[MattBK]

Fridge is a GE GDS20SCSSS (~11 years old)

Came home from a trip and fridge was totally warm. The condenser fan was not spinning. I put a table top fan behind the fridge and this helped the fridge get slightly-cool. I replaced the fan motor. 24 hours later, fridge still not cool - not even slightly cool, just warm. The condenser coils were a little dusty but not caked on like I've seen in some videos so I don't think that's the main issue.

So I took another look, the new fan was running, but compressor was not. While I was just looking at it (did not turn power on/off), the compressor made a click sound, did not turn on. Made another click 5 seconds later, did not turn on. Then about 5 minutes later the compressor turned on. Is there a way to tell if it's a refrigerant problem or mechanical? Or if compressor is turning off from being overheated?

While the compressor is running, the air blowing inside the fridge is slightly cool. Not warm and not COLD.

How can I tell if it's a compressor relay problem or a I need to recharge the refrigerant?

 

[HotRodx10]

If the compressor runs, but it doesn't get cool, it's likely the compressor is done for; it's possible it's lost refrigerant, but not likely. If it's not a really expensive fridge, it would likely cost more to fix it than it's worth. You can have a tech look at, but my opinion is that is likely a waste of money.

 

[MattBK]

More info: The cold line coming out of the compressor feels pretty cold to the touch and has some slight condensation. The compressor itself feels pretty hot. The condenser coils feel warm.

 

[HotRodx10]

The line coming out of the compressor (to the condenser coils on the outside) should be hot, and the evaporator coils (on the inside of the freezer) should be cold. The return line from the evaporator coils to the compressor will be cool to cold.

 

[MattBK]

Decided to peek at the freezer evap coils. There is ice over the cold line but the coils do not feel cold. Photo:

 

[HotRodx10]

Most likely that means the compressor is worn out and not compressing the refrigerant hardly at all. I believe the 'wide spot' in the cold line is the expansion valve (high pressure, 'hot' side above, and low pressure 'cold' side below). If it's only getting cold right at the valve, the refrigerant is barely getting compressed, so there's virtually no pressure differential being generated. No pressure differential=no temperature differential.

 

[MattBK]

I let it run a little longer and the frost made it around the bend.

 

[Old Thomas]

I would buy a new fridge today.

 

[surferdude2]

If it was mine, I'd find the leak and fix it but I have the means. It looks like the cap tube or liquid line is leaking where it exits or enters the filter/drier but the leak may be somewhere else. Soap bubble it to see. Might be a little oil at the leak site also, that's a useful clue.

It's obviously low on refrigerant, any tech will know that right away. Even so, paying someone to fix it may not be worth it. If it is indeed the simple leak possibility that I mentioned, it will cost you roughly $300 +- I think that's a 20 ft³ bottom freezer drawer so they go for around $1000 new. Your money and your choice at this point. But wait, there's more! There's the overload clicking... that's not normal unless you just happened to shut it down and tried to restart it without waiting 5 minutes or so to allow the head pressure to bleed down to the low side. If it's going off on overload from a cold start, you have a second problem. The overload might be defective (not a very good chance) or the start capacitor could be defective (possible). Then you mentioned that the condenser fan was not running which could mean you need to replace it.

With all that baggage together, I'm inclined to agree with @Old Thomas since paying a tech to troubleshoot and fix all those issues would far exceed the value of an 11 year old frig/freezer.

If you aren't ready to buy a new one, you can buy a used one that works for cheaper than fixing this one.

Sorry.

 

[MattBK]

The fan motor could have been an easy DIY fix but decided to pay someone to fix it because I figured they have the motor on hand and it would be quicker and they would diagnose any other problems. I was wrong on all fronts there so now I'm tackling it myself after paying them $250 to do the motor.

A quick search shows that this type/size fridge costs about $1500 before tax/installation. So even if I pay a few hundred more to fix it, I rather do that. I'm in NYC where buying a used appliance would be a real pain to locate/move myself.

I tested the relay: 5 ohms. Test the overload: 0.5ohms. Could not test the capacitor with my check multimeter but the compressor starts so it's fine?

