confirmed the 1-year warranty was long over. They offered us a deal: either pay $75 for a service call and estimate, or pay $260 for up to $500 in repairs (or credit) and a year's warranty. But - we had to decide BEFORE the serviceperson came in the door, on Thursday, four days away.
"They want us to place a blind bet", said my wife. "Nuts to that", I said. We don't bet without knowing the odds and outcomes. But first, we had to deal with groceries and frozen goods.
We always have a Plan B for appliances - they all fail eventually. By good fortune and lack of effort, the old "basement refrigerator" was still there. I opened it and saw it wasn't moldy for years of non-use, and it was still plugged in. I powered it up, and it simply worked. Not fast, it's half a century old, but working. We moved the critical stuff and frozen meats and sauces in from our tepid Kenmore.
It was getting late, well past midnight, when my wife got on the Internets, to Sears' Web site for parts and repairs. Sears is good for parts on their products, from decades back, I knew that. We got the parts lists and diagrams (no repair manual). But she also checked the "U Fix It" kinds of discussion groups, and came up with discussions of similar kinds of failures. In the meantime I was asleep.
Aug 2nd: The next morning, I reviewed her findings. Discussions of similar failures with similar 'fridges zeroed in on an overload/relay, the parts which immediately control the compressor. For some 'fridges, these parts are intended to fail to protect the expensive compressor. SOme of them burn up - a few catch fire! - and that was also consistent with our experience on Sunday.
My guess from the Sears parts diagrams, also supported this kind of failure. But it may have failed because of a starter capacitor, or worse a compressor motor failure. Sears lists a part called a "start dev" which I think is an assembly including the overload device. Hard to read these parts lists and sketches....These are all $50 or so parts. But the compressor is $300 - too much for a $900 old refrigerator, and beyond my capacity to replace anyway.
I'll probably replace a few of the parts and see if it fixes it. That will take a few days to resolve. Meanwhile, to examine the compressor and to make repairs, I have to get help to move the refrigerator, that eats a day. But there's still time to cancel, or accept, the repairperson and their "bet".
"Refrigerators are not heavy, but they are bulky" said my neighbor, as I told him that afternoon about my situation. I needed another person to pull the refrigerator out of its "nook", because it was mounted six inches off the floor. He couldn't help me, but he usually had someone who worked for him. "He's at the Shore this week, I can't reach him." My other friends were unhelpful - one had a broken forearm.
Between looking for help, I prepared some lumber to "catch" the refrigerator when moved off its pedestal. Stacks of boards of the same height. These would sit on teh lip of a hand truck, so they can be lifted and rolled across the vinyl floor of the kitchen. A 1 X 2 board would be my lever, to lift a corner of the 'fridge and move it along.
It's the same technology they used to build the Pyramids. Levers, blocks and sledges. Only I had no slaves.
Aug 3rd, Tuesday morning: I'd done all I could, without actually looking under the 'fridge. I couldn't find anyone to help me move it out to see. But I was running out of time, so I decided this morning to move it myself, anyway, before my wife left for work. I set up the stacks of lumber on cardboard, and started moving the fridge forward, inch by inch.
Sure enough, the lumber "caught" the front edge of the fridge, and the assembly rolled over the floor without toppling. With several inches out, my wife put a second stack underneath. I moved the whole thing out, with less fuss that I expected. Then, I tipped the refrigerator forward, my wife removed that center stack, to put the back edge on the floor. I tipped it back, my wife pulled out the lumber from the front, and the 'fridge sat on its wheels again, on the kitchen floor, for the first time in seven years. My wife left for work.
I removed the back panel to expose the compressor and wiring. It was dirt simple: a compressor, with some modules mounted on it; a fan; and the dusty evaporator coils. And, the modules were burnt! I could see the soot, smell the material. Problem found, or at least the last problem. The module was designed to fail, to protect the compressor. But did it fail, or the compressor?
