Aller au contenu principal

Repair and technical information for the GE Refrigerator PSB48******, a side-by-side model with an in-door ice dispenser and automatic defrost system, produced by General Electric. This page covers information for model numbers matching the pattern PSB48******.

2 questions Voir tout

Fridge temperature display doesn't match reality until power cycle

I have a recurring problem with a GE Profile PSB48LSRABV. Here’s an example sequence:

  • Fresh food compartment set to 37F on the control panel (for months)
  • Thermometer in fridge reads 46F (with lots of water inside per manual)
  • Press “Display Temp” button, display shows fridge at 37F
  • Turn power off, count to 15, turn power on, wait for control panel to boot
  • Press “Display Temp” button, display shows fridge at 46F (matches thermometer!)
  • Compressor runs and cools fresh food compartment
  • Hours later thermometer in fridge reads 37-38 (yeah!)
  • Press “Display Temp” button shows fridge at 37F
  • Within days, the fridge is warmer inside again, thermometer inside reads 45+, but display reads 37 - unit needs to be “rebooted” again!

Fresh food thermistor from top of cabinet measured properly in ice water (and warmer water), the unit has always passed all the self-tests, and condensor coils and fan are clean, so I had techs come give it a shot under home warranty:

  • Trip 1 - Tech came to assess, ordered new evaporator fan
  • Trip 2 - Replaced fresh food evaporator fan (note new fan is working in fridge, and old fan works fine on bench)
  • Trip 3 - Ordered main computer PCBA
  • Trip 4 - Replaced main computer PCBA

Unfortunately, it continues the same cycle of misreading the temperature until power cycles, and taking a day off of work for each visit is getting painful.

Any ideas?

Répondre à cette question J'ai le même problème

Cette question est-elle utile ?

Indice 1
5 commentaires

Someone suggested that I read the thermistor at the control board when the fridge is misbehaving. I don't think that's a valid reading on the meter because both the meter and the fridge are running current thru the sensor, and the voltage will not correspond to what the meter expects.


Taking a look at the schematic in the Technical Service Guide (page 57), I'm suspicious that the ground to the thermistor may be floating up due to load of two motors and 4 thermistors if there is a weak connection somewhere with increased resistance. Pin 7 of CN30 of the main board provides the ground. That grey wire goes thru a bunch of connectors and splices before it hits the sensor. Assuming the PCB drives a constant current thru the thermistor and measures the voltage at pin 8, a slightly higher ground reference would lead to an equally higher voltage. This would mislead the board to think the resistance was higher and thus the temperature lower that it really was. Hopefully there aren't splices burried in spray foam insulation of the cabinet!


Hi @sbpgh ,

I was thinking about taking a resistance reading with the power turned off. That way you're measuring the circuit involving the sensor where it goes to the temp control.

If you do this with the power off it is only resistance and if you do this when the temp indicator is 46F (when the fridge is 37F) and then do this when the temp indicator is 37F (and the fridge is 37F) then if there is a difference then the problem is in where the sensor output leads connects to (is there a temp control board?). If there is no difference then it is back towards the sensor or the earth connected to it


I gathered some more data with settings FF=37, FZ=0 watching typical behavior:

- Display shows 38

- FF thermistor reads 2.29V from CN05-3 to CN05-2.

- FF thermistor also reads within 14mV from CN05-3 to CN02-7

(I think this shows that the problem is not a ground float of the intended ground reference since I used a ground wire that is only used for door switches)

- Compressor running, FF fan and feedback read 0V @ CN05

- Power cycle the fridge

- Display shows 43 and FF thermistor still reads 2.29V

(I think this shows it's not a thermistor input problem, and the PCB is connected and biasing thermistor correctly)

- FF fan reads 7.32V and FF feedback reads 2.5V (within spec)

- Fridge cools down until it reads 36, FF thermistor reads 2.60V max, then starts dropping (warming), fridge fan and feedback now read 0V

- Compressor continues to run

- Fridge evaporator reads 2.9V, freezer reads 3.7V, freezer evaporator reads 4.1V for a long time.

- The next day, display shows 37 and 0, but FF reads 2.300V


Today I found it displaying FF=37, FZ=15 with the compressor off. Curiosities:

- Why does the display fail to update? The thermistor is clearly putting the right value into the PCB, but the display (or the PCB) gets confused

- Why does the fridge fail to turn cold again?

- Could the system be low on freon and not cooling things down well enough before the mandatory defrost kicks in after 4-5-10 hours?

- Is the computer refusing to cool the fridge until the freezer is cold after the initial chill?


Ajouter un commentaire

1 Réponse

Réponse la plus utile

As I suspected, the unit was low on refrigerant. Apparently the firmware in the PCB satisfies the fresh food temperature and then tries to satisfy the freezer. It doesn’t seem to read the fresh food temperature during that time. If the freezer can’t actually reach temperature, the system eventually times out (not sure if it’s freezer evaporator defrost or freezer defrost?) and tends to the fridge again.

Service tap was less than -10psi initially. Adding some R134a refrigerant to bring up the pressure to between 0psi and +2psi (depending on how cold things were) made things happy again.

Cette réponse est-elle utile ?

Indice 2

2 commentaires:

Hi @sbpgh ,

Good work on finding the problem.

Did you find out why it was low on refrigerant though?

It is a sealed system so there should be no loss of refrigerant in the system, unless there is a leak somewhere.

Unless it was like this from new, if you don't know why it was low perhaps it will be low again soon.

Just a thought


yeah, I know it shouldn't leak down, but no, I didn't search too hard for the leak this time around. The fridge came with the house purchase with a tap on the service tube, so it's probably been an issue in the past (or has a leaky tap). I'm just glad to have proven that the sensors, PCBA, and compressor are actually functional and will eventually get to sniffing for any leak... probably more motivated by the next time it acts up.


Ajouter un commentaire

Ajouter une réponse

Scott B sera éternellement reconnaissant.
Nombre de vues :

Dernières 24 heures : 4

7 derniers jours : 14

30 derniers jours : 108

Total : 10,277