Funny Fan

[beljames]

I see that as we move towards summer all of the fan and cooling posts are coming out! Here's another!

For some time now my fan has had an intermittent fault where it fails to cut in when very hot. When its just merely a bit hot, there is no problem and the fan either comes on naturally or I can operate it from the over-ride switch. However, after a long run when it is frighteningly hot (you know, you rag it solid for 1/2 hour and then traffic jam), the fan fails to cut in on either the switch or naturally.

Unless... I switch the engine off and back to ignition only, it then happily comes on, but actually start the engine again, and the fans cut out once more.

All of the wiring is straight (and you'll appreciate the unique nature of the symptoms make it somewhat hard to test, but I've been around withe multimeter!!), so it must be a screwy otter switch or relay.

Anyone seen this kind of thing before?

[MajorClanger]

With the override switch wouldn't this suggest a loose contact or wiring problem? Presume the override switch bypasses the otter switch but not the relay, so have you checked the relay (although why the relay would be affected only when the engine is hot I don't know)?

MC

[GreenV8S]

How bizzare, I can't think of any single fault that would cause those symptoms. If I were you I'd get it to 'go wrong' and then short out the otter switch with a screwdriver, just because it's a ten second job and rules out funnies with the otter switch and override switch. Then I'd concentrate on the fan relay. I would expect to find the input pin permanently live, the output pin live when the fans are 'on', one side of the coil live when the ignition is on and the other side of the coil floating when the otter switch is cold and pulled down to earth when it is hot. If you can confirm similar behaviour when the fans are working normally, hopefully you will then be able to see what has changed when the fault occurs? When I had a similar problem with the fuel pump relay I found it very helpful to make up a short extension lead for each terminal of the relay so I could wire the relay in with it physically removed from the socket. This allowed me to get at the terminals to mreasure things. Obviously need to be wary of changing the system you're trying to measure if the problem is with the socket itself.