Troubleshoot iced up evap in home fridge

[gangemcs]

hi.  I've seen a couple of your videos.  Very useful.  I have a 7 year old GE refrig. - GSH25JSTASS.  The evap coil is frosting up and then the temps inside fridge and freezer were lower than they should.  Comp and fans work.  I defrosted it then tested the thermister at the top of the evap and got close to 10K ohms at ~ 50 degrees F, the heater sensor got about 2 ohms at ~ 55 degrees F and the heater element itself (one that is long and screwed in with 2 screws at bottom) was ~ 16ohms.  So all those usual suspects seem to be ok.  What I cannot figure out how to test is the defrost timer on the MB.  Is that testable?  Should i just replace it?  I also know there are at least 1-2 more thermisters not connected to a connector inside the freezer that I don't know where to test on the MB.  They physically look ok so I suspect they'd ohm out ok.  I'd appreciate your thoughts.  My next thought is the defrost timer or something else on the MB but before replacing it I'd like some professional feedback. 

[DurhamAppliance]

the board uses an adaptive defrost to determine when it should defrost.... time is only a very small part of its algorithm.  A combination of run time,  previous defrost time,  sensor readings and door opening counts are used to make this determination.

Determining board failure on this model is done through pricess of elimination.  If heater,  evap sensor and defrost bimetal are within spec,  then that leaves the board as the fail point.

You have successfully ruled out the heater and the defrost bimetal/limit but your thermistor test is not complete.   You have shown your thermistor registers ok at 50f, but suppose it fails when it hits 32f, or 10f?  Suppose it wont register anything lower than 50f? If that's the case ,  the board would think the evap is always 50f and would never need to turn in defrost,  right? 

So you will need another temperature reference point.... ideally,  they shiukd8be tested at the normal operating temp of the compartment or item it's sensing.  0°f is ideal but sometimes difficult to recreate.  You could try getting a reading while the sensor is on a cold evap that's not encased in ice... what reading do you get?  Many times we use the ice water test.... ie submerge the thermistor in a glass of mostly crushed ice with a little water for 5 minutes.  This should be 32f and your thermistor should read about 16.3 k ohms.  If this target is reached,  then you probably have a bad relay on your board but I would still change the thermistor first just in case it has an intermittent failure due to water intrusion.  If your fridge still doesn't defrost,  replace the board.  This way you haven't lost anything since it's a good practice to replace the evap thermistor anyway if it hasn't been changed in years,  especially if it's one with the black sealant where the wires enter the thermistor body. 

[gangemcs]

Thanks for the helpful tips!  I re-tested the thermister (one clipped onto the top of the evap tube) and heater sensor but both right at 32°F this time. I used an ice bath with a thermometer in it.  The heater sensor is ~ 1.7ohms (should be 0?) and the thermister is ~ 14.1K ohms.  Sounds like I should replace at least the thermister since it's off by more than 5%.  To replace, I simply cut it out and splice in new one since the other ends are going to a connector and it's hard to get the spades out?  When installing new one, you recommend simple wire nuts and electrical tape to connect wires again or but connectors?  Is it worth testing the other thermisters?  If I still don't have de-frosting them also replace the MB?  

When I try to order from your website, it sends to me Repairclinic.com.  Are you related to that website too?

Thanks again for the help.  I know a little about refrigeration as I work for a compressor and component OEM (Danfoss) but I don't get into details too much like our technical guys.  Avoiding a service call by doing some simple testing myself will hopefully save some money and time.

[johntech]

I reccommend using crimp on bell connectors then seal with silicone on any splices inside the refrigerator. That way you can keep moisture out of the connections.

[Chat_in_FL]

15 hours ago, DurhamAppliance said:

the board uses an adaptive defrost to determine when it should defrost.... time is only a very small part of its algorithm.  A combination of run time,  previous defrost time,  sensor readings and door opening counts are used to make this determination.

Determining board failure on this model is done through pricess of elimination.  If heater,  evap sensor and defrost bimetal are within spec,  then that leaves the board as the fail point.

You have successfully ruled out the heater and the defrost bimetal/limit but your thermistor test is not complete.   You have shown your thermistor registers ok at 50f, but suppose it fails when it hits 32f, or 10f?  Suppose it wont register anything lower than 50f? If that's the case ,  the board would think the evap is always 50f and would never need to turn in defrost,  right? 

So you will need another temperature reference point.... ideally,  they shiukd8be tested at the normal operating temp of the compartment or item it's sensing.  0°f is ideal but sometimes difficult to recreate.  You could try getting a reading while the sensor is on a cold evap that's not encased in ice... what reading do you get?  Many times we use the ice water test.... ie submerge the thermistor in a glass of mostly crushed ice with a little water for 5 minutes.  This should be 32f and your thermistor should read about 16.3 k ohms.  If this target is reached,  then you probably have a bad relay on your board but I would still change the thermistor first just in case it has an intermittent failure due to water intrusion.  If your fridge still doesn't defrost,  replace the board.  This way you haven't lost anything since it's a good practice to replace the evap thermistor anyway if it hasn't been changed in years,  especially if it's one with the black sealant where the wires enter the thermistor body. 

 

1 hour ago, johntech said:

I reccommend using crimp on bell connectors then seal with silicone on any splices inside the refrigerator. That way you can keep moisture out of the connections.

whatever happened to the "like" button?

[evaappliance]

I've been wondering about the like button too   

[Space]

Have you checked your condensate drain to make sure it's not plugged up with gunk or ice?

[gangemcs]

Thanks for the comment.  The condensate pan is not clogged or anything.  I think it's either the control board or thermister clipped onto the top of the evap tube because I could only get ~ 14K ohms across it and it's supposed to be 16.3K.  I tested it while in an ice bath for several mins @ 32°F.  If not that then it's the main control board.  There's not much else that they put in these fridges.  GE is sending me a new board in case it's that but I'm gonna splice in a new thermister anyways and see how it goes before I put in the new board.  The original comment was that it's good practice to replace these thermisters anyways after so many years and this fridge is almost 9 years old.