Welcome to our resource for MG Car Information.
MG MGF Technical - Overheating in traffic: radiator fan not working
traffic on the north circular was even heavier than usual as I travelled home from Silverstone.
Whilst moving very slowly in stop-start traffic, was shocked to notice that the temperature gauge needle was making a slow but inexorable move to join the red line at the top of the scale... 8oO
Being a hot and muggy day, the heater was switched off, but decided to see whether turning the heater and fan speed to maximum would at all help, and to my surprise and relief the temperature started to fall fairly rapidly. But it wasn't until we finally cleared the traffic foul up that the temperature properly returned to its 'normal' position.
Got home, parked and unpacked the car. Decided to run the car at idle with the heater off, to see whether the radiator fan would kick in.
Hot idle speed is normal.
So I don't suspect the ECU's temperature sensor. So I'm guessing a fuse or a poor connection?
Oddly, during the sprints earlier in the day, the water temperature had remained absolutely fine whilst idling in the start line que...
So what's the consensus on the most likely problem before I head home this evening to investigate further?
|Rob - Same thing happened to me last year. Got so hot the valve on the expansion tank blew. Very impressive cloud of blue steam when the coolant hit the engine!|
Turned out to have been the fuse, AA man had it on his check list as 'common fault'. Left me with a couple of spares.
|Cheers Rex. |
Was it the under bonnet fuse (fuse 6) that went, or the one in the passenger compartment (fuse 15)?
|Hi Rob - Under the bonnet one. |
You mean there are fuses in the passenger compartment?! Only had mine 3 years...doh
|I had the same issue myself a week ago. It was the first time (during our ownership) that the MGF had ever been stuck in traffic. Like you, turning the heater and fan on full worked.|
Took the car over to Trident the next morning and the guy on the service desk was out replacing the under-bonnet fuse almost before I'd finished describing the issue.
Replacing a fuse is an easy fix but I'm always left wondering why it blew in the first place.
|Cheers guys - I'll be checking this as soon as I get home tonight :o)|
Adrian, like you, I am concerned that the fuse blew in the first place. First one in over 7 years mind. I'll replace it, and if the fuse goes again (assuming that this is the problem), then I'll investigate further... either a short or the motor was drawing too much current?
The underbonnet fuse is rated at 20 Amps - I presume that this is a standard fuse type? I which case I've got plenty (I think).
Rex, according to the wiring diagram I've got, the under bonnet fuse is the main power link between the bettery and the radiator cooling fan. The passenger compartment fuse is for the cooling fan relay that is earthed via the MEMS.
|When my fuse went it was over the Jubilee BH last year, very hot day. I was sitting on the A303 for an age when it happened. The AA man reckoned the fuse blew when the fan started, due likely to the fan not having been used for a long time. I agreed, not having been in that situation before, v. hot day, stuck in traffic etc. |
Checked fuse under bonnet and it is 15amp on mine, at position 2. For info I've a 2000 VVC.
|The A303 can be a real pain can't it? Seems to foul up big time as it goes from two to one lane on the approach to Stonehenge...|
Well, I can't claim that my rad fan won't have been taxed recently, but I can't say for sure.
Looked up the MY2000 wiring diagram - not sure why the later cars use a 15 Amp fuse rather than the 20 Amp on the earlier Mk1 cars. Perhaps a different motor is used?
|Just to add to the confusion, the fuse on our W-plate 1.8mpi was 17amp, IIRC !|
|Now that IS wierd!!! Wiring diagram shows fuse 6 to be rated at 15 Amps for the MY2000 - irrespective of 1.8Mpi or VVC...|
I concur, the rad fuse blowing is a very common fault. I was on the Regency Run last year and there was an F in distress, so i stopped to help and was told about a cloud of steam...
I feared the worst, but it turned out to be the rad fuse under the bonnet. Since then i have heard of two other Fs with blown rad fuses so if the fan isn't on, then this is usually the culprit.
