MG-Cars.com
MG MGF Technical Temperature teaser BBS discussion at MG-Cars.net
MG-Cars.net
Welcome to our resource for MG Car Information.
Recommendations
Parts
MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.
|
|
MG MGF Technical - Temperature teaser
| Following the steam erruption following the eariler CCF (coolant cap failure) and the associated worries about a potential HGF, I am still finding that the water temperature starts to rise in traffic- particularly of the slow stop-start variety. Interestingly, thrashing the car around Donington, the temperature guage stayed in the 'normal' position all day long! So it appears that there is a lack of moving air through the radiator? Or is there another potential explaination? How do I check the operation of the rad fan? I can't hear it running- so is there another way of checking its correct operation? (It nestles behind the radiator and can't be viewed from above) Is there a specific fuse? Okay, okay, I need to get a workshop manual!!! Thanks in advance for any help Cheers Rob |
| Rob Bell |
| Rob, do you hear a fan at all, the only one I've heard is the side intake fan whirring sometimes. Usually after a jaunt in stop-start rush hour traffic. But like you said, I doubt you would actually hear the front fan unless your ear was pressed to the bonnet anyway. Steve. |
| Steve Childs |
| The engine bay fan seems to be operating normally Steve. Not sure how to check the rad fan short of crawling under the car and waiting for it to come on. Doesn't sound like my idea of fun. |
| Rob Bell |
| Rob, I park the car in a car lot and if I leave the car on in there after a drive, after a while the radiator fan comes on and you can easily hear it when it kicks in. Perhaps you are now finding the true cause of the overheating. David |
| David |
| I'm sure there have been threads in the past about the front fan either not working or working all the time: time to trawl the archive! I seem to rememeber that its been a faulty sensor.... Neil. |
| Neil |
| There were also stories about it being wired in reverse, so that it goes backwards and effectively does nothing. |
| Ed Clarke |
| Thanks guys. Yes, I remember those threads blaming a faulty sensor. Corroded sensor terminals perhaps? I think that the non-operation of the rad fan is the prime suspect- thing is how do I check its function? Answers on a postcard to the usual address ;o) |
| Rob Bell |
| Rob: ecard! There's a good thread from Aug 2000 in the MGF Technical Archive entitled "Radiator fan thermostat" which describes in gorey detail what's what and how to test it! Neil. |
| Neil |
| Nice one. Thanx Neil :o) |
| Rob Bell |
| I know from my experience the testbook cannot detect fan failures even if tests exist for it. I saw it at dealer when mine failed ;) Gauge went never up tho, you just saw grey smoke coming out of engine bay due to overheating. But the gauge will also move up if their is NO water passing the sensor. Airbubbles perhaps? If the sensor is really busted, it will do anything anytime, not just in traffic jams. Once mine dropped down to zero (!) every time passing 5500rpm. Never knew why, but 3 weeks later another head gasket went up in smoke, dealer never wanted to relate the two. Better have this checked Rob, perhaps your gasket is already failing, but not enough to have a continuous failure. |
| Dirk |
| Thanks Dirk, I intend to. I just want to check out other possibilities first... Okay, checked the fuses: underbonnet fuse "2", rated 15 amps- OK. The main cabin fuse box, fuse "15" (also used for the aircon fan, if fitted), rated 10 amps- also OK. So the main controlling fuses are okay- so that leaves the sensor/ fan motor and ECU. May replace the sensor- anyone know the part number/cost? Oh heck, I'll phone up the spares dept tomorrow! Rob |
| Rob Bell |
| Rob, You probably already know but the car has two temperature sensors- one for the ECU and one for the temperature gauge. The rad fan is controlled by the ECU. So either of this sensors could be faulty ie you could be getting a false reading on the gauge because its sensor/connections are bad or the ECU could be getting a wrong temperature reading because its sensor is faulty! The following is a reply from Nick Adams a Lotus Engineer on a related question on the Lotuslife BBS. I think you will find it very usefull: "Hi Per, right, I now have access to the manuals and may be able to shed a little light. There are two temperature sensors fitted to the engine, one which is orientated vertically and which controls the temperature gauge on the dash and one which is orientated horizontally and provides coolant temperature information to the engine management system. The EMS should switch the radiator fan on when the coolant temperature reaches 102 degrees C with the engine running and off when the temperature drops to 96 degrees. When the engine is turned off the fan will switch on automatically at 112 degrees and off