Temperature sender and Temperature sensor

KEIHINCARB

MEMBER
1992 HONDA ACCORD
2.2L
4 CYL
2WD
MANUAL
275,000 MILES

I know the location of both sender and sensor. When cold
temp. Sender measures approx. 240 ohms and when engine has warmed up 30 minutes later, it measures approx. 100 ohms. Don't know the specs? I've verified that the gauge is in good working order. Problem is (after engine is
warmed up) the gauge just sometimes stays stuck at below the C mark, sometimes it moves up a little (1/4") then goes back to below the C mark, and sometimes it'll move to the proper location as it should (1/4 up the scale). This all affects (rpm's) ie idle. Computer thinks it hasn't warmed up yet, thus idle / rpms on the high side 1200 rpms. Vs normally
at 700 rpms. One day all works well, next day not so much.
Does this mean that my temperature sensor (two wire) is going bad and need replacement or is it my temperature sender (one wire). Thanks

1 Reply

CARADIODOC

MECHANIC
33,763 POSTS

Two separate sensors were used on older models. The single-wire sensor is for the dash gauge. It doesn't have anything to do with engine performance. The two-wire sensor is for the Engine Computer

It takes a few minutes for coolant temperature to rise to normal. Idle speed doesn't remain at 1200 rpm that long. From what you're describing, it sounds like the thermostat is stuck open causing coolant to not get to the desired temperature. The place to start is by watching the actual temperature on a scanner to see what the Engine Computer is seeing. If it is normal, to check the dash gauge, remove the wire from the single-wire sensor and ground it. That should make the gauge read full hot.

Measuring a sensor's resistance is only useful in verifying it is defective, but not as a diagnostic step. Ford had a huge problem with failing two-wire sensors in the early '90s, but other than that, temperature sensors have an extremely low failure rate because there's just one part in them. When they do fail, they are usually intermittent due to a broken internal connection. They can read fine with an ohm meter, but go open circuit in operation, so the test gives false results.

When the gauge reads full hot when grounding the wire, but too low when connected to the sensor, the first suspect is corrosion on the terminal, either on the sensor, the connector, or both. Also check for a spread terminal in the connector, especially after someone probed it with a meter probe. If the gauge reads low right after the sensor is installed, check if someone used teflon tape on the threads. That will insulate it. There has to be a good electrical connection between those threads for the gauge to read right.

There's two things you can do to see if a stuck-open thermostat is suspect. One is to feel the temperature of the air coming from the heater. The second is to feel the upper radiator hose after the engine has run long enough to be warmed up. That should be too hot to hold onto for very long.