|
Post by turbokinetic on Sept 13, 2018 23:16:22 GMT
It seems that many of us have encountered broken off or cracked thermostatic bellows tubes on these machines. For those who aren't into the technical aspects of these machines; the bellows and bellows tube work to sense the temperature the refrigerator is running at; and operate the main contacts to cycle the compressor off and on.
When this part is damaged, the compressor will not cycle off and on automatically. In other words, you can get it to run, but it runs constantly. If you stop it, it won't re-start by its self.
In this series of videos and discussion, I am trying to repair a broken bellows that was sent to me by a fellow member. In this case, the bellows its self has some dents in it; but the main problem is the completely broken off tube.
The parts.
First thing was to swage the end of the broken tube and prepare it for reconnection.
The issue with this repair is getting a charge of gas back into the bellows. It originally had some gas in it (maybe SO2?) so something has to be charged into the sealed bellows and tube after the repair, for it to work. The factory charge was put in through the end of the sensing tube, then the end was crimped shut and brazed. I don't know how much gas will need to go in, so I want an accessible service port of some sort.
It seems that a tire valve, with the rubber removed, will do nicely! The brass part sitting next to the bellows is the modified tire valve.
The service port is soldered in place, and the tube is re-attached!
Next, I need a way to evacuate and charge the bellows. That is an adaptation of a refrigerant hose, and a bicycle pump chuck! It will lock onto the tire valve and allow me to use the customary vacuum pump, gauges, etc. to charge the bellows.
I charged it with butane. This will have a little more pressure than the original SO2 (if that was the original charge) so some adjustments of the control will be necessary.
There are more details in this video. As said the hard part will be calibrating the control after it's back on the fridge!
|
|
|
Post by birkie on Sept 14, 2018 0:00:42 GMT
Nice work!
Yes, the DRs used sulfur dioxide in the bellows. Cks and I think CAs used methyl chloride
Was the can isobutane, or n-butane? Isobutane is very close to SO2, and would probably require minimal adjustment. N-butane is farther off, and may require more tweaking.
That charging valve is a nice addition, especially for experimentation. What I've always wanted to to (ever since someone else on the form posted about a similar problem) was figure out precisely how much to add.
In the theory - and I don't know how this would end up working in practice - you could calculate the saturation pressure at something like 35F ; slightly above the control range. Assume that it the bellows is full of saturated vapor at that point (no liquid). Temps below that point will be saturated (which is what we want), with liquid pooling in the cap tube as the temp goes down. Temps above that point will only slightly raise the pressure via the ideal gas law PV=nRT. So given the 35F saturation pressure, use the ideal gas law to calculate the pressure at room temperature then, with the bellows in place (and the control in the "on", bellows-expanded position, fill up to the calculated pressure
The above would minimize the bellows pressure at room temps (machine unplugged), yet allow the control to operate in the expected saturated range. Or at least, I've wanted to see if it ends up behaving that way.
Very nice post.
|
|
|
Post by turbokinetic on Sept 14, 2018 0:15:12 GMT
Nice work! Yes, the DRs used sulfur dioxide in the bellows. Cks and I think CAs used methyl chloride Was the can isobutane, or n-butane? Isobutane is very close to SO2, and would probably require minimal adjustment. N-butane is farther off, and may require more tweaking. That charging valve is a nice addition, especially for experimentation. What I've always wanted to to (ever since someone else on the form posted about a similar problem) was figure out precisely how much to add. In the theory - and I don't know how this would end up working in practice - you could calculate the saturation pressure at something like 35F ; slightly above the control range. Assume that it the bellows is full of saturated vapor at that point (no liquid). Temps below that point will be saturated (which is what we want), with liquid pooling in the cap tube as the temp goes down. Temps above that point will only slightly raise the pressure via the ideal gas law PV=nRT. So given the 35F saturation pressure, use the ideal gas law to calculate the pressure at room temperature then, with the bellows in place (and the control in the "on", bellows-expanded position, fill up to the calculated pressure The above would minimize the bellows pressure at room temps (machine unplugged), yet allow the control to operate in the expected saturated range. Or at least, I've wanted to see if it ends up behaving that way. Very nice post.
Originally I had thought it was isobutane, but after getting home and looking at the can I am afraid it is n-butane. Normally when they don't specify, it is the n-variety IIRC.
It would be good to scientifically charge just the right amount in the tube, as you describe; but I don't have any instrumentation to make that happen.