again at 106 degrees. The sensors are both resistance based devices which offer a lower resistance as the coolant temperature rises. The normal resistance at 88 degrees C coolant temp should be in the order of 300 to 400 ohms. Any increase in the earthing resistance due to corrosion or excessive sealant will make the sender output reflect a lower coolant temp than is the case. This would give the symptoms you describe. To check if this is the problem run an auxiliary earth line from the body of the horizontal sensor to the negative terminal of the battery. If the fan switches at the right temperature you simply need to remove the sender, clean the threads and reasemble with a little thread sealant. If the fan still switches incorrectly try replacing the sender unit itself. Good luck, Nick" Spyros |
| Spyros Papageorghiou |
Hi all, I am at the moment testing a small wiring exercise ending up in a "panic switch " ! This simple setup enables both engine bay fan and front radiator fan to be engaged at the flip of a switch. If left in that way after parking and switching off the engine the fans will run for a few minutes.... This wiring does in no way interfer with the normal procedure commanded by MEMS -unit. After some prolonged testing this easy mod. will probably be at Dieters Tech. site if he aproves ! MGB owners will know what I have done.. Regards , Carl. |
| Carl |
| Rob have you looked at the condition of your radiator ? |
| Jon Baker |
| Hi, Some interesting stuff here. Carl, switching the radiator fan on with the engine off will not do a lot as there is no thermo-syphonic coolant circulation (unlike the B!). However you know this already and I'm just nit-picking as usual. One aspect which I am theorising about (as ever) is the open bypass pipe. When the engine is at low revs the coolant circulation rate is reduced, and the bypass pipe offers the easiest path whether the thermostat is open or not. I wonder whether some engines can suffer from this open loop circulation and possibly have an occcasional fast and uncontrolled rise in temperature at tickover? There is probably some other combination of circumstances with the breather pipe and heater circuit too. This may explain the sudden burst of coolant from the expansion tank that has been experienced by a few after a fast run. Of course if this were always true then the radiator fan wouldn't come on at tickover, but then no theory's perfect. Regards, Kes. |
| Kes |
| Rob, You are experiencing EXACTLY what I have been suffering since my 36K service. My problem has never been sucessfully diagnosed, and I've put it down to air in the coolant system - which has now rewarded me with HGF. Would be most interested if you do find an electrical or mechanical problem. Paul P9 VLS |
| Paul |
Hi Kes, yes, I agree 100% on what You say. The thought about also letting the main radiator fan work is that this can give some additional cooling even if there is no or minor circulation in the system. Probably the nicest thing is that You get a confirmation that both fans are working OK ! About the cooling system in general; yes, there is something dodgy in that system and itīs bypass etc. Why does a lot of HGF and general overheating occur at this time of the year ??? My thought is that now when we slowly are aproaching warmer weather people tend to shut the best bypass there is in the system , the heater valve !! We have a IR remote temp sensor at work and when time and wether permits I will take some measurements on diff. engine parts and coolant hoses just to check how efficent the system is at idle etc. Best regards , Carl. |
| Carl |
| Paul, I certainly keep you posted WRT investigations on going. Jon- only a cursory inspection of the rad- looks okay. Why do you ask? Spyros- actually, no, I wasn't aware that there are two temp sensors. However, the apparant non-operation of the fan makes me suspect the ECU sensor- especially as its failure has the severest consequences. Thanks for the information you've provided, it should be enough for me to make a dianosis. Carl- another nice idea- as usual! Look forward to reading abot it. Cheers Rob |
| Rob Bell |
| Rob - You are experiencing exactly what happened to me last summer. eg temp gauge went up in traffic but never when moving, no matter how hard the driving. Also you never hear the main rad fans running but the engine bay fan is working overtime. Problem was the sensor, changed in about 10 minutes by the mechanic. Checking is easy. When engine is hot and you come home leave it running for a few minutes. If the sensor and fans are working the rad fans should start up very quickly. If you are outside the car you will easily hear them. They will come on well before your gauge will move. If they have not come on in 10 minutes of tickover running then there is a problem. The lack of liquid around the block on the engine means temps rise very fast. If the fans don't come on to test that the