I recommend to checking the wiring to the fan. There are some sharp corners where it can wear and make a short cut.
Not fuses. Checked them both - both are okay.
Interestingly there is a discrepency between my owner's handbook and the workshop manual. In the handbook, the fuses are numbered in the wrong direction! D'oh!
Also, there is an error in the workshop manual - at least in application to my car: fuse 6 in the underbonnet fuse box is rated 15 Amps - not 20 Amps as in the workshop manual. Double D'oh!
And if all that was not enough, fuse 15 in the passenger compartment fuse box is stated as being the aircon cooling fan - and despite what both the handbook and workshop manual claim, is rated as 10 Amps rather than 20! Triple D'oh!
Recipe for confusion me thinks.
As mentioned before, despite the confusions over ratings and locations, all fuses check out fine. So I ran the engine to see if I could get the engine to over-heat once again. Took over 30 minutes! Long wait...
Car over heated. So I pulled fuse 6 and put an amp meter across the terminals. No current. With the engine still running hot, I pulled fuse 15. Again, no current.
Seems to me that the MEMS isn't triggering the fan.
Must be the ECU temperature sensor. Will get a new one, and will 'interogate' the existing one in the car. If I short the temp sensor plug - so the MEMS reads a resistance of close to 0 ohms, the MEMS should think that the engine is hot, and trigger the rad fan. This should confirm whether or not the sensor (or it's electrical connection) is the culprit.
Dieter - I hope that it isn't a chaffed wire. For starters, the rad and fan is very difficult to get to. I really don't fancy dismantling the whole front end of the car!!! But given the lack of a MEMS radiator fan 'on' signal, I don't think that this is the problem...
If the relay were shot the you would get no current when switching, to find out if the relay is working you have to get a voltmeter on the brown/black wire to the mems, it should be 12v when cold, and 0v when hot, if one of these values in not maintained then the relay is dead.
It is unusual for a relay to die like this, usually the contacts 'burn out', but it is just as lightly as a MEMS fault (but not as lightly as a temp sendor failure)
|Rob - I think it is most likely to be the connection to the sensor rather than the sensor, fan or fuses. I have had this many a time now and basically every 6 months or so I take the connections off, spray some contact cleaner around, replace and all is well again.|
Over the 7 years I have had the car now I had 4 new sensors put in before realising it is just the contacts. That was about 4 years ago now and touch wood not a problem since other than doing the regular maint.
As a part of this I do check that the fan works on a regular basis. If when you come back from a good long hard run (at least 1 hour) you park it and leave it running for about 5 minutes the fan should come on, even in the winter. I do agree with you though that if you start it up and wait for it from cold it can take a surprisingly long time to really get hot. As mine also has the uprated MS radiator now it takes even longer than it used to.
|Alan, thanks. I was wondering if the contacts might be the problem. What you've said will save me purchasing a new sensor, and I'll clean the contacts.|
A spray on contact cleaner? Not come across this before. A Halfrauds product?
Will, thanks for the info. I'll check the contacts and ensure that the fan comes on as it should when the resistance across the sensor is sent towards 0 ohms. If it doesn't then I'll check out the relay function as you suggest.
BTW, I thought that the MEMS connection to the relay was blue /slate grey?
|>BTW, I thought that the MEMS connection to the relay was blue /slate grey? |
Do'h, reading the wrong side of the sheet ! (rest of info is right though!)
>A spray on contact cleaner? Not come across this before. A Halfrauds product?
