Fan won't switch off!

[spud]

Hi all,

Here's a strange one that's got me totally baffled:

Whilst recommissioning my HPE VX the cooling fan, which had been working correctly, stopped cutting in. I checked all the connections and found that the wiring coming out of the radiator sensor itself had broken insulation and frayed wires, with one of the three wires virtually severed. I repositioned the wire in question so as to remake contact and then the fan came on... but wouldn't go back off once the temperature guage had gone back down. It stayed on even with the ignition off and the temperature virtually at zero. I got it to go off by disconnecting the battery and when I reconnected the battery the fan was off. With the wiring so dodgy I figured that at least the sensor was junk so I sourced the correct one from my local factors and fitted it. I then started the engine and sat waiting for the temperature guage to rise to the usual fan cut in point which is a fraction above halfway. However, the fan cut in much earlier than this and... refused to switch off. So I thought... hmm, maybe the relay? I found a spare relay of the correct type and fitted it but... no change. Fan stuck on permanently. Whether ignition on or not. I can't get it to go off! If I hadn't disconnected the battery it would still be on now, whirring away all night...          

Anyone got any suggestions?

This car, restored at the previous owner's VAST expense at Betacar, is beginning to drive me nuts!!! I just wanna drive the b4$"ard.

Andrew.

[thecolonel]

I would imagine that somehow you have by-passed the sensor/switch.
engine fan sensor works by going to ground

normally of the three wires one is attached to an earth ring around the sensor this then links to
a second wire and the third goes to the fan (often via a relay)
If you have a test light connect to bat + then touch to the sensor wires to find which are earthed
remaining cable should be fan wire.

Hope that makes sense.
Geoff

[Neil-yaj396]

Did you replace all the damaged wiring? If not that could be earthing?

[peteracs]

Hi Andrew

I only have the Haynes wiring for the earlier cars so may be different for yours, but in the diagram it does appear that the relay is powered from the ignition switch for the relay power, so as soon as the ignition is switched off the fan should stop (not sure if this makes sense as would expect the fan to run after switchoff, anyone else comment?).

However I would suggest you invest in a cheap multimeter if you do not have one, always useful to have.

This is what I would do to track down the issue, which I suspect is either earth or shorting connection (or incorrectly wired up).

With the fan running and the ignition off, remove the single wire from the sensor (which goes to the relay), does the fan stop?

Next remove the wire from the sensor at the relay does the fan go off?

Let me know

Peter

[spud]

Thanks for your thoughts guys.

Geoff, it would seem from reading my first post that I must have bypassed the sensor, but I haven't. Not by any of the connections that I have made anyway. But perhaps it is effectively bypassed because, despite being brand new and unused, it is faulty...
The sensor, as I understand it, works like this: One wire, connected to a body earth, goes into the sensor. There is a thermostatic switch inside the sensor which, when it is activated by the water temperature reaching a certain point, makes a connection from the aforementioned earthed wire to the second wire coming out of the sensor. This wire connects to the relay and causes the relay to switch on the power to the fan. Correct or not?
The sensor has three wires, as you described. The shortest one does indeed connect to an earthing ring round the sensor. What does it do though? In my description above it isn't needed, is it? It certainly doesn't alter the way my fan works whether it is connected or not...
I do have a multimeter and can understand the basic tests of establishing voltage and continuity. The earthing tab washer that fits around the sensor does make a continuity connection to the radiator body, but the radiator body isn't connected to earth of course. The sensor wire that connects to the body earth isn't connected to the wire which connects to the earth tab as you suggested Geoff, it's connected to the wire which goes to the relay. This, I assume, should only connect when the thermoswitch is activated by the water temperature. On my new switch it is connected all the time. Is this wrong? Have I purchased a dodgy sensor? Saying that, I kept the old sensor... and it's wires are connected to each other in the same way... but perhaps that sensor really is broken...
The wire which comes out of the relay to connect to the sensor is registering 12 volts. Is this correct?
I have tried wiring the sensor up in every permutation possible with the three wires but it makes no difference- it's either on permanently or doesn't work at all. And anyway there is really only one way to wire it as it uses a combination of certain wire lengths and connector types that only allow one way of connecting it.