fans work just close the fan relay under the bonnet (I can't remember which one it is). If the fans run then the fault is either the sensor or MEMS. Most likely the sensor. Sensor cost last year was about Ģ20. Mechanic fitted it from above (lift back of hood and remove cover plate) by feel without draining coolant. He first checked pressure cap was ok then just unscrewed the sensor and put the new one in. Probably less than a spoonful of coolant leaked out. Because we were linked up to the testbook at the time it was immediately apparent on restart that the temp was rising rapidly and at 102 degrees the main fan started. It stopped when the temp had reduced to 96 degreees. The old sensor had failed at 78 degrees. Not enough to seriously upset the mixture but looking back over fuel consumption records it was easy to see that the car had been 1 to 2 MPG down for a while from average. (MEMS thinks engine cold so putting in more fuel). With the new sensor consumption immediately restored to normal. By the way my temp gauge before this was fixed had moved up to one division below the red portion, so by my reckoning the engine had got seriously hot seeing even when the fans start at 102 degrees the temp gauge has not moved. However I never lost any coolant and have since put another 8000 miles on the clock without problems. I do however make sure the fans are working now regularly, just by leaving it running for a while after a run and hearing them startup and stop. It does provide peace of mind. Hope your problem is this simple. Regards, Alan |
| Alan |
| Thanks, I hope so Alan. Just being a bit dim, how did the chap close the relay? Did he just short it? Maybe I'll invest in a new sensor, fit it, and see what happens... |
| Rob Bell |
| Hi, Carl, if we make some assumptions for the sake of discussion, and they are that a) the K-series head gasket and head/block joint combination is a little fragile, and b) that there is a higher incidence of gasket failure in the F than in a front-engine installation (and they are not immune, as I know to my cost), then we could try to establish why. Whether we would be any better than Rover at this is another matter. There appears to be a number of peculiar characteristics of the F's cooling system which distinguish it from a front-engine installation. There's no thermo-syphonic circulation to the radiator, so there's no 'buffer' for the coolant when the thermostat is closed, and no heat-sink when the hot engine is turned off. There's twice the amount of coolant to circulate (10.5 litres). I don't know enough about kinetic energy in closed systems to know whether this is a problem or not. There's much longer pipework creating more resistance to flow. Then there's the open bypass pipe enabling circulation through the engine when the thermostat and heater valve are closed. As this is always open it forms a short loop which the coolant will always whizz round whether the engine is hot or cold. I'm not the only one who notices that the heater performance varies according to engine speed. As for that damn breather pipe, to return '..any excess coolant created by heat expansion..' to the expansion tank - my eye it does. The thing runs almost constantly, adding to the bypass pipe problems and introducing air bubbles into the coolant. As much coolant is going back into the system as is being pumped into the expansion tank from the breather pipe, so in go the air bubbles. Then there's the radiator fan being controlled by a sensor in the engine block! Hmmmm. Which, if any, of these causes a problem, and what that problem is, is open to conjecture. I guess one or more (or another not identified) tends to push the head joint past the point of no return. I too always have the heater valve open when I start the engine, hot or cold, and only close it when the engine is good and hot. The theory is that this will give the greatest coolant circulation when the thermostat is closed. It's probably better to close it when the engine is hot so that the coolant is forced around the radiator circuit! Oh yes, does anyone know why the coolant pipe from the engine (the hot stuff!) goes to the bottom of the radiator and the return pipe (the cooled stuff) is taken from the top of the radiator? And whilst I'm at it, can any of the lurking Elise owners provide a detailed layout of the Elise's cooling system? Yes, every nut and bolt, please. Well, what would we do if we didn't have something to worry about? Regards, Kes. |
| Kes |
| I only ask in case you problems are due to wear and tear rather than sensor faults or air locks. If the fins on your radiator have become damaged or clogged through time then of course the radiators efficiency will drop and there will be no tell tale leaks. Sorry about the HGF Paul ! Hope you hadn't already had your timing belts changed then at least you can save some money on the 60,000 mile service. |
| Jon Baker |