Fraid not, Maplin or RS / good electronic shop will stock it, or maybe a good 'band'/DJ music shop as it is used cleaning sliders in mixing desks.
|Actually Rob EFFECTIVE contact cleaner is hugely difficult to get hold of these days and I haven't seen any in the normal shops for years. I actually get mine from an engineer who works for the BBC. You actually only need a tiny amount so a small can lasts for ages. I doubt anywhere like Halfrauds would be likely to as it is something only needed in very small measures and wouldn't have the turnover to support the big boys. Specialists should have though. Really good HiFi shops are a possibility.|
Hope it all works. Alan
|Thanks guys. Depending on what time I leave work today, I'll try and pop down to Tottenham Court road.|
Failing that, anything wrong with some wire wool?
|I spent a long time looking for contact cleaner and was told that it's essentially the same stuff as carburretor cleaner! Of course, these days that might be just as hard to find... :)|
|Hmm - maybe the same as the de-greasant brake cleaner as well perhaps Tim???|
|Try CTC as used in fire extingishers BUT not the coloured stuff and only use in well ventillated areas.|
|> essentially the same stuff as carburretor cleaner!|
that wouldn't surprise me, but carb cleaner probebly would be more difficult to use (not as easy to direct spray), but carb cleaner removes carbon, grease etc. much like you woul expect contact cleaner to do.
>anything wrong with some wire wool?
getting inside the connectors woul be tricky, and you might leave a thread in there shorting things out.
>maybe the same as the de-greasant brake cleaner as well perhaps Tim???
I would expect the brake cleaner to be much harsher, but I'm no chemist!
|How about just scratching the contacts lightly with a sharp knife (like the tip of the small blade on a Swiss army knife)?|
|I did read that cleaning the contacts mechanically was a bad idea - but I can't find the reference now.|
When I suspected that my engine bay fan didn't work properly, I found that a standard potentiometer plugged directly into the loom (instead of the sensor), and I could then easily test the circuit by varying the resistance.
|You may be right Steve - I guess scratching the surface only increases the surface area to corrode in future :o(|
At least brake cleaner has a directional nozzle on it ;o)
Will try and get some electrical connector cleaner though.
|Yay! Radiator fan now works! :o)|
Purchased some contact cleaner from Maplin's on Tottenham Court Road last night.
I thought I'd try out my idea for a fan diagnostic - namely inserting a length of copper wire between the sensor plug contacts to reduce the resistance to close to 0 ohms.
Interestingly, this wasn't necessary: removing the plug, and presumably sending the resistance to infiniti had the same effect as the copper wire: the engine revs dropped on the cold engine, and the radiator fan kicked in. Maybe a fail-safe function, or a radiator fan diagnostic: remove plug, and the rad cooling fan should kick in. If it doesn't then there is a problem with the fan or it's switched power supply...
Liberally applied the contact cleaner to both the sensor and the plug, dried them off, and re-inserted. Quick drive around the block, returned and left the car to idle. STILL took an age for the water circuit to really warm up - but instead of overheating, the rad fan kicked in. What a sweet sound! :o)
Contact cleaner cost 2 quid - so thanks guys! :o))
Will be keeping a regular eye on this Alan, thanks for the tip :o)
Another success chalked up to the BBS ;o)
|Great result Rob - establishing the test process on the sensor and that your car's OK. After Neil's comments it makes me wonder wether the sensor I replaces a while back is actuall faulty or just dirty!|
One last thing for the archives - name and source of the cleaner product?
at a guess.
|Rob et al|
For those of us less familiar with our cars but always ready to learn from our peers - on really hot days, e.g. Saturday morning arriving at Silverstone, I stop the car, switch off the ignition and then hear a fan cut in.
Is this the same fan as you were discussing? BTW I have not yet seen any 'higher than normal' temperature readings.
I've never heard a fan cut in while driving even in central France last summer, although this may be a function of hearing failing with age plus the general noise level from an F with the roof down.
There are two fans:
1. Radiator fan - at the front
2. Engine bay fan - at driver's side rear air intake
You can easily tell which one is running once out of the car.
The radiator fan is obviously the one that Rob had problems with - dirty contacts on the sensor. This should cut in after 20-30 mins of the engine just sitting idling.