Should I try a DIY recharge of refrigerant? There are plenty of tutorials for this on YouTube and looks like it would cost me about $100.

In the meantime, I'll try to find the leak.

 

[surferdude2]

It'll be harder than you think to pull off a successful repair with less than the necessary tools. Depending where the leak is can make it nearly impossible for an inexperienced person to keep from screwing things up worse than they are. You'd have to use the low temperature solder that's made for the purpose rather than silfos to reduce the possibility of creating scale and plugging the tiny cap tube. With that a Bernzomatic torch can do the job and mimimize the damage.

It's best to use the process tube to recharge the system rather than a saddle valve. It'd be nice if you had a couple of pinch off tools but I have used vice grips in an emergency and made it work very well to isolate a section for repair without having to purge the system. But then I had an oxy/acetylene torch to heat the tubing back up and reform it enough to be usable and I wasn't working on a cap tube system.

I can't tell you to blow the system charge since it's not a good thing to do unless you may decide for yourself how much charge is left in there and whether it is reasonable to conclude that it's already dumped for all practical purposes. If you had a proper empty jug to use for withdrawing the charge, you could warm the system up to room temperature or higher and connect a jug that has been in a freezer long enough to get below zero. The charge will migrate to the jug. I've done it with a jug in slushy ice in a tub but it's a lot of trouble and your time has to be worth nothing to you.

Then you'll find that a jug of gas may cost a little more than you like but that will depend on what's in there now. A standard jug is 15 lbs. so you'll have enough left to go into business.

Good luck... post back if you need further advice or encouragement/discouragement.

 

[MattBK]

Hi surferdude2 - Do I have this right? The main pain points are 1) closing up the leak with low temp solder 2) removing existing charge/refrigerant from the system (since I cannot release it into the air). I assume there is some charge because there is some frost on the lines.

If there was no leak and no charge, then it would be straight forward to charge the system with the appropriate tools? refrigerant, recharge hose, gauges, vacuum pump, piercing valve - is my understanding correct?

Also, if there is a charge in the system can we assume there is also air in the system because of the leak and I can't add charge without removing the air AND the existing charge? Or is it possible to only remove the air with the vacuum pump and keep the existing charge. Then measure existing charge with the gauges to determine how my of new charge to add?

Thanks for your help

 

[surferdude2]

The solder is a tin/silver alloy. Use Harris Stay Brite 8 for the good stuff. Ordinary solder is not suitable... it must be tin/silver with no lead.

Yes, you have to either remove the charge or isolate the leak by pinching off the lines on either side of it. You can't solder it with it leaking out. In your case, probably just determine that the system is already flat, which it likely is, and open it up for repair once you find the leak. I doubt you need to worry about the tiny amount of charge that is left in there.

Yes, evacuation is required and charging will have to be done slowly unless you can find the factory recommended charge and have a scale. Find out what type of gas it uses so you can know if can be vapor charged. I'd charge it through the process tube and pinch it off with vice grips to allow soldering it shut. You can use a saddle valve if you prefer... they can be a problem but not usually if installed properly.

No, you can't separate the air from the charge with a vacuum pump. Your system will need to be evacuated fully down to 500 microns, or lower if you have a good enough pump, since it operates in a low temperature application. I always shot for 100 microns for freezers but it takes a good quality 2 stage pump with clean oil to get there.

To repeat, I doubt there is enough charge left in your system to worry about trying to save it, small amounts are allowed for purging your hoses and you may find that is all it takes to discharge the system.

 

[surferdude2]

I will add that if you can't find the leak, there is no point in doing any further work on it. These systems rarely get leaks and when they do they tend to get worse in short order. I'm hoping there is enough charge left in there to allow you to soap bubble test it and find the leak. Sometimes that's easier said than done. It may take a shot of nitrogen or other inert gas to jack up the pressure.

 

[MattBK]

Thanks for the thorough reply. The fridge surprisingly has the charge amount (4.06oz) on the inside sticker. And I do have a food scale I could use. I'm going think about this for a week or so as the fridge is at a weekend home not currently in use - so I have some time.