The module failed in flames, that's for sure - I removed it in pieces. Parts of it were stuck, welded to the three pins connecting it to the compressor. As other people's notes described, the compressor has a motor with two "windings" or coils; a starting winding, a running winding, and a pin common to both. Controls in one of the modules start the motor by powering both windings; then current is removed from the starting winding. There's also a starting capacitor, another module, in series with the starting winding. That module had soot on it but appeared otherwise undamaged.
Tests for the capacitor was simple: I have a digital meter that measures capacitance (look it up) and it read as it should. Testing the compressor motor is a pass/fail proposition: there should be several ohms resistance through both coils in series; the one with lower resistance is the starting winding. Sure enough, I measured about 9 ohms across two pins, and smaller resistances to a third pin, about 4 ohms and 5 ohms for each. That was the "pass" test which told me the coils were not open. I'd have to run the refrigerator to see if the motor still worked, and the compressor still did its job.
Aug 3th, afternoon:I knew roughly what was wrong, I did not know exactly which parts were needed, by name, or how to get them fast. Sears at least gave me some clues.
Sears provides on the Web drawings of parts, with numbers linking them to names, and names linked to part numbers and prices and availability. Sears has parts stores locally, or ordering by Web and shipping. Pictures of parts let me select some obvious possibilities; I could read numbers off some parts that weren't destroyed. Prices were $30-$60 per part. The compressor - a sealed unit, with refrigerant running through it - cost $300 but I could never remove or install one, anyway. So, buying that was out.
But the Sears Web page also provided a photo list of "parts other customers have bought" which led me to an assembly of parts called "start dev" which seemed to cover most of the parts destroyed. It's almost as if they made a repair kit for just this problem....the price was under $60.
I could order online, or check if the parts were available locally. From my ZIP code, they located the nearest Sears parts store - about 8 miles away, across the Delaware River in Pennsylvania from my New Jersey home. It's near a regional shopping center I use occasionally. Google even showed me the store - their "street view" had a grainy photo.
But first I had to check with the accountant. I called my wife and reviewed what I found, and prices for the parts most likely needed. I couldn't know EXACTLY which parts to buy, until I saw them at the store. She asked "what's the chance these will fix the refrigerator, AND fix it safely?". That's a reasonable concern and I said this. "My tests suggest that the compressor is not fried outright, and that a few of these parts will cost under $100, $150. If they fix it, I can monitor the fridge and, with time, confirm it will be safe to use. But if the compressor fails when I repair it, it's not worth further time or money.
I called ahead to verify the store was open and had the parts I THOUGHT I'd need. Google Maps gave me the route. I was there in 20 minutes and had the parts in 10 minutes more. Examination of the "start dev" kit showed it had all the parts I needed. In fact, it had two sets of parts, for different revisions of the compressor! I did not buy the capacitor, if mine was bad I'd have to come back, but I saved $40.
I actually had time to check the neighborhood - found a "Tools 4 U" used tool store across the street, and checked them out. Got a tour from the owner, I think he wanted to sell me a franchise! I got home with the parts (no new tools), but that was only the beginning of my repair work.
Aug 3th, evening: That evening, I had to install the new module onto the compressor. Problem was, the old parts were practically WELDED onto it! I could see the new module had three sockets, for three simple straight pins from the compressor. But my compressor had pieces of metal on the pins. I thought they were simply soldered on - but I used a "soldering gun" with a very hot tip, and could not un-solder them. I actually spent hours filing off the pins even using a Dremel grinding wheel at the end (my wife's suggestion, that's why she makes the big bucks). It took FOUR HOURS with breaks to clean those three contact pins!
By midnight I was ready to attach the module, the capacitor, and then start up the 'fridge. It was a difficult fit, the pins were bent and still rough. There's a kind of "mechanics feel" in installing any part: when I "felt" the fit was acceptable I wriggle and pushed the module in place, and installed the capacitor and power cable.
Now, I still had to clean the coils - they were choked with "dust bunnies". The coils are arranged in a series of "W's", along the bottom of the refrigerator, kind of like waves in the ocean. (Here's how a refrigerator works, if you care to know.) With the coils choked with dust, there's no air flow, the 'fridge can't get rid of heat and cooks itself.