Hi Kes, yes there are indeed some very akward soloutions in the cooling system of our beloved "F" ! I know of one guy here in Sweden that I intend to contact. He has had a complete rebuild of the cooling system ! AFAIK he took out the original thermostat and have instead the thermostat(the type with an inbuilt "jiggle valve" ) on the outlet integrated in the hose. Donīt know if he also rerouted the IN/OUT to front cooler. Anyway this was done after he had discovered that the car could not be left idling for more than 6-10 min. before overheating occured. One can just wonder if the original waterpump is up to the job shifting around that amount of water either at low or high RPM .. If I get any feedback from him I will respond here, Regards , Carl. |
| Carl |
| Rob - To close the relay he had a piece of insulated fairly thick gauge wire in a horseshoe shape which he pushed into the relay housing to short across it. It looked home made (the insulation looked like 4 or 6mm thick heat shrink tube). Changing the sensor is definitely the cheapest, easiest and most probable solution. Apparently these things fail quite a lot according to the man. Regards, Alan |
| Alan |
| Jon, the rad seems okay, but I'll have a closer look... I drove the car to the Half-way house meeting last night. No signs of over heating- although I seem to be convincing myself that the oil temps are a few degrees higher than normal (paranoia?) Parked the car, left it running. Waited and waited. No rad fan, no engine fan. Internal temp clocks read fine. Okay, it was cold- probably approx. 5 degrees celcius, but SURELY the fans should have been triggered? Hmm. I can see a need for a new sensor and further investigation. Rob |
| Rob Bell |
| Thanks Alan, I reckon that is what I'll do... "common things happen commonly" as they say in my trade! Rob |
| Rob Bell |
| had a new inlet manifold gasget,new head gasget, re-sleeve to replace cracked sleeve,new radiator core,still the damn thing overheats in traffic. think the only thing i can now replace is the car |
| howard |
| Just bought a new water temp sensor- part# GTR206 (cost 10.99 including VAT). Alan- is this the same item as you had replaced? Either I am paying much less for the part than you did, or it is the wrong one! (might it be GTR185?) Fingers crossed that it will be a dry warm weekend! Will have camera at the ready!!! ;o) Cheers Rob |
| Rob Bell |
| Rob, I don't know the part numbers but the ECU temperature sensor has a BROWN body and is mounted horizontally. The temperature gauge sensor has a BLUE body and is mounted vertically directly below the ECU sensor. Both are located on the engine outlet pipe at the flywheel side so you will have to remove the engine cover for access. Replacement is easy- Make sure the engine is cold, disconect the battery earth, disconnect the harness from the sensor, put an old cloth under the sensor to catch the small quantity of coolant that will escape, uscrew the old sensor, clean the threads and screw the new sensor in tightening only to 6 Nm. Don't forget to top up the coolant at the expansion tank - no air bleeding is necessary. You may want to apply a very layer teflon tape to the threads of the sensor before screwing in. Make sure it is very thin though so you will still get a good earth. Spyros |
| Spyros Papageorghiou |
| Thanks Spyros- the sensor that I now have in my posession is the brown probe that you describe. I'll keep you all posted as to whether this "fix" works or not. Cheers Rob |
| Rob Bell |
| http://www.lame-delegation.de/mgf-net.de/wreck_engine/dkcp_3909.jpg Far below on the right side of this picture, is it ? I'm sorry, but also havn't got the number. May be Rog can find it quick. Just drop him a line. Dieter |
| Dieter Koennecke |
| That looks right to me. Anyone else have a comment? Cheers Dieter Rob |
| Rob Bell |
| Yes, check out what I have written in, 'Coolant -auto air bleed check' thread. In addition feel over the whole rad for even temps. variations indicate blockages or air locks (difficult to reach I know) The fan takes an age to operate, but you may want to speed things up by holding 2 to 2500 rpm for 5 to 10 minutes to put sufficient heat and circulation through the system to awaken the fan. Rog |
| Roger Parker |
| Thanks Rog- yes, I read that post with interest and will check it and re-bleed the cooling system anyway. Need some new antifreeze- should I get this from the Rover parts dept? |
| Rob Bell |