The engine bay fan doesn't cut in easily - the engine bay has to be pretty hot, although the VVC cuts in at a higher temp to the MPi and so is even less commonly heard.
|Never noticed the radiator fan on my VVC, but the engine bay fan quite often cuts in on a hot day when I reverse into parking spaces (I sit in traffic for at least 1 1/2 hours in the morning). Temperature guage has never shifted from one division below half. Is the rad fan noticeable if you're in the car?|
|The rad fan is inaudible whilst sitting in the car. You have to be standing outside to actually hear it.|
It is never in operation after the ignition has been killed. I guess it would be pointless given that there is no discernable coolant circulation to the rad once the engine (and therefore mechanical water pump) has been switched off.
You may never even notice that there is a problem, because with the heater and interior heater fan on, enough heat appears to be dissipated to keep the temperature 'normal'!
I only noticed that there was a problem because of being stuck in very slow moving traffic for over an hour, with the heater circuit and fan switched off...
Regarding the contact cleaner that I used, Will, I did look at that switch and contact cleaner (£2.50) - but in the end, just purchased the standard contact cleaner (£2.00). Oddly, this product does not seem to be listed on the Maplin's on-line store. However, I am sure that the switch and contact cleaner shown will do the job nicely! :o)
>>After Neil's comments it makes me wonder wether the sensor I replaces a while back is actuall faulty or just dirty!<<
Makes me wonder too. I tested the resistance values of a couple of supposedly 'dud' ECU temp sensors a while back - my old one, and one from Per's Elise. Both tested with expected resistances in hot and cold water... so perhaps this is not a sensor problem at all - rather a problem with dirty contacts!
I'll be amending my web site accordingly soon...
|The radiator fan is a hard one to hear. The car needs to be hot and running. The engine bay fan can (and does) cut in when the engine is either on or off and so more commonly heard. |
It took me 5 years before I atually heard the fan in the front and only then because I was checking to see if it actually worked.
|Thanks Guys, I am hearing the engine bay fan just behind my right ear (as I suspected) and fairly frequently (not scientific). Don't think I've ever sat still with the engine running for long enough for the rad fan to turn on.|
Thanks for help.
I remember when you tested the sensor I sent you. I got very surprised when you found that it was actually fine! Since then, I always take the plug from the sensor off and put it back on a couple of times, each time I open the engine compartment, just to maintain good connections.
Some of you may find the following description of how the system works useful:
The sensor is actually a "thermistor" (a resistor the resistance of which varies with temperature). When the coolant is cold or at normal temperature, the resistance to electrical current (which is monitored by the ECM) is high, which the ECM translates into a fan turned off. When temperature is high, the resistance becomes low. When temperature rises past a certain threshold, the ECM notices that the resistance becomes so low (i.e. the temperature so high) that it switches on the fan. The problem is that corrosion on the contacts of the plug f*cks up the monitoring, since it leads to higher resistance which can fool the ECM into believing that everything is fine with the coolant temperature, whereas it might be way too high and in great need of a helping hand from the fan.
In other words: good electrical connections = correct resistance readings by the ECM = fan switched on when needed.
|So what Per is saying is it is worth checking the connections from time to time. Have we managed eradicate all future HGF perhaps... :0)|
Good tip Per.
well done ... again.
Did I get this right ?
> remove plug, and the rad cooling fan should kick in. If it doesn't then there is a problem with the fan or it's switched power supply...
Is it approved ?
I think if the connection id the *true problem* then most of the other sensor *changers* will get new problems. Or bettter said the old problem with bad female spade connectors will get back.
Not aware of the 'remove sensor and the rad cooling fan should come on' being an official test of the cooling circuit. At present, that's just my observation (I was surprised!) Anyone able to confirm?
Your point regarding the connections and those who have already changed their sensor is well made - and is the same point that Alan and Per are making, namely...
... if you've changed the sensor, and a poor connection was the actual cause of the problem, then you can reasonably expect the problem to recur. It did for me, for example.
Regular checking of the connection may indeed be a relevant check - perhaps to be performed at every service interval?
|Thank you to all on the BBS.|
I took my car for a good blast and returned slowly for the last couple of miles. Reversing into my garage the engine bay fan cut in (I have heard this many times) so I left the engine running and walked to the front of the car and >>>>>>> I have a working radiator fan. Thank you all for teaching me something new about my little car. I shall be cleaning the contact at the weekend to be doubly sure that it works well in the future.