But, just to see if things would "work", I could start it up. At 11:30PM, we plugged it in and turned the control on. I watched from the back and told my wife, "Go ahead!" as my wife turned the knob to set the temperature. It "kicked" the motor on, and I could hear it "purr". I asked for the flashlight, and looked on either side of the compressor. "No sparks....Still have to clean the grate...pretty dirty...." But it ran for a few minutes before I turned it off: I wasn't finished yet.
I had to clean seven years of dust before running it for use. But most of the coils are under the fridge, inaccessible. I vacuumed the front and back of them, but I found that an artist's brush (with bristles about 1/2 inch round) did a better job of "grabbing" the dust and loosening it up to be sucked out. Even so, I had to tip back the fridge so my wife could crawl under it to clean the rest. She insisted on a "block" to stop the fridge if it got loose - no dummy, she! Vacuuming and brushing took most of an hour.
And, I wanted to set up some instrumentation. I have a "Kill-O-Watt" device, that measure appliance AC current, voltage, and power (even phase angle, look it up) over time. Anyone who wants to know how much power is used by their washer, refrigerator, toaster, and so on has got to get one of these. They cost maybe $30-$40 bucks. And I have digital thermometers, so I can see how cold the 'fridge will get. I can monitor the current to tell how the compressor is running; my ears will tell me when it's running. And my nose will tell me if it smokes.
I set those up, ran the fridge again, and took notes at intervals. An hour's run time told me the fridge was cooling down. When it stopped, that told me the controller was working, it saw the temperature was cold enough (about 35 degrees as I measured it). I watched it run for a few cycles, off and on, saw that it would not explode. Then I told my wife to check the fridge and turn it off when SHE went to bed, and I gave up for the day.
But we agreed that I'd call Sears tomorrow, and cancel the repair call. Either it was fixed, or it would prove uneconomic to repair - no blind "bet" was needed.
Aug 4th: The next morning I got up and read my wife's notes. All was well, she wrote, with some details, before she shut it down. I powered it up and monitored it as I had breakfast and did my morning's work from home. Things looked good: it ran for about 14 minutes, stopped for 10 minutes, as the temperature varied over about four-five degrees. The current "draw" for the compressor was about 1.2 to 1.4 amps. And it did not sound bad (pretty quiet in fact).
The temperature of the compressor worried me: I have an IR remote thermometer (you point it at something and it tells you its temperature) and it said 130 degrees! But it did not get hotter, and it cycled between 110 and 130 - that could be "normal".
Only time would tell, so for the rest of the day I watched and took notes. Up and down, off and on. The good news was that the compressor slowly got COOLER over the hours, not hotter. That told me the heat was from cooling the 'fridge, not from the compressor. By early evening, I concluded I "understood" the operational cycle of this empty fridge, and it passed the "likely OK" test.
Fix the alcove, too
We still had two loose ends. One, the fridge was too hard to move in and out of that alcove, lifting it or dropping it six inches. Two, what if it smoked again - how would we know? The first issue was resolved this same day: I sawed out the platform in that alcove, covered the floor under it with flooring (to raise it to the same height as our vinyl floor). A trip to the hardware store got a five-dollar piece of that. It was some sweaty work to unscrew and chop out some wood, and to cut up the underlayment to fit the space.
The second issue was also solved at the same time: a kitchen smoke detector, battery powered. Mounted inside the alcove, it would most likely not see cooking smoke. And it said it was resistant to such things. I also got a THIRD digital thermometer, to monitor the surface of the compressor with its remote sensor. It read cooler than the IR thermometer, but it did not make good contact with the case. That's OK, it's a kind of "fire detector" - placed near the new module, if the sensor burns up with the module, the temperature will go wild, and never be normal again.
Aug 4th: That evening the alcove work was finished, and shortly before my wife got home. She saw the 'fridge back in place when she came in the door, and that made her whole day. I explained all the monitoring devices now in place, showed her the smoke detector before I installed it. She said it was OK.