| Kes, As another "lurking Elise owner" (and former MGF VVC dito) I can confirm what Spyros has written above (Nick Adams' comments were in response to a question I had asked on the Lotus Life BBS). Actually, I am just about to replace this brown temperature sender myself this week-end on my 111S, having experienced similar temperature rises in my car). An interesting point that we Elise owners can contribute with, having very precise digital temperature gauges, is that the deterioration of the sender seems to be gradual. When my car was new, and during the first months, the radiator fan would always kick in at an indicated 104C. After about a year it started to do so only at an indicated 107C and then, on one occasion, at an alarmingly high 120C (which is why I turned to Nick Adams). In other words, it appears that a gradual build-up of deposits and/or corrosion on the sender gradually alter the quality of the signal sent to the MEMS and, thus, the point at which the fan kicks in. Maybe a routine yearly change of this sender is in order...? Per |
| Per |
| Or two years/ 24000 miles? Interesting obseration Per (I recognised your name from the cut&paste above, BTW!). How many vehicles is this observation based upon? |
| Rob Bell |
| >Need some new antifreeze- should I get this from the Rover parts dept? Can't suggest a source, but recall to get the SAME STUFF which is already in. No mixes allowed and the colour of the stuff is no evidence for equals. |
| Dieter Koennecke |
| Hi, Per, thanks for the comments and carry on lurking: like a few others I occasionally lurk on the Elise board (and even posted once!). Now I have your attention there are a few points about the Elise's cooling system I'm curious to resolve. If you could possibly answer them I'd be grateful. Actually I seem to remember reading that Rover and Lotus worked together at Gaydon to develop the Elise's cooling system. Anyway, Does the Elise have the very short bypass hose linking the cylinder head manifold with the heater return pipe? Does the Elise have the high-level breather pipe from the same bypass hose area feeding back to the expansion tank, and if so does it seem to circulate coolant most/all the time? Do you know of anything peculiar to the Elsie cooling system that isn't present on the F's? Is there a diagram of the Elise cooling system components which you could scan and mail to me, or host somewhere? You see, lurk around here and someone gives you a job to do! Rob, yes, I would only use the Rover/Unipart antifreeze (which I believe on your F is Unipart Superplus 3). It's just not worth using anything else. The Kidderminster Rover dealers hadn't even heard of Superplus 3, they only went up to 2. I don't know what the MG branch at Bromsgrove use, I'll ask them next time I'm in. Oh yes, I'd get a gallon of distilled water from Halfords too. I also read (I do a lot of reading) from various manufacturer's websites that it's the additives that go, not the Ethylene Glycol, that you can't easily test the additives, just the coolant concentration, and that a reputable coolant will last far beyond the 2 yr change period, up to 5 years or 50,000 miles quite easily, so I don't think that there's any great panic on late changes. Why don't they ask these questions in pub quizzes? Regards, Kes |
| Kes |
| FAO Rob, I just got the bleed document updated after some researches and help. Got all in and updated. http://www.lame-delegation.de/mgf-net.de/bleed/index.htm The picture I mentioned was the right position of the stub ! I only was confused until I got that the picture was taken from my disassembled engine. The small 17mm hose to the coolant expansion tank is already removed and the location of the oil dipstick assy looks also not common, cause its loose and swaped to the left side. This auto bleed term is a vital one as mentioned already by Roger after EACH coolant drain down. Didn't Carl also write about it already several month ago ? Guess that some more websides need an update though. And probably HGF reasons need to be updated also. :( Notice: This term is currently NOT included in Workshop Manual RCL0051xxx Issue 7 !! So earliest appearence may be in Issue No 8, dated 2001 !! My humble comment on future coolant changes if NOT done at actual MG-Rover registered workshops.... Oh, chaps we will need a new laaaaaarge HGF HOS for you. :( Should be named HOI = Hall of Ignoring HTH Dieter |
| Dieter Koennecke |
| Rob, I'm afraid I only have my own experience to rely upon on this one. There really seems to be much less discussion about HGF and cooling issues among Elise owners. This may of course in itself indicate that engine bay ventilation/cooling is better in the Elise (and is the root of the problem in the MGF). Kes, I will look into all this during the week-end and I'll get back to you next week. Cheers, Per |
| Per |
| Hi Per, wasn't it you that put extra sealant around the 111S headlight covers to prevent so much condensation & muck getting in? To avoid sabotaging the MG board & turning this into an Elise thread, could you drop me an email to let me know if you can recommend this treatment? Cheers, Dot |
| Dot |
| Dieter, thanks. :o) I'll look at the auto-bleed valve function at the weekend, and perhaps take some more piccies with everything in-situ. Worrying that it isn't mentioned in the workshop manual isn't it? :oO HOI indeed :o( Rob |
| Rob Bell |
