This fridge came in with several problems. The worst of which was the cabinet being bent into a diamond shape, so that neither the freezer nor the fridge compartments would close properly. I hadn't seen this sort of damage before, so it took some thinking to come up with a way to fix it. In the end, an extended hydraulic ram (porta-power) was easily able to straighten the cabinet. Can't imagine how this amount of bending happened, but it must have been an exciting day for someone.

The defrost problem was with the defrost control. These fridges have a time and temperature based defrost control. The defrost cycle initiates after so many hours of compressor run time. Then, the defrost cycle ends once the evaporator warms up to a certain temperature. I tried my best to get the defrost timer motor working; however the plastic gears had shrunk over time. They no longer mesh properly, and for that reason there is no way to consider that control reliable. 

I know that it is possible to install a "regular" defrost timer in these, however there were some concerns with this.  The timers I have seen all have a minimum operating temperature rating of 32°F; meaning that they are not intended to be mounted inside the freezer compartment like the original GE timer.  Another concern is loss of the temperature controlled defrost termination. General Electric engineered this to be a cost-effective design. If they thought it would be OK without the temperature sensing defrost control, they would  have eliminated that cost from the design.  Without engineering documents it's hard to say for sure, but I expect that there are several reasons for this. One would be the inconsistent nature of the hot gas defrost. The amount of heat available will vary considerably based on how long the system had been running before the defrost cycle started; ambient temperature, and other factors. For that reason, the time necessary to defrost will not be constant.  Also, since the compressor is running in a loop with little restriction, the suction pressure returning to the inlet will be very high. This will put a lot of load on the compressor. I am sure that GE wanted to be sure that the defrost cycle was ended before the return gas temperature was too warm for the compressor, under these operating conditions. 

If there was a happy medium to be found using a fixed defrost cycle time, they would have found it, I expect. But they went to the cost and development time to use a time and temperature based control.

Having said that, the solution I used to replace the existing defrost control emulates the same functionality. There is a timer monitoring compressor run time. After 12 hours, the timer will enter a defrost cycle. The end of the defrost cycle is controlled by a defrost termination thermostat located on the evaporator, in the same location as the original temperature sensing bulb from the original defrost control. The timer and the thermostat work together  (with a relay) to accomplish the same functionality as the original control.  In the video, I explain this and draw a simplified schematic for this system. I do realize that some people will consider this overkill and a waste of time; but as said previously I am only keeping the intent of the original design, as opposed to losing some functionality which the factory had included. 

Sincerely,

David 

Thanks Ralph, for the kind words!  I did combine learnings from my automotive days, and industrial control work I do for my real job now. I guess one could say this was a cross-genre repair LOL! 

Thanks Louis.  I try to preserve as much originality as possible when I repair things. There is a balance to be struck, though. That balance is between true originality at all cost; versus giving the user of the appliance a good ownership experience.  If this were a museum piece, I would have called in favors and located another working defrost control timer motor.  These defrost controls have been out of production for decades, so any found now will have the same age-related deterioration problems. This fridge belongs to a fellow enthusiast who lives out of state, so if anything goes wrong it will be a problem. You know, the usual lack of willing service people and shipping costs and hassles. So, I wanted to use off the shelf parts which can be easily obtained and replaced if something happens; down the road.

Hi John.  I had a feeling you would have some comments on this repair. We have talked before about this system and you have been instrumental in helping me get the original system working on others. Remember one of them had the fan wiring tied together, actually. Thanks to your guidance we were able to locate that previous service error. I am aware of the later repairs with a non-thermal timer being sanctioned by GE; however that doesn't necessarily mean that is optimal under all conditions. 

As for my reasons for using this design as opposed to a fixed defrost timer - there are several.  I'm not under time constraints on these projects. There is satisfaction to me; had in recreating something with all the original functionality. This sort of problem solving is one of the skills I use at my "real job" in industry. Much of the industrial equipment we work on is going on 20+ years old and original parts are not available. Conversely, some is brand-new and still being developed. It's important to ensure that any changes don't alter the control philosophy in a material way.  Small projects like this are a way to keep my problem-solving skills sharp.  

This is along the same lines of why rebuild a belt-drive Frigidaire compressor versus installing a hermetic unit. 

Again it's not that I don't agree with your advice or think it's wrong - this project was done to do things my own way as an exercise in engineering. 

Sincerely,

David

Hi John; thanks for the follow up.  I certainly expect that the simpler, later GE defrost timer is more reliable than the thermal one; for the simple reason that it's not nearly as miniaturized. Had the original been a little more robustly constructed, it might have been more reliable, but it was miniaturized before that concept was fully developed in the world of engineering (other than military stuff.)   I'm hopeful that the solid-state components used on this project will last the life of the fridge. There is no reason for a solid-state defrost timer to go bad, ever - but we all know how that goes with the Chinese made parts. I expect it contains at least one or two electrolytic or tantalum capacitor; which tend to go bad over time. They are also sensitive to temperature, therefore the temperature rating put on the package by Ranco. Assuming the electrolytic capacitors in the timer and one-shot relay are not a perishable part, the only thing carrying any measurable current is the plug in relay. That's a $10 part available at any Grainger or industrial electrical place.

Another modification that I believe GE recommended on all of these designs is to connect the condenser fan in such a way that it runs all the time the compressor runs, whether in defrost or cooling modes. That will protect the compressor in the event that it gets "stuck in defrost" and runs too long. The original design switched the condenser fan off during the defrost cycle. 

Kenny; if the hot-gas defrost solenoid goes out, you could in theory modify the evaporator to use an electrically heated defrost system. However, that would be a whole lot more work than just replacing the solenoid. The hot-gas defrost circuit has a special, dedicated tube which is routed in such a way that it clears the evaporator, and the defrost water drip trough, and the drain line. Whatever you used to replace it would have to warm all the same areas or else the defrost water would freeze in the drain and tray area and pile up in the freezer compartment. The hot gas system is really simple and should work if the solenoid coil is functional.  The solenoid valve assembly is located at the rear of the compressor area in a very accessible area. Removal of the evaporator to make a modification to use electric heat would be a terrible job with many possible failure points. 

Bob, it would seem that the current-day refrigerators are engineered to a very narrow cost margin and want to employ the cheapest design possible. They also have been driven by environmental and safety rules to use absolute minimum refrigerant charge volumes. The added sealed system parts and required refrigerant probably are one significant factor discouraging them from designing a system like this.  Also, the point at which the solenoid switches over to hot-gas mode, there is a rush of return pressure to the compressor suction. This is stressful to the compressor. GE over-engineered the compressors on the earlier designs to a degree where they are able to tolerate brief excursions outside of their normal operating pressure range, and tolerate some liquid refrigerant returning to the compressor during defrost initiation. Most of the newer refrigerators have very compressor designs running with little engineering margin for this sort of "added stress." The manufacturers would have to use a commercial-grade compressor in order to get a similar level of longevity; which would cut deeply into their profit margin, and also result in a fridge which lasts longer than they want it to last. 

Sincerely,

David 

Kenny; completely failing to make any start attempt in spite of bypassing the defrost timer and the thermostat would indicate that possibly the compressor overload breaker is faulty (open and failed to reset); there is some other broken wire in the unit; or the compressor motor has a burned-open winding.

Have you tested that you're getting power at the compressor?

Understand that! Limited tools always is frustrating. Hope to hear more after while.