Anablog

As an electrical engineer I am cursed with electrical failures in my life. I already told you about my two failed fluorescent light ballasts. I suffered for five years with failing Harley generators and regulators (until I found the frayed wire under the fender that was shorting the generator out, but only intermittently). So a few weeks ago I am sitting upstairs in the office/loft/warehouse/lab/shop/consulting-place megaplex and suddenly my Maytag refrigerator starts running the icemaker and plopping ice cubes onto the floor via the convenient refrigerator door-mounted delivery system. I ran to the fridge, poked the door-switch and few times and then yanked the freezer door open to stop the cubes from raining down on the floor. I soon found a cut-off switch under the bezel that could be thrown to stop the icemaker. The implications of this are mind boggling. It did this just sitting there; I had not used it in a day. If I was not home or worse yet, on vacation the icemaker would have just kept flopping ice onto the floor until the waterlogged mezzanine collapsed or perhaps the motor in the fridge would have burned up. With my luck it would have caught the whole place on fire, water logged or not.

I pulled the unit off and the burn mark by the switch was not encouraging. Now the point here is systems-level design. I am astonished by how some component manufacturers seem to think system-level design is nothing more than catalog engineering. They think we just pick the parts out from a catalog and presto! Your design is done. They think designing the components is what is hard. Actually the exact obverse is true. Ask anyone that has to design a system on a chip how complex things become with just two or three subsystems.

With this refrigerator problem I have a hard time blaming Cherry. I really doubt the Cherry switch was defective and I really doubt that the Maytag engineers are so stupid as to put an under-rated switch into the application. I suspect the failure mode was due to water being trapped inside the rubber boot meant to keep water out. Since the icemaker just started running by itself I have to believe there was a creepage breakdown across the switch that conducted enough current to start the motor. Either that or the contacts were actually closed with some high-impedance film that suddenly broke down and started to conduct. See why systems-level design is so tough? I learned about waterproofing at Ford. The taillights had a little hole in the bottom. I pointed out that this could allow water in. The designers said that no matter how hard they tried to make the taillight hermetic, water always got in and then the seal would hold the water in, not keep it out. That would corrode the bulb socket and cause failure. After following a few Chryslers with tail lamps half-full of water I saw their point. I suspect this Maytag would be better off with no rubber boot trying to shield out the water. The humidity in this area had to contribute to the failure. See the pictures and judge for yourself. Oh, and yet another shout-out to Digi-Key. It did take 15 minutes but I found two replacement Cherry switches, both of 15 amp instead of 10 and one having a high actuation force (pdf). That is the one I will use.

Out comes the door icemaker delivery system. Yes it used the dreaded Torx fasteners.

The black smoke-marks cannot be good. You can see the backside of the rubber water shield. It forms a cup and I suspect it just held water and created a high humidity environment that caused creepage in the switch.

Sure enough, the switch was toast. I suspect it had a creepage failure that make the motor run and then when I ran up and pressed the switch a few times that is what caused the contacts to be closed. If Cherry or Maytag wants the switch I will save it for them. It really is hard to diagnose—all I know is that you have to be very careful to not say—“oh it was a bad switch”. Cherry makes really good stuff. It is conceivable a bad one got out but you also have to look at the environment that the switch worked in. I thought those ten Harley generators and regulators were junk too until I found the frayed wire under the rear fender.