| Coolant temp. sensor MEK 100170 is the missing part No. just found in the workshop manual. Torque to tighten is 6 Nm Good luck for the weekend works. Me is off for a long weekend CU Dieter |
| Dieter Koennecke |
Hi all, Yes - as well as cooling /bleeding problems I think we have to fight the "sensor syndrome ". Unfortunatly the used sensors ( NTC or PTC ceramic covered with thin metallisation for bonding the connection cables ) are prone to deterioate with age. This is why they never are used in the industry ,at least not where one is depending on accurate readings ! There IS a series of semi-profesional sensors (the small bead itself and connecting cables) that maybee could be implemented into an old sensor housing. If You want to see what I mean , just unplug the engine compartment air-temp. sensor ,remove it and have a look from the outside how the bead is mounted ! No way this thing will survive more than a few years.. regards , carl. |
| Carl |
| >>Coolant temp. sensor MEK 100170<< Oo-er, no I am worried- the sensor I have has the part number GTR206. It has a brown plastic housing, with a brass cylindrical probe. It has two electrical connectors. Same thing? I wonder when a complete sensor re-fit becomes necessary to avoid 'sensor syndrome'? Probably at the 5 year mark? |
| Rob Bell |
| Just replaced the sensor- and took some pictures doing it. Unfortunately, a little more than 'a few drops of coolant' came out. More like 300 + mls infact. Cause to be worried? Will post images later- perhaps this evening if I get a chance to work on it this evening... Shall have another go at bleeding the cooling system- if I can get hold of unipart antifreeze. :o/ Not sure if Halfords stock this... Hope this works! Cheers Rob |
| Rob Bell |
| Best of luck Rob, I can read up the bleed info from my NEW workshop manual if you need it? |
| Paul Lathwell |
| >MEK 100170 Entry from Edition 7 WS-manual. Don't worry. May be that is the new. GTR206 ...hmmm... colour is Ok. Lets see how it works. Rob, if you can't get the coolant just now, .... clear water is better than nothing. You can exchange it later by getting out one filling of the expansion and refill with pure coolant. Dieter |
| Dieter Koennecke |
| Paul, I'd be VERY interested in what it says (as I am sure would Dieter, if he doesn't already that is!)- thanks. Dieter- good suggestion- I'll top up with water for now, and bleed again later with pure antifreeze. Sadly got to the dealership shortly AFTER the parts dept had closed! Still, reconcilled myself with a quick test drive in a 200 BRM. Hmm nice! Yimmy's car could be about to be traded in ;o) Rob |
| Rob Bell |
| "Unfortunately, a little more than 'a few drops of coolant' came out. More like 300 + mls infact. Cause to be worried?" That's strange. Was the engine completely cold (ie overnight)? Did you have the expansion bottle close when removing the sensor? Only a little coolant came out of mine when I did it. I would not worry though - no air should have entered the system as long as the expansion tank was not completely empty. Any air that entered through the sensor hole will have escaped to the expansion tank through the jingle valve and the top bleed line. Is the fan working now? Spyros |
| Spyros Papageorghiou |
| Rob, I'll get back to you with the info, a little tied up right now. :( |
| Paul Lathwell |
| Not quite 'stone cold' Spyros- probably the explanation. Thanks for the kind words of reassurance! Paul, no worries mate. Let me know when you can :o) |
| Rob Bell |
| Rob, I have no mention of that specific MEK prefixed number on the July 99 parts slides, only a close number relating to the 2000MY oil temp sensor. GTR 206 is listed as the correct one for your application. There is a later dated parts slide, which I haven't got, but your replaceing like for like, so no problem. Coolant will escape in volume from the sensor stub whne it is removed if there is any residual pressure in the cooling system and also if the expansion cap is removed. Rog |
| Roger Parker |
| Why not remove the cap at the start of the job to release the pressure then replace the cap to stop coolant loss? (assuming the engine is cold and will not build up any additional pressure that is) |
| Stuart Elliot |
| Stuart, of course the right way. (IMO) Also release the coolant from the axpansion tank before the thin 17mm hose of the stub gets pulled off. Anyway, lets see Robs report. Dieter |
| Dieter Koennecke |
| Well, I just couldn't leave alone could I? I chamged the spark plugs at the same time, and now the engine is missing slightly... Hurrmph! Not sure of the cause, as the spark plugs (K&Ns as I have a set to try out) were correctly gapped. I'll whip them out again to re-check, and probably replace the ignition leads at the same time. Upshood- the engine hasn't been run long enough to get it warm enough to trigger the fans- but at least the idle is more reasonable. Thanks for all the advice Rog- the GTR206 probe fitted nicely and is identical to the one I removed. More updates to come I think... |