Neil, the damaged wiring was only part of the old sensor- the one wire still in use is the wire that runs from the relay to the sensor and this is perfect. I bypassed it anyway with a new wire just to check and the fan operation remained the same.

Peter, the fan on this car, when it was functioning correctly, did run on after the ignition was turned off. Perhaps the later cars are indeed different.
The answer to both of your questions is Yes.

Any more comments/help very gratefully received. My thinking at this moment is the new sensor is suspect...

Andrew.

[thecolonel]

Hi Andrew,
beacause the rad isn't earthed and the sensor needs to be, you will have an earth wire (from car)
going to the sensor this needs to go to the body of the sensor via the earth ring, the third wire goes
to the fan. It's been a while but as far as I remember the relay operates on the + side of the fan and
is ignition controlled. I'll try and dig out a spare sensor and check again.
I presume all wires on the sensor are black and the connecting loom has a black and a Black/white ?

Geoff

[peteracs]

Hi Andrew

You are correct in saying that the sensor is simply an on/off switch which is controlled by the temp of the water in the rad. Once it goes down below a certain temp it should go open circuit and hence turn off the fan, same as wiring in a switch. As the fan goes off after you disconnect the wire, then the fault has to be with the sensor which is obviously happy to 'switch on' (albeit at a lowish temp from what you say), but does not appear to 'switch-off'. I, like you, would conclude that the sensor is faulty.

It is fairly simple to test assuming that when you disconnect the sensor the fan switches off AND when you reconnect it switches back on again AND the temp of the rad is low (ie not hot enough to switch it on in the first place), simply disconnect all the wires and using the multimeter on Ohms setting, take a reading across the two terminals. This will probably read a fairly low value.

In answer to your question about the wire to the sensor being at 12V when disconnected, that is correct, the switch simply pulls it to earth when on and this is what activates the relay.

My guess as to the fault in the sensor is that they normally have a temp to switch on and a different temp to switch off, the ones you have probably will switch off eventually, but the temp to do this may be lower than the ambient temp, hence never do so.

Peter
 

[spud]

I presume all wires on the sensor are black and the connecting loom has a black and a Black/white ?Geoff

Geoff, that is correct.

Ok guys, before I pull out what little is left of my hair could one of you please test your fan switch/sensor with a multimeter?

I have just received another one (same type- intermotor) and rather than just fit it I've tested it with a meter. There are three wires as you know. Wire number 1 has a male 'bullet' type connector- this is designed to connect to the white/black wire that goes to the relay. Wire number 2 has a female 'spade' type connector- this is designed to connect to earth- one of the earthing tabs located below the right hand side headlamp to be precise. Wire number 3 also terminates in a female spade- this is a very short wire designed to connect to the copper earthing tab that is fitted around the sensor. Now, when I connect my meter to wires 1 and 2 I get a beep which is signalling that they are connected. Ohms resistance zero. Surely this is wrong? This is the same as the sensor that I have just fitted- I know the fan will immediately switch on when I fit it.
So if one of you helpful souls out there could test yours I would appreciate it.
AND... if yours turns out to be the same then could someone explain to me how the fan switch is activated? I'm feeling a little stupid about this, electrics aren't my strongpoint but I do know 12V DC only needs a + and a -...

Thanks guys,
Andrew.

[thecolonel]

Ok, it's possible that the switch is designed for various cars.

as you know the bullet is to the relay. Boil a kettle, fill a cup and place sensor in it,
check resistance between bullet and either of the other black wires. You can now
determine which are the switch wires and the other black goes to the ring.

connect the two switch wires (Black and bullet) to car loom wiring.

I've sent you my phone number if you want to chat about it.
regards
Geoff

[spud]

I've sent you my phone number if you want to chat about it.
regards
Geoff

Thanks Geoff, very kind of you.