What I could do is chill just the lower two or three inches of the sensing tube, and allow liquid to condense there only. This could be accomplished by having the butane can cooler than room temperature, but not as cold as the chilled tip of the tube. This would ensure that the entire thing wasn't charged with liquid, as well.
|
|
|
Post by turbokinetic on Sept 14, 2018 1:38:48 GMT
I do remember seeing a store with isobutane. Just trying to remember where that was now and what they were selling that for......
|
|
|
Post by coldspaces on Sept 14, 2018 1:58:27 GMT
Somewhere I have info about how to charge them, if I can remember where I read it.
|
|
|
Post by turbokinetic on Sept 14, 2018 2:09:25 GMT
Somewhere I have info about how to charge them, if I can remember where I read it. That would be great. I know it can be done through experimentation; but any sort of procedure would be helpful. I don't have a lot of instruments, as said, to do anything overly scientific. Unfortunately.....
|
|
|
Post by birkie on Sept 14, 2018 3:01:52 GMT
It would be good to scientifically charge just the right amount in the tube, as you describe; but I don't have any instrumentation to make that happen. I just did some math - really all you'd need is a pressure gauge. So if we pick a "highest temperature still in saturation" (e.g. 35F), then pressures at higher temperatures are determined by the ideal gas law and are proportional to the ratio of absolute temps, assuming constant volume (bellows is in the control and doesn't grow any bigger). P2 = P1 (T2/T1) At 35F, the saturation pressure of isobutane is 24.14 psia, that'll be P1 An 85F room is 302.59 K, 35F is 274.82 K. Those will be T2 and T1 24.4 psia * (302.59K/274.82K) = 26.58 psia To get gauge pressure from absolute, subtract atmospheric pressure; ~14.7 psi 26.58 - 14.7 = 11.88 psig In other words, in an 85F room, fill with 11.88 PSI of isobutane as a gas. Overcharging should be fine; the only real consequence is higher pressures in the delicate bellows when the machine is off. Because those bicycle fittings go pssst when disconnected, probably overcharging a bit is wise. So my hypothesis is to charge with isobutane gas to some pressure slightly greater than 11.88 PSI with everything at ambient temperature, and with the bellows in the control. I'd love to see the contemporary recommendation was for charging these things. A different gas (e.g n-butane) would have a different P1, of course
|
|
|
Post by coldspaces on Sept 14, 2018 3:12:32 GMT
|
|
|
Post by coldspaces on Sept 14, 2018 3:29:14 GMT
Don't forget a fully charged bellows should not be removed from the control without first chilling the bulb. It then must be clamped while still cold to prevent over expansion of the bellows to be stored at room temp. Wouldn't this mean you should have it clamped to charge it at room temp?
|
|
|
Post by birkie on Sept 14, 2018 3:45:29 GMT
Wouldn't this mean you should have it clamped to charge it at room temp? Thanks for the info! I guess that confirms that this kind is charged as a gas. Yes, I agree that the bellows need to be clamped if filling at room temp - or placed in the control (which has the same effect). Placing it in the control seems ideal if using the calculations for minimum charging pressure above - that way you don't have to worry about differences in volume beween clamped vs the volume it sits when in the control at room temp.
|
|
|
Post by coldspaces on Sept 14, 2018 4:00:30 GMT
Wouldn't this mean you should have it clamped to charge it at room temp? Thanks for the info! I guess that confirms that this kind is charged as a gas. Yes, I agree that the bellows need to be clamped if filling at room temp - or placed in the control (which has the same effect). Placing it in the control seems ideal if using the calculations for minimum charging pressure above - that way you don't have to worry about differences in volume beween clamped vs the volume it sits when in the control at room temp. Good point! Having it in the control would be ideal.
|
|
|
Post by ckfan on Sept 14, 2018 12:39:12 GMT
I hate to post a gee whiz comment but I had to. This thread is awesome. You guys rock!
|
|
|
Post by turbokinetic on Sept 14, 2018 14:14:46 GMT
Guys; thank you all so much for the great replies! I really appreciate it. Somehow I got the quotes out of order below, so just bear with my ineptitude LOL!
Thanks for the info! I guess that confirms that this kind is charged as a gas. Yes, I agree that the bellows need to be clamped if filling at room temp - or placed in the control (which has the same effect). Placing it in the control seems ideal if using the calculations for minimum charging pressure above - that way you don't have to worry about differences in volume beween clamped vs the volume it sits when in the control at room temp. Good point! Having it in the control would be ideal. Yep, one of the things I considered was that the control could be assembled / disassembled with the port in place. Also I was considering putting the port on the tip of the sensing line, but that would have prevented removal of the bellows from the control without reaming the hole where the tube passes through. If it were my control, I wouldn't be worried about that; however I also don't know how large the hole is in the top deck of the refrigerator, where the tube passes.
