Amana SBI20JW refrigerator (defrost?) problem

I've just done overdue cleaning of the unit, but a recent problem that seems intermittent continues. For the most part it seems to be operating just fine, both fans & compressor running, no undue ice buildup anywhere. Once in a while though I notice that the temperature, particularly in the freezer, spikes leading to partial thawing of less dense foods and the ice in the bin. I'm beginning to suspect the defrost timer, which I've not yet been able to locate. It just went through a defrost cycle with only lightly frosted over evaporator coils. This resulted in ~1/2 cup of water in the catch basin above the condenser. The defrost cycle though lasted fully 1/2 hour; is this approximately normal or too long? As soon as I noticed the unit had shut off and water was dripping into the basin I adjusted the thermostat to be sure it would come back on when the cycle completed. If 1/2 hour isn't too long maybe I should look at the terminating t'stat and try to confirm that it's not stuck in the closed position...? Any other suggestions?
Thx, -joe

 

How do you know it is defrosting for 1/2 hour? Just from the fact you hear when the compressor shuts off and then again when it comes back on?

Maybe the defrost element itself is only actually heating for half that time. But when it comes out of defrost mode, there still is no call for cold as the stat is likely in the fridge section.

I like to use the Kill-A-Watt plug-in watt/amp/voltmeter to see what is actually going on. When in defrost you can expect something in the neighborhood of 450 watts, give or take a little. The meter is digital and resembles size off cigarette pack or voltmeter or lamp timer device. You plug it in any outlet and plug any appliance cord into it to find out your draw. I believe Harbor Freight sells them, reasonable.

Location of defrost timer may be encased in same plastic housing that main control thermostat is. If not located at front or rear bottom of chasis.

Terminating t-stat? You speaking of the same stat I just refered to(that sets fridge temp?), - or the defrost stat? If the latter, that would not be able to lengthen the defrost time, but it could shorten the time the element is on.

If you have a newer style fridge that has digital control and control board, and thermistor or ? -you could have some other problem. I mainly work on the older or basic bottom-line ones without all the bells and whistles to complicate things, in rentals.

Sometimes in refrigeration, sporadic warming of freezer unit is due to being marginally low on freon. One minute all appears fine, but then it may start to warm up - then get cold again. In such a case though the situation gets worse, and you find the compressor trying to run more and more and getting overheated.

 

Thx for the response. I certainly don't know at this point that the heating element was on for that entire half hour. What I was wondering about was whether 1/2 hour is normal for a defrost cycle that's controlled by a timer. The terminating thermostat may well have shut off the heating element before the cycle was complete. That particular thermostat as I understand it is common to most units and sits adjacent to the evaporator coils and opens when temperature rises beyond a set point, killing current to the heating element but not terminating the cycle.

By 'main thermostat', I assume you mean the device you adjust to control the temperature inside the unit, also known as the 'cold control'. I assumed that because I turned the cold control from its normal setting up to near maximum while it was in the defrost cycle that it would start back up pretty much as soon as the cycle terminated, not necessarily when the heating unit; anything wrong with this logic?

The Kill-A-Watt sounds like a good idea. I haven't used my ammeter because it's not very convenient, especially as I've yet to locate the defrost timer, not for want of looking I might add. And the possibility that it's a smidge low on freon has also occurred to me but the jury's still out on that.

Thx again for the response. -joe

 

Nothing wrong with 1/2 an hour defrost time. The defrost cycle will not cause the freezer temp to spike. The best place to set the cold control is at the midrange or slighly colder. A lightly frosted evaporator coil is normal, so it sounds like the defrosting cycle is ok. If the defrost cycle was in trouble, the freezer coil would be frozen over. My suspicion is that there are times when the compressor is not starting or is overheating which would account for the freezer temp spiking. If the ice cubes in the ice bin are turning to water, the temp in the freezer is rising to at least about 32 degrees. It could be because the coils underneath were dirty, it could be there are times when the fan motor underneath poops out which could have been related to excess heat coz of the dirty coils; translation- the excess heat would overheat the lube in the fan motor bearings. Another possibilty is the cold control is intermittent. I wish I could give you a more definite answer; next time you see the temp rising quckly check and see if the compressor and fan underneath are running. If the fan is runing but the compressor is not, the cold control is probably ok. Take off the grille in front below the doors, the defrost timer should be in there.

