Refrigerator Water Valve Circuit

[wilheldp]

I'm trying to troubleshoot a problem with a refrigerator where the water dispenser valve sticks in the open position.  The signal to the valve is controlled by a microcontroller through a triac.  This particular fridge has a primary solenoid that opens whenever the icemaker or water dispenser calls for water, and individual signals for both.  The control of the primary solenoid is done through diodes attached to the individual solenoid signal lines.  I have attached a diagram of the circuit from the service manual.  Is there anything that jumps out at any of you that would make the dispenser valve stick on?  If there's nothing in the circuit that could get stuck, then the fault lies in the program of the microcontroller which is far harder to track down.

[Fank1]

If ONLY the dispenser valve sticks and not the primary valve also, you just have a "sticky valve".

[wilheldp]

No, both the primary valve and the water dispenser valve are getting "stuck."  I have measured the voltage at the connector on the circuit board (J3-7 on the diagram), and I'm getting about 70 VAC on it.  When it is properly activated, you get 120 VAC on that pin.  That's why I think it might be something with the design of the circuit instead of a failure of the valve or microcontroller.

[apelly]

Half broken triac?

[wilheldp]

Tests fine.  And this is a systematic problem with a certain model control board.  There are 3 identical triacs on the circuit board, and the other two don't exhibit any failures.

[bobcat]

On the schematic, J3-7 is not connected directly to the valve solenoids. What else is on that circuit?

[wilheldp]

On the schematic, J3-7 is not connected directly to the valve solenoids. What else is on that circuit?

J3-7 is connected to the solenoids via wires and connectors.  I think they just put that gap in the schematic to show that it is a removable connector.

[SeanB]

Could be leakage in the triac holding it partially on, or there is a need for a snubber across the load side. Is the 70VAC the voltage open circuit or with the 2 solenoids connected? If open it is leakage, and if it istill there with the coils connercted the triac is either breaking down or triggering from noise.

Measure the voltages with the solenoids and show a clear photo of the front and back of the board, especially the area where the triac and connector is. As well the triac part number.

To see if it is leakage use a 5W appliance lamp across the coils to show if there is enough current to light the lamp with the fault present.

[wilheldp]

Could be leakage in the triac holding it partially on, or there is a need for a snubber across the load side. Is the 70VAC the voltage open circuit or with the 2 solenoids connected? If open it is leakage, and if it istill there with the coils connercted the triac is either breaking down or triggering from noise.

Measure the voltages with the solenoids and show a clear photo of the front and back of the board, especially the area where the triac and connector is. As well the triac part number.

To see if it is leakage use a 5W appliance lamp across the coils to show if there is enough current to light the lamp with the fault present.

Unfortunately, I don't have the refrigerator that I originally tested this on any more.  I just have the circuit board and photos from my inspection.

From my photos, it looks like the 70 VAC is present at the solenoid connector without the dispenser solenoid attached, but I never disconnected the primary solenoid.  However, I did remove J3 from the circuit board, and didn't measure any voltage across the solenoid connector.  This indicates to me that the fault is on the circuit board itself.  After some further inspection and circuit tracing, I found that the solenoid signal comes from a pin on a microcontroller, goes through a darlington driver, then the darlington is attached to the gate of the triac.