To find the leak, could I not use plain old air? If I attach a piercing valve and connect a low pressure air pump?

 

[surferdude2]

I wouldn't use air... you can use a small amount of R-134a for testing... you'll need a can of it later anyway if you get the leak repaired. You can get it at Walmart for about $5. You'll need the piercing valve for the can that has the 1/4" male flare to accommodate your gauge manifold hose.

Valve

Refrigerant

 

[HotRodx10]

I don't think you want to pump air into the system, especially if you don't have a means to pull the required vacuum on it afterwards. There's obviously a little refrigerant left in the system.


I noticed it uses R-134a refrigerant. Would it work to add a low side port from an automotive A/C system to recharge the system? Or does it already have a similar port? I don't know how you'd connect it, since auto A/C systems are usually aluminum, and the lines on the the fridge appear to be copper until you get to evaporator coil.

 

[MattBK]

Is this the spot where I could braze on a service valve? Looks a bit deformed to attach a piercing valve. Also how do you use a torch in such a small enclosed area?

 

[surferdude2]

If you want to use a piercing valve, put it on the larger tubing at the compressor, the tubing that runs to the evaporator. If you decide to braze a pre stubbed schrader valve access port onto that stub, it requires the system to be void of charge. Then cut the existing stub off and braze your new port in its place. You can unplug the electrical connectors and get that wiring out of harms way and then, if necessary, use a non flamable shield if any further protection is needed from the torch flame. A #2 tip is all you need if using oxy/acetylene... otherwise a propane torch with a turbo tip is good.

The piercing valve may be the best way to go. There is always the possibility that this unit may not be worth fixing so you won't be out much cost.

One thing that comes to mind that you need to check is whether the present frost accummulation is immediately at the outlet of the filter drier. My experience with filter drier failures is that if they are several years old and all of a sudden get exposed to extraordinarily high temperature, like yours did when the condenser fan failed, the dessicant can flake off and partially plug the outlet. It can also partially plug the capillary tube. Either way it will cause the system to pump down all the refrigerant into the condenser coil and sometimes that fools a tech into thinking all is needed is to add some refrigerant. Obviously not the right move in a case like that.

That's why I want you to find the leak before breaking into the system. If you can't find one, then it may not have one and you may have a plugged system as outlined above.

I'll leave the gamble up to you. You can attach a piercing valve on the suction line of the compressor and connect a can of gas and see if adding some brings the temperature up. Be sure to purge the line of air before add any gas to the system. If that brings the temperature down, then you have lucked out and the filter/drier is still ok and the system isn't plugged. If instead it causes the compressor to bog down and stall from high head, it's pretty much game over and not worth fixing.

If it does accept a charge and starts cooling better, you will still have a leaky system and need to find the leak. Since this is an 11 year old system, I can tell you with certainty that there is no such thing as an 11 year leak... it didn't come from the factory with a leak. If it isn't a plugged system and it actually has a leak, you need to find it or the charge may not last long enough to make it a usable frig/freezer. Time will tell, charge it and see.

 

[MattBK]

Finally back at the weekend house. So I decided to do the easiest thing possible and see what happens. I put a piercing valve on the suction line, connected a charging hose with r134a, purged the hose of air, open the piercing valve, turned the fridge on. The gauge was showing about -15 psi. So I let small bursts of charge in (checking gauge in between) until I got to -2.3oz on my scale. I stopped because psi kept settling back to -15 psi and total charge the fridge takes is 4.06oz. (the internet told me I should be able to get to +1 or 2 psi when fully charged). I sprayed soapy water on the areas I could, including the evap coils and did not see any leaks. Figuring I need a new fridge and have nothing to lose at this point, I decided one more time to connect the gauge and try charging and watch the gauge. This time adding about 1.8 or 1.9oz (now overcharged?). Same thing happened. Then I just let the fridge run overnight because what the heck. Now the fridge is actually cooling. The top is about 39 degrees and the freezer is freezing ice. However, the compressor is louder than normal and has not stopped running even though the temperatures seem good. I turned the dials down to 1 out of 9 and the compresor is still running.