 

Thx. Is the cold control at all a common failure? The unit is pushing 20 yrs old, but until now has been trouble free. All seemed well for a few days after I cleaned the coils but then the temp spiked again. Is an overheating compressor sign of low refrigerant charge? -joe

 

I doubt you have a refrigerant charge problem. All of the refrigerant would have leaked out by now. Cold controls do fail, 20 years of on/off wears them out. The trick is to catch it rather than just replace parts. The part # for the cold control is R0161092

ebay $30
Sears $38
repair clinic $45

 

Thx, John. When I got up this morning the unit seemed to be running normally, but only showed 10F in the freezer with the cold control set a bit colder than the midpoint on the dial. I let it run for a while and then pulled the inner cover to expose the evaporator to find essentially no frost on the coils. So I pulled it out from the wall and took off the back cover. The fan was running and the compressor was fairly hot to the touch as was one of the 2 refrigerant lines. Put it back together and pulled the inner cover to find then a light frost on the coils. It had been running for quite some time before I first exposed the evaporator and there was no water in the drain pan so I don't believe it had just completed a defrost cycle. Since I last checked the coils, about an hour ago, freezer temp seems to be slowly heading down, having dropped a good 5F in that time. It's been running constantly all this time. Any further clues/suggestions? -joe

 

Well it is 20 years old, so it won't run like a new one. Light frost is normal. Frankly, when the coils go underneath go a long time between cleanings [like a few years] that's tough on compressors. Dropping the temp 5 degrees in an hour is a good thing. I think the best thing to do is monitor and see if you can catch it when the temp spikes. At this juncture, I'm leaning towards an intermittent cold control, but I'm loathe to just change parts without proof.

 

You could place a thermometer in the fridge section, where you can easily read the numbers by opening the door only for like 2 seconds, (to keep from raising the temp in the fridge section) See if you find the temp conclusively higher than what it should be, according to where you have the cold control set.

You could then use an ammeter to clamp around one of the wires leading to the compressor relay box. Between that and a Kill-A-Watt meter plugged in to the outlet (with fridge plugged in to it), you could conclude several things about if fridge is in defrost cycle (without opening up the back of the freezer), or if compressor is out on thermal overload, and if the cold control is working or not (i.e., if power is going through the cold control when it should be calling for cooling, as opposed to it malfunctioning and not allowing the current to flow through it, you would show amps in the wires leading to the compressor relay box, even if the compressor was off) - before suspecting marginally low freon.

 

Temperature in the freezer is all over the lot. It was in shutdown mode a while ago at something over 10F. When it came on it dropped to well below 0F fairly quickly and is still running. I'm going to throw a part at it and see what happens. I'll order a new cold control tomorrow and plug it in as soon as it arrives. We like the unit and even if it does unexpectedly need a new compressor I'm reasonably sure that'll cost a good deal less than a whole new machine. I'm also fairly well convinced that the defrost is functioning properly. I'll post the result when the new cold control has a few miles on it. Thx again for all the help! -joe

 

Sounds good. My gut feeling is erratic cold control as well. Thanks.

 

That be great if you do, as that helps everybody.

 

Jury's not quite back in yet but it's looking like we've found the culprit. I found a new cold control locally & installed it early this afternoon. Buttoned everything back up and let her rip. Freezer temp has stabilized in the vicinity of 0F and seems to be holding well with the dial slightly below the mid point on the scale. I'm a late bloomer at this sort of thing but with your kind assistance, my first attempt is beginning to look like a success. I'll keep the fingers crossed for a bit longer but right now the signs are good. Thanks so much for all your help! -joe

 

Thanks for the update. Would you be willing to share what the parts cost you? It helps others with their decisions.

 

Sure. Bottom line with tax, over 9% now here in the PRK, was $51.xx, With shipping that's about what it would have cost from Sears.

Temp spiked again early this afternoon, but came right back down. I'd maybe - hopefully - turned the new cold control down just a tad too far. Compressor seems to run as long as needed to bring temps down, except for the occasional timeout for defrost. If it spikes again I'll spring for a Kill-A-Watt and monitor power usage more closely. -joe

 

how high did the temp spike?

 

It was up near 32F, but an hour or so later had dropped back below 10F. Might've been more than an hour I suppose, but it didn't seem like long. I was distracted by another project so wasn't monitoring as closely as I might have. -joe

 

This is the part of the post (see quote above) that has scared me as I first read this thread, and became suspicious of a marginal leak, possibly causing pockets of refrigerant that can compress one time and not at others.

The fact Joe had it running for awhile, yet no frost.

On some models where you can easily get to all the condensor coils(the hot coils), it's also a tip off if the entire coil circuit never gets fairly uniformally warm and only the first number of feet of the coils do.

