Water pressure related to refrigerator inlet valve leak?

[Stephenson]

Hi All,

I have not been on the site in over a year - I am saddened by Terry's passing.

I have a sort of plumbing related question that I'll bet someone has some idea of ...neighbor gave me a nice refrigerator with a bad control board. Troubleshot the board to the power relay for the compressor and condenser fan (they share power) - ended up destructively removing the top of the cover and noting burned contacts from arcing - was able to sand points and bend the relay frame a bit and got it working consistently enough to confirm nothing else was wrong. Ordered a replacement and will install - all OK, so far.

Wanted to test icemaker to make sure it worked, so rigged up a garden hose to 1/4" copper to the inlet fitting - no issues - BUT, got a leak from the inlet valve outlet to the refrigerator - the push pull type connection with a release collar. Pulled apart and other than a ridge on the tubing where the locking mechanism inside the outlet fitting grabs the tube, all looked fine. Screwed around with it for awhile - could not get it to seal off - just very slow drop ever few seconds. Pulled, again, and tightly wrapped a couple of turns around the tube where it would contact the o ring in the fitting - did not affect the tiny leak in any way.

So, do I simply have too much pressure into the valve? 3/4" garden hose to the adapter - usually, there is a 1/2" supply line in the house to a 1/4" valve connection to the supply to the refrig. Could I be providing too much pressure from the garden hose?

 

[Breplum]

If your pressure is under about 80 psi, then that is not a factor.
the fitting is typically called John Guest type and the little o-rings deform readily and then leak. Needing to replace fittings of that type is normal and common.

 

[Sponsor]

 

[Stephenson]

Thanks, Breplum.

Read a bit more on these single oring John Guest connectors - seems like a common failure mode relates to the single o ring deforming from lateral pressure related to the stiff tubing. This particular valve connection has a tube that curves rather aggressively and at a bit of angle, inside the alcove then outside the case ... pretty stressing. These devices should clearly use a two oring connector.

There was some discussion relating to the replacement of the oring, but limited examples.

I hate throwaway technology, especially for something if this nature!

 

[Slomoola]

Any easy "convenience" pipe or fitting like SharkBites or Pro Presses, most use O-rings. Most city water is chlorinated. Not a recipe for longevity. I would avoid anything with O-rings if at all possible. All you can do is slather with 100% silicone grease (Magic Lube) and hope for the best. Here in OKC, I can smell chlorine in the AM hours in our cold water.

Replace what you have leaking. Move on in life. BTW, good catch on that PCB repair.

 

[Stephenson]

Thanks, Slomoola - I know most folks don't have time to play with electronics - and, I know jack about repairing, but learning. I know cost is the answer, and I know most appliance electronics are more reliable each year ... and, I know cost is lowered by making them smaller, and including as much as possible on one board - BUT, it would seem reasonable given this to make the PCB components easier to replace, not harder. And, yeah, think of the money in the secondary and replacement PCB market!

 

[Fitter30]

John Guest and others that are push to connect fittings most failures are caused by end not square and burr end that wasn't cleaned off.

 

[Stephenson]

So, I did another pull and examine ... end of the male fitting is a bit scratched from where is it held in.

So, carefully placed some silicon grease on the oring area, sanded the male fitting area with 2000 grit paper, greased the fitting then reinserted and water on. Tiny drip. Rotated collet while under pressure and leak stopped. No way it will stay that way, but fun to investigate.

 

[Slomoola]

Remember on all new type refrigerators and freezers. You MUST use a good surge protector LOCALLY AT THE ELECTRICAL PLUG. Just as Stephenson is working on, the circuit boards in these new computers slash refers get taken out by surges. Even if you don't "think" yours doesn't have circuit boards, put a surge protect on her anyway.

You get surges when something turns on AND off. My house has a 5 ton compressor unit outside. Inrush current draw is 150 amps at 240 volts. Simple math on that one is 36,000 watts of inrush power. I put a type 1/2 surge protector on that one too.

 

[Stephenson]

I installed a large capacity, whole house "surge protector" - my garage refrig is on a single outlet with a quality surge protector. Unfortunately, the main one is house (three years) is so difficult to remove I haven't tried to pull it out - some built in Miele monster.

Around here in Florida the most common losses with our frequent electrical storms are garage door openers and irrigation systems.

 

[Fitter30]

So, I did another pull and examine ... end of the male fitting is a bit scratched from where is it held in.

So, carefully placed some silicon grease on the oring area, sanded the male fitting area with 2000 grit paper, greased the fitting then reinserted and water on. Tiny drip. Rotated collet while under pressure and leak stopped. No way it will stay that way, but fun to investigate.

If printed circuit boards had replacement components that would put technology back in the 1960s with plug in parts. Where computers were as big as a building and a simple calculator has more computing power. Can't compare even retail cost with manufacturer cost.

 

[Stephenson]

Agreed - but, for some devices - washing machines come to mind - the more basic approaches were pretty reliable and easy to fix. Now, they didn't look as good, weren't as energy efficient - and the buzzers were waaaay more basic

I like small computers and slick communications devices ... pehaps they are more reliable because they don't mix line power with the very small feature electronics.

 

[Stephenson]

Well, the silicon grease didn't fix it, the 2000 grit smoothing didn't fix it ,,. how about cleaning with spray brake cleaner to remove silicon grease residue, blowing out with compressed air, then spreading RTV or marine silicon, on the male fitting end and installing - leave water off for a few hours?