Don't forget a fully charged bellows should not be removed from the control without first chilling the bulb. It then must be clamped while still cold to prevent over expansion of the bellows to be stored at room temp. Wouldn't this mean you should have it clamped to charge it at room temp? Yea, I remember reading about several sad stories of people having the bellows all stretched and sprung out because of this mistake! I had planned to charge this one in the control, for this reason and as well as to be sure the volume was going to be the same during charging as it will be in operation.
Excellent! That is extremely helpful.. Thanks for putting it in a new thread so it will stand alone and be easier to find.It would be good to scientifically charge just the right amount in the tube, as you describe; but I don't have any instrumentation to make that happen. I just did some math - really all you'd need is a pressure gauge. So if we pick a "highest temperature still in saturation" (e.g. 35F), then pressures at higher temperatures are determined by the ideal gas law and are proportional to the ratio of absolute temps, assuming constant volume (bellows is in the control and doesn't grow any bigger). P2 = P1 (T2/T1) At 35F, the saturation pressure of isobutane is 24.14 psia, that'll be P1 An 85F room is 302.59 K, 35F is 274.82 K. Those will be T2 and T1 24.4 psia * (302.59K/274.82K) = 26.58 psia To get gauge pressure from absolute, subtract atmospheric pressure; ~14.7 psi 26.58 - 14.7 = 11.88 psig In other words, in an 85F room, fill with 11.88 PSI of isobutane as a gas. Overcharging should be fine; the only real consequence is higher pressures in the delicate bellows when the machine is off. Because those bicycle fittings go pssst when disconnected, probably overcharging a bit is wise. So my hypothesis is to charge with isobutane gas to some pressure slightly greater than 11.88 PSI with everything at ambient temperature, and with the bellows in the control. I'd love to see the contemporary recommendation was for charging these things. A different gas (e.g n-butane) would have a different P1, of course Again, very good info. As I said (sorry for repeating myself) I will have to get a better gauge capable of reading to 0.1 PSI to do this. I don't trust my old manifold gauge set. These were never meant to be that precise or accurate. I appreciate your research on the formula. Math never lies, but schitty instruments can feed you bad starting numbers.... you know GIGO effect!
I would rather error on the side of having an overcharge to the degree that some liquid is in the sensing tube, as long as it's not so overcharged that liquid is trapped in the bellows. As you said, the coupler does release some gas as it's disconnected; and there is really no way to quantify this nor to release the coupler with exactly the same speed each time.
There will always be some uncertainty involved doing this sort of thing outside of a laboratory environment. I am confident that the control I strong enough to constrain the bellows with some liquid in it at room temperature. Seeing how far the bellows expands when someone accidentally takes one out of a control without clamping it; it would seem that there would have had to be liquid boiling off to provide that level of extension. My point being, I believe that they planned for that contingency in the control design.
I hate to post a gee whiz comment but I had to. This thread is awesome. You guys rock! Thanks; I am very hopeful that it'll be a solid operating control after this repair! Time will tell.
I plan to get isobutane today (if I can remember where I saw it for sale); a metal valve cap, and (possibly) a better electronic gauge.
Originally, I thought the cans I had were isobutane but they are not. I will locate isobutane for this. The isobutane should be much closer to the SO2 in temperature / pressure curve. I believe this will allow the control to work without any major modifications to the springs or mechanism. The knob is already removed, so the calibration is going to have to be done anyway. I think that by running the fridge with the knob removed, and finding a point where it cycles at the correct temperature, it can be calibrated without modification.
Thanks again for the great info!
Sincerely, David
|
|
|
Post by coldspaces on Sept 14, 2018 16:37:37 GMT
Good Luck David, glad I could help. It is going to good info to know how to recharge one of these, thanks for documenting the work.
|
|
|
Post by bhammer on Sept 14, 2018 18:00:30 GMT
Well. I can't express how fortunate I feel for David reaching out to me and helping get my unit back operating. I am optimistic and at the same time flabbergasted by his abilities. I am equally impressed by his taking the time to document and share his work and progress. This knowledge is of great value to me and I hope other forum members appreciate it as well and find it helpful. Thank you thank you thank you.
John
|
|