Aug 5th We agreed that the next day, we could move our food from the basement fridge and the college fridge. And, the refrigerator could now be rolled out for cleaning and inspection - and replacement, the next time it fails - without risk of falling on me. I also used part of that day to clean the fridge before putting the food back. My wife was surprised that evening with the results.
- Herb Johnson
As of April 2011, the repaired Sears Kenmore ColdSpot is continuing to run, but it has "issues" at times. I continue to monitor the compressor case temperature: it now cycles between about 85 and 95 degrees F. But on occasion the inside temp of the refrigerator compartment rises to the 40's and 50's.
On one occasion, the problem was a blocked vent. Cold air enters the refrigerator side from the freezer through a small vent. Moving the item blocking the vent solved that problem. There's also seasonal changes: when the house is warmer in the summer, I need to adjust the refrigerator thermostat.
When temps seem too high, I clean the coils underneath, either with a vacuum or with blown air. There's poor air flow under the 'fridge, no easy way to clean the coils. The small fan moves air slowly. I've considered adding a secondary fan.
In April 2011, one morning I thought my cold cereal was not cold enough. Sure enough, I measured temps in the fridge in the high 40's. I tweaked the control for "cooler", but the temps did not go down. Yet the compressor was cycling, and not overheating. Cleaning the coils did not do much.
Looking inside the refrigerator compartment, I identified the cold air vent, and there was no blockage. I disassembled the covers and discovered how the refrigerator side was controlled. There's a mechanical linkages (the metal wire in the photo) that opens a flap, over a powered air vent from the freezer. To bring more freezer air into the refrigerator side, the knob adjusts the flat. However, this control is identified as "FREEZER" and is marked "COLDER" when the flap is CLOSED. That's because when less cold air is routed from the freezer, it lets the freezer get "colder".
So I re-labled the control to describe its function - to route air from the freezer. As a "cold air to refrig" control, it allows "more" or "less" air intot the refrigerator compartment. Previousl set to the center, I"ve now turned it to the left (more air = warmer freezer). If the freezer gets too warm, I'll increase the "Refrigerator" setting to "colder" - which cools both the freezer AND the refrigerator compartments, in that order.
Even an hour later, the refrigerator compartment is down to 42 degrees F. The panel says "wait 24 hours between adjustments. In other words, it's underpowered. That's OK.
For those of you who think this is a lot of fuss....replacing this refrigerator would cost $500-$700, and the replacement has to fit both the width of the cabinet space, and "fit" my wife's requirements.
P.S. I told my wife later that morning about the refrigerator and my repairs. She said "well, you know I put my home-made spagetti sauce into the 'fridge last night". You mean that gallon of sauce you worked on all night?. Did you put it in steaming hot? "Well, sure...".
- Herb Johnson
Here's how a refrigerator works, if you care to know. There's a compressor, that pumps refrigerant around. There's some condenser coils to cool the compressed refrigerant, an evaporator device or valve, some evaporator coils that chill the refrigerator contents. A fan blows air across some coils to cool them. Compressed refrigerant is hot, expanded refrigerant is cold, it goes around and around.
Condensor coils dissipate the heat from the compressed refrigerant, cooling it down before it enters the expansion coils. The fan pulls room air at right angles across the W-shaped "waves" of condensor coils. An expansion valve expands the liquid to a gas, chilling it. That gas in the expansion coils chills the contents of the 'fridge, with another fan. That cooling then warms up the refrigerant before it enters the compressor. The compressor compresses the gas to a fluid, heating it up. It loses that heat plus the refrigerator's heat, as I just described, round and round the loop.
When the 'fridge is cool enough, the compressor stops; when it warms up, it starts. That's thanks to a controller with a temperature sensor. There's some electrical heating coils to defrost the freezer, they run at intervals for your "frost free" refrigerator.
Original text about 3100 words, 3 hours for 1st draft. Copyright © 2011 Herb Johnson