Thanks again and hoping for fine weather, enjoy driving your MGF and TF's
quick note. Super Servisol 10 ( as used above) is available from Farnell Electronics @ £2.15 per can - order code 278786 (and no I do not work for them).
so confirmed. I only wanted leaving out misunderstanding :)))
However. Now the idea......
I will get an Aircon switch (for Aircon fan)
Mount the relais near the brown ECU coolant sensor.
- Connect a relay (NC contact) between the hot wire of the sensor.
- wire the relay coil to the aircon fan switch (push button activates relais)
- wire the aircon fan switch light parallel direct to the front fan.
If the push button is engaged the fan MUST come of due to interuption of the brown sensor. This must be indicated by the switch light.
Another advantage is for tests and normal running.
- the switch light will show whether the fan gets powered or not.
- if light comes off you *should hear* at last whether the fan works. And it MUST work, or anything is wrong with the fuse or wiring.
etc. etc..... Please think about.
I mean it's a cheap and useful solution for nearly all failure cases of the fan, fan fuse, ECU sensor, ECU itself and affiliated wiring.
Is this worth getting a patent ? ;)
*Dieter's magic fan test and over-heat emergency switch*... LOL
(OK, Carl suggested nearly similar the other year)
This might be an easy and cheap DIY solution.
PS. will do that job ASAP at my MGF and suggest here and in Clubmagazines if it does what it should.
|Sounds fine - but remember that after 256 tests on that button the MEMS will do a "format C",shut down and leave You stranded in the middle of nowhere :)))|
Just joking, seems a good thing to do! Keep us informed about any gremlins or odd behaviour due to the breaking of the sensor circuit.
Would have done the mod just today but I need a good relay at first (IP6x with mounting bracket or other with gold plated contacts). I know I have one anywhere in the cellar, but where ? ..
>after 256 tests
LOL... I think the ECU internals look like 64-technique only ;)
|Rob - Glad the contacts were the problem. Judging by the comments made after maybe I should have taken some shares out in the makers?|
As I said earlier I haven't now replaced a sensor in years. I just keep the contacts clean on a regular basis and check the fan operation regularly. It could be worthwhile asking dealers/mechanics to do it when having a service as when the car is on a lift it is easy to get to the connector and just to disconnect and reconnect a few times to scrape the contacts clean would take seconds, and wouldn't be significant on a bill. You would need to trust them though. Equally a cleaner spray is just seconds when you have the access.
|Alan, a good idea that the sensor should be a part of the regular service. It ought to be suggested at least.|
I wonder why this sensor plug, rather than any of the others, causes so many problems?
Dieter, great idea :o) Look forward to hearing how you get on!
>>Super Servisol 10 ( as used above) is available from Farnell Electronics @ £2.15 per can - order code 278786<<
Thanks for this David - used half the can on my Triumph ensuring all the electrics were up to an MoT pass today - proved to be a worth while investment as it turned out! ROFL
|Where is the sensor whose contacts need to be cleaned?|
Marvellous, further to my earlier praise of this board, I have today removed front bumper from my '99 VVC and discovered......TWO fans. Is this because my car has Air conditioning?
Air conditioned cars have two radiator fans. Either one or both of them will work at the same time depending on coolant temperature.
I concur with the advice above regarding the sensor conections. In fact I clean all connections I can see every time I service the car.
the sensor is located on the left front corner of the engine on the coolant outlet pipe from the head.
See http://www.mgf.ultimatemg.com/hgf_pages/related_problems_overheating.htm for more information - I haven't yet updated the site with the sensor plug cleaning advice...
|Wow! I miss logging-in for a few days and the thread grows from 5 to 50 posts! Looks like I'll have to have a careful read and then run some checks on the car.|
|>I haven't yet updated the site with the sensor plug cleaning advice... |
LOL :))) and not the *bracket push* advise on pic *D*
>Looks like I'll have to have a careful read
this is IMO the most important cause for HGF after all fixed other terms from the last years.