I think though, I THINK, I've solved it.
I did a similar experiment to your suggestion Geoff. I suspended the end of the sensor into a pan of water and boiled it. I attached the multimeter to the short earth lead (the one intended for the earthing tab around the sensor) and the long lead (the one intended to go to the body earth under the headlight). When the water approached boiling the meter beeeeeeped, indicating a connection had been made. When allowed to cool down the beep ceased. Success! The sensor comes with incorrect connections on it! Simple as that. The short wire with the female spade on it, which is designed to fit to the earth tab around the sensor, should actually have a bullet connector on it to fit to the white/black wire from the relay...
When I tried this experiment on the car with the first new sensor the fan didn't work so I can only assume either a) It wasn't hot enough or b) That sensor is faulty.
I'll fit this second new sensor on Monday and fit the correct connectors to the correct wires and hopefully it'll work as it should. I'll let you all know but it seems to be problem solved.

I hate new stuff that doesn't work properly. There's only three wires- you'd think a manufacturer of the size of Firstline, who produce a huge amount of electrical bits n bobs, would be able to get that right... If I hadnt persisted with this I'd have been relying on a manual switch.


 
Andrew.

[spud]

 

"seems to be problem solved"... famous last words...
The sensor works PERFECTLY in the bloody pan. Doesn't work at all on the car!
Wiring is now correct.
All connections tested and proved to be good.
Fan works when the relevant wire on the sensor is grounded.
Doesn't work when the car is warmed up and at the supposed switch on temperature. 
I allowed the temperature gauge to rise to three quarters; at that point I took evasive action and grounded the sensor manually to get the fan on to cool it down.
(For those that don't know- be careful if you're ever in a similar situation because if you just merely switch off the engine the temperature will rise significantly after that because there will be no water circulating anymore and the hot metal will begin to heat it up further, possibly taking it into the red).
So- the only possible explanation for this is that the water isn't getting as hot as the guage says it is... When I stated at the start of this post that the sensor works perfectly in the pan I actually meant it does switch on and then has to cool significantly before it switches off: I actually don't know the temperature of the water when it switches on... I may have to invest in a temperature guage that I can hold in the water to check it. Can't say I've ever seen one but I'm sure they exist...

Life is never dull with a Lancia...  Heyyy, get onto FIAT with that one; that slogan is a winner!

Andrew.

[peteracs]

Hi Andrew

Seriously frustrating I agree.

Did you double check the wiring, especially if the earth/ground side actually goes to earth? (sorry to ask but sometimes the obvious ones have to be asked)

Is the temp of the water next to the sensor getting v hot, or is there a chance the rad has a blockage which allows water through, but not much?

For a temp sensor I bought a multimeter with temp sensor with it as well, cost £30 from memory off Ebay (new).

Peter

[Thotos]

The temperature gauge shows temperature at the cylinder head while the fan switch detects temperature at the bottom of the radiator. If you don't have good flow of water round the engine and the radiator there could be a vast temperature difference between the two.

[spud]

Fair comments both. All the wiring is good, all earth connections go to earth.
Because I wanted to flush the whole system and refill with new water/antifreeze I used a radiator/water system descaler flush additive before I fitted the new sensor. It is from Halfords and is in two bottles- you run the engine for half an hour with each bottle added to clean water with the thermostat removed. (It was at the end of this operation that I noticed the broken wire insulation on the original sensor). There is also a (tested) brand new thermostat fitted. The water was flowing freely around the whole system when I flushed it- also the radiator flowed perfectly well and when I changed the sensor for the second time the water emptied out of the sensor hole pretty sharpish. I think the system is as clean as possible and free from any restrictions. The rad gets hot all over- no cold patches...
Peter, that multimeter sounds interesting, I'll have a look.

Andrew.

[peteracs]

Hi Andrew

This is the one I bought, somewhat less than £30...

http://cgi.ebay.co.uk/DIGITAL-MULTIMETER-TEMPERATURE-PROBE-LEADS-CASE-/400154778101?pt=UK_ConsumerElectronics_SpecialistRadioEquipment_SM&hash=item5d2b1559f5

Peter

[spud]

Thanks Peter,
That looks like a handy piece of kit to have even without my current problems so I'll get one.

Cheers,
Andrew.