 

I've wondered about that as well, but when it's working it does seem to cool very well, and I don't think there were any cold spots around the perimeter of the condenser. Poring over the wiring diagram on the back of the box, there appears to be a relay that controls compressor function; aren't relays sometimes subject to intermittent failure? Looks like access is tight, but I do find a relay for this model at one of the online vendors. Also a capacitor, although I don't yet see where that might fit into the equation. -joe

 

Yes, the relays can be intermittent. If the capacitor is bad it could cause the compressor motor to have trouble starting and running. It's usually the relay that poops out.

 

Thx! I just finished removing and reinstalling the relay & condenser just to make sure there aren't any major gotchas involved. If the symptom, intermittent temp spikes, continues to present do you think it a good bet to throw another $50-60 at it and replace either or both of these? How about the overload? It's mounted to the compressor right next to the ptc relay; I left it alone this time as the spring retainer had me slightly baffled. Back is buttoned back up and unit is running again at the moment. -joe

 

It depends on how much you like the fridge. The relay and overload are usually replaced together. Reason being- a marginal relay stresses the overload. I found the relay/overload kit at Appliance Parts from RepairClinic.com for like $61 and some change which is high. If you look at their parts pics you see it's a part # 12002782, 12002783 or 12002784. I found the 82 and 84 on ebay for $40 and the 83 for $58. The problem is which one fits your box? A phone call to a local parts store or an e-mail to an ebay vendor should answer that question. I'm loathe to just throw parts at an appliance but, if a relay/overload fixes it, you're done and your total outlay is a lot less then a new fridge. Otherwise, it would have to be a worn compressor. In that case, run it until it dies or budget a replacement.

 

Just giving this thread a bump to revive it. The beast is still running but not perfect. I'm just not ready to panic yet. Does anybody have a clue as to about how many kwh/yr a 20 year old 20cuft side by side might have drawn when new? I finally picked up a Kill A Watt at Costco and it looks as though it's pulling close to 1500. thx, -joe

 

I know many of our 18 cu's are about 150watts an hour. And that is when it is cooling. When defrosting, it might be about 450.

 

Thx. It's pulling ~220 when it's running, but I'm thinking a yearly rate might be more useful in trying to determine condition. The Kill A Watt is telling me it's using something like 3 times more juice annually than a current comparably sized energy star should pull. And it's not like ambient here is anything like sweltering. The thing just seems to spend a lot of time running. -joe

 

Anybody got an idea as to why it would be pulling 29W when it's off? Lights aren't stuck on & ice maker is off.

 

No. Can't recall how fancy or modern a unit you have. On the plane jane non-digital ones from the 90's and earlier, I have had my Kill-A-Watt plugged in, and it is 0, until I open the door, and with the light on it reads 39 or 40.

Either you have a short, or somethign based on some bells and whistles feature is pulling power all the time. Then again, this could be because of your problem with yur light circuit/board?. Maybe that is all the current it needs to sap some relay on the board, making it so the firidge can't turn on, with the bulb screwed in.

Unscrew the bulb and see if it still reads 29.

 

Thx. It a 20 year old side by side with no ice or water dispensers. So it's pretty basic. The Kill-A-Watt jumps by ~40w when I open either door whether it's running or not. The bulbs are indeed 40w so that much at least makes sense. Issue several months ago was that the temp would spike once in a while pretty randomly. At the time I wound up replacing the cold control and at first though I'd fixed it, but no such luck. It doesn't do it very often or on any schedule I've managed to uncover. There don't seem to be a lot of choices out there right now for counter depth French door counter depth models without ice&water in the door which is what we'd prefer so I'm still on the fence so far as replacement's concerned.

 

But what is with your 29 watt question? You did not address that in the last post. I wanted to know if that still would remain 29 watt whether or not the bulb was in or screwed out. And how now did it get to be 40?

Or, is what you mean by jumping 40 watt, is that you get a reading of 69 watts with the door open? Still, unscrew the bulb and see.

 

The refrigerator and freezer each have a single 40w bulb and the K-a-w jumps by approximately that amount whenever I open either door. Doesn't that confirm that the lights are switching one & off as they should?

 

More than likely, yes.

But is the reading 69? (29 +40) with the bulb in and door open? What is the watt reading with the bulb unscrewed and door open or closed?

 

It's running at the moment, reading 223w. When I opened the door it jumped to 264, the dropped back to 223 when I unscrewed the bulb. Remains there with the door closed. I'll check with the bulb out when it shuts off.