 

[Stephenson]

So ... some tips from a youtube video noted simply cutting the tubing and direct inserting it into the valve connector ... my tubing has a make fitting in the end of it, so not certain this will work based on diameter difference? Same fellow in video had a leak in the angle fitting that was used to change direction into the valve fitting - to relieve stress, likely, since thus may be cause of a lot of failures? So, he spliced a new piece of tubing using a John Guest splice, then added a piece of 5/16 tubing run directly into the valve without a male fitting on the end.

Question ... is the male fitting there to reduce stress and allow repeated replacements of the inlet valve? It is much harder than the tubing, right? My male end fitting has circular scratches where it is captured by the female Guest fitting, but this is downstream of the oring location.

Thinking about taking valve to big box store and trying 5/16 tubing fit ... what do you think?

 

[Fitter30]

Refrigerator tubing for ice maker 1/4" od. Also have inserts for vinyl tubing both in brass and plastic.

 

[Stephenson]

Fitter - 1/4" the tubing TO the inlet valve - OD is 5/16 FROM the inlet valve to the water filter - I think all the tubing in the fridge is 5/16.

So far without success:
- Cleaned and lubed with silicon grease
- Cleaned and sanded male fitting with 2000 grit paper and lubed with silicon grease
- Cleaned and sanded male fitting with 2000 grit paper, cleaned again, cleaned female fitting, carefully applied marine silicon sealant (way stickier and stronger) to the oring, and to the male fitting - had removed water and dried everything - reinstalled with a bit of tape to try and relieve some of the lateral stress. Will wait a few hours and see what happens

I don’t think any of this will work - and, in the last case, above - will likely connect the fittings together so completely as to need to replace the valve - at least. Perhaps both the valve and the tube. Tubing runs from the inlet valve to the water filter. Together about $60 or so.

Here’s the issue - as I see it - the single oring female fitting is simply not holding the tubing in alignment, the misalignment causes the lateral stresses from the stiff tubing to press against the single oring which flattens it out enough to preclude perfect sealing. It gets worse over time as the tubing is always putting that lateral load on the oring. Once the tubing is moved, or even after some period of time, it causes the leak to start. There are also theories re the chemicals in water which cause the oring to deteriorate - most older ones are EPDM - new ones I think are Viton (seems like this material is used in HVAC systems).

This video is interesting and instructive - the replacement tube has a formed in curve where it connects to the same type of fitting on the water filter. BUT - the other end going to the inlet valve is straight - it has to be because it must be snaked through the back of the case to get down to the inlet valve. BUT - this video shows the inlet valve - and the other valves - having clip on stress relief devices (“tube retainer” which clips ONTO the collet and ALSO establishes the alignment with the valve!) which are absent on my frig - an update to stop the issue?

And, here is a creative fellow dealing with same basic design, but with an angle coupler for inserting into the bottom of the inlet valve - I'm assuming to relieve stress? In this case I am assuming the other end does not have a curved fitting or pre-bend as it would not go through the case.

Tried to upload photo, but size limits for server are circa 2005 cameras, I guess.

 

[Stephenson]

Continuing the saga …

Was leaking after the marine silicon - no real change in amount when I flexed the line from the valve - could still get it to drip faster or slower based on position.

Next stage of investigation becomes potentially destructive of the valve and the plastic male end on the water tubing.

I measured plastic fitting diameter at 0.3100” and the tubing at 0.3115”.

Removed the collet on the valve with a small flat blade, and the o ring (thicker and stiffer than most o rings of this diameter), removed the cured marine silicon I had used. Note the indentation on the o ring of the collet pressure down on it. (photos)

Soaked the o ring in alcohol and silicon spray for an hour, flipped the o ring upside down, put valve back together - reinstalled - still leaked, but a bit less, I think.

Removed tubing, carefully examined the plastic fitting - by now has been in and out of the o ring and collet capture clips about 20 times - so is pretty discolored. Cut 3/16 of the tip off, reinstalled without lubricant. No leak so far … have cycled valve via the water dispenser a couple dozen times.

IF it doesn't leak, here’s my take on failure mechanisms:
- only the contact point between the tubing fitting and the o ring (fixed by cutting tip and establishing a new contact point)
- o ring became brittle/damaged over time by water chemistry (potentially fixed by flipping it over, and/or soaking in alcohol/silicon)
- combination of one or more of the above

If it starts to leak, I will need to decide whether to cut the entire tip off and use without it (the plastic tip is 0.0015” smaller than the tubing), or buy another valve.

Discoloration indicating some kind of rubber deterioration -

Indentations from the collet inside the valve -

Smooth side when flipped over -

Collet with four retaining clips on each arm -

Pretty messed up looking plastic fitting - groove is caused by movement in the collet retaining clips - some residual marine silicon -

Cleaned out of silicon, but o ring deterioration mark showing on green plastic?

O ring without collet -

Collet installed -

Dark areas where oring contact, crisp ring from collet retainer clips - but some indication of o ring wearing the plastic?

3/16” cut off tip - sanded the edge to ease entry -

Edge of plastic fitting same 3/16” shorter as amount cut off - no leak so far ...

 

[Stephenson]

No leaks, so far!

So, this seems pretty silly ...
- $0.10 o ring would likely fix
- changing the oring sealing surface (cutting tip so it slides to new position)
- having positioning stress sleeves to ensure/maintain alignment
- new design with double o rings
- threaded fitting with cone shaped seal like supply line