Coolant level sensor - 1993 325i
 

I have a 1993 325i sedan with 118,000 miles on the clock. Roughly 15,000 miles ago the low coolant level indicator started coming on frequently and pretty consistently. In all cases the coolant level was always just fine. I replaced the coolant level sensor and the problem persisted. I should have checked the old sensor because a simple OHM meter check showed that it was working fine.

I have noticed that there is a lot of turbulence in the reservoir bottle. The nature of the circuit is that if the coolant level drops (opens the circuit) even momentarily while driving, the low coolant warning will flash on when you turn off the engine. So my theory is that the turbulence is causing the float to bounce around. Just one bounce sufficient to cause the float to drop and open the circuit will cause a false alarm.

Questions:

1. Any one else experience this?
2. What has changed? I figured the new sensor might resolve the problem thinking the float no longer had sufficient buoyancy. But the replacement seemed to make no difference.

As a bit of a hack I simply installed wiring with slide connector that allows me to just short the coolant level sensor, effectively causing the sensor to report full coolant level at all times. Not really a good way to resolve the problem but not much different than a constant false warning either.

Any suggestions or insights would be appreciated.

Search for recent threads about this.

It is caused by a cracked wall inside the coolant reservoir.

Thanks for the suggestion and pointers. I have thoroughly examined the coolant reservoir. I see no evidence of cracking anywhere. The partition inside the bottle looks solid and intact. I have even probed with a long screwdriver and cannot detect any problems. I see that the other contributors are typically the sensor or the wiring. I have changed the sensor (as described in the original post) and it had no effect. I will check the connector. I imagine that could cause the problem given the nature of the circuit ... if there is a partial break in the connector, then all it would take would be a momentary opening of the circuit to have a similar effect as the sensor bouncing momentarily low (due to turbulence or other undetectable problem). Right now I have a "bypass" wired across the two conductors leading to the sensor connector. This would mask a connector problem. Will advise the posting if I have success.