- gasket should be stable re-designed with steel dowels (2002)
- coolant cap tooling reworked in 1998 (first better batches marked with white dot)
- inlet manifold leaks solved with green gasket (1998)
- overfilling of expansion tank *should not happen any longer at good dealers*
- advise on 4th bleed location special after body repairs with coolant drain (1999?)
Now they need better connectors surface plating at the sensor spade connectors and plug female connectors.
- Gold plated (galvanic not the cheapish crap chemical gold)
Female connectors are from Tyco-AMP
The sensor is another question.
PS. additional to Rob's look at this
|>>LOL :))) and not the *bracket push* advise on pic *D*<<|
ROFL Indeed not Dieter! ;o)
Funny thing though, with the sensor screwed in the way it is on my car, it is really difficult to push the clip in to remove the plug. Might actually have been easier to remove it as shown in pic D !!!
Should have taken a picture whilst performing the job, but the angles weren't easy :o(
Dieter, do you have a picture of the 'correct method' that I could use? Cheers :o)
>> http://www.mgfcar.de/sensor/index.htm <<
Can definitely recommend this very useful page - perfect for checking the performance of the sensor :o)
|It must be the weather !|
Mine did the same this evening - got stuck in a long queue coming off the M27, and just as I started moving again, noticed the water temperature right up just below the red. Immediately turned the heater full on & the temp dropped rapidly back down to normal.
Now got two jobs to do.......check fuse and relay, etc, and finish off the audible temperature alarm I started to build in the cold winter months and should have finished !
|>Should have taken a picture whilst performing the job, but the angles weren't easy :o(|
>Dieter, do you have a picture of the 'correct method' that I could use? Cheers :o)
Mate, I woun't break _my_ fingers ... LOL :))
Seriously, it isn't easy. I agree, everyone should take his own way. Flying spring or broken hand :)
|>>Flying spring or broken hand :) <<|
LOL! Absolutely ;oD
>> ... and finish off the audible temperature alarm I started to build in the cold winter months .. <<
This sounds interesting Tony. Look forward to hearing the details regarding this project :o)
I have to declare that the design is not mine. 'He' who designed it must claim the credit. My intention was, and still is, to put it on my site once I've finished it. However, since I have not got my finger out, and taken so long about it, perhaps it would be an advantage to all for me to go ahead and load it up before I've finished it.
I will mail 'he who designed it' and see if he agrees.
Excellent! Checked the sensor connector this morning - engine running and wobbled the connector - fan cut in, out, in, out, etc. Disconnected it, cleaned with contact cleaner, re-connected it - perfect. Obviously something to check periodically.
Well done :-)
|The Water Temp and Low Fuel Alarm circuit diagram is now up on my website. Go to....|
|Great news Tony :o) Glad to hear that normal service has been resumed with your F's cooling system :o)|
Wowzers - circuit looks complicated - good work Bruce. Probably looks worse than it is on that diagram - but I look forward to your pictorial instructions Tony :o)
|Tony, No ground connection?|
|Well spoted Will, but I'm sure you knew where it should be! :-))|
The circuit has now been updated. Thanks
|Meeting of the minds !|
I have been thinking to keep my MGF on the condition that I could install an audio/visual signaling system to monitor the coolant temperature.
Now, Tony and his "connections" have done it; Kostas will keep his MGF.
I wish the developer of the system would come forward and start providing complete DIY kits, please.
Perhaps, Rover could buy the rights (do they exist?) and save the otherwise magnificent MGF experience from being spoiled by the eye strain - one eye on the road; the other on the coolant temperature gauge.
This thread was discussed between 23/06/2003 and 09/07/2003
MG MGF Technical index
This thread is from the archive. The Live MG MGF Technical BBS is active now.