| Rob Bell |
| For those interested, I can tell you that the replacement of this sensor (Rover part number GTR206), which I did yesterday, appears to have solved the "fan not kicking in" problem on my Elise 111S (VVC). I went for a long drive afterwards, to get the car really warm, and then stopped while letting the engine idle. The fan kicked in at exactly 104C, bringing the temperature down to 93C, just as it always did when the car was new. :o) Per |
| Per |
| Per, I am going to check the resistance range of the probe I removed- in ice cold water and boiling water to confirm that it was a sensor glitch as opposed to a problem with an electrical connection. Fancy doing the same? It'd be interesting to see whether there is commonality in the failure. |
| Rob Bell |
| Rob, I'm afraid I wouldn't know how to do that... but I can send you my old sensor if you like. In any case, mine hadn't failed completely, it had just become too "optimistic" regarding the temperature of my coolant and, thus, not bothering to send the signal which would activate the fan. According to my Lotus workshop manual, a sensor which has given up completely results in the MEMS leaving the fan on all the time. Per |
| Per |
| Do you have access to a multimeter Per? All I planned to do was measure the resistance across the two terminals at close to zero (ice cooled water) and at 100 degrees (boiling water). Frustratingly, I wanted to do this with a new sensor- but couldn't 'borrow' the work multimeter in time... :o( |
| Rob Bell |
| Interesting point from Per - fail safe on the Elsy is fan ON - is it for the *F*? Ted |
| Ted Newman |
| Nope- the thing doesn't (didn't) come on in my car. The problem may be the 'default' signal from the sensor- if stuck 'low' the ECU would be fooled into thinking the engine was cold- as has happened in my case. If the sensor signal is stuck on 'high' then this would fool the ECU into thinking that the coolant temp was persistently high, and run the fans the whole time. This should be apparent from measuring the resistance across the probe terminals and observing the change of resistance in cold & hot conditions- or lack thereof. Rob |
| Rob Bell |
| Rob, Per, I wonder whether it will really deliver an analogue signal but a digital ON/OFF with the known range of hysteresis. Such cheap made but expensive sold sensors are often only made from Bimetal' (bends under temperature) A hacksaw will give you the last evidence ;) |
| Dieter Koennecke |
| From the stuff that has been archived, and the quote from Nick provided by Spyros ("The normal resistance at 88 degrees C coolant temp should be in the order of 300 to 400 ohms") it would appear that the sensor actually provides a measurement of the water temperature. Therefore, this enables the ECU to act as the switch- as opposed to the probe performing this function. Rob |
| Rob Bell |
| As quoted above in Spyros' message: "The sensors are both resistance based devices which offer a lower resistance as the coolant temperature rises. The normal resistance at 88 degrees C coolant temp should be in the order of 300 to 400 ohms. Any increase in the earthing resistance due to corrosion or excessive sealant will make the sender output reflect a lower coolant temp than is the case." In other words: the MEMS interprets the resistance provided by the sensor. As corrosion/deposits increase on the latter, it will provide higher resistance which the MEMS interprets (wrongly!) as "fairly low temperature - no need for any fan". The system enters "sensor failure mode" (i.e. Fan On all the time) only if no or ridiculous resistance values are recorded (i.e.when the sensor is clearly broken). It will miss the more common (and much more dangerous!) situation in which the increasingly corroded sensor fools the MEMS into believing that the coolant temperature is 15 or 20 degrees lower than it really is. Per |
| Per |
| Based on these comments Per, one would expect that your temp sensor to record an exceptionally high value, whilst mine would record a high, but inappropriate value and a 'good' sensor a lower value. Unfortunately, we do not have a reference range for the 'home mechanic'- and that could be very useful for diagnosis of a faulty sensor. Rob |
| Rob Bell |
| Rob, The workshop manual gives detailed figures for the coolant temperature GAUGE sensor but I cannot find resistance figures for the ECU sensor. It says the following though in the cooling section: "The temperature of the cooling system is monitored by the ECM via signals from an engine coolant temperature sensor, which is mounted in the cylinder block, outer elbow. When a temperature of 102C is reached the ECM switches the fan on via a relay. The fan switches off at 96C." The Elise manual however (I only have parts of it) says "The sensor is a thermistor (a resistor that changes value with temperature).......Low coolant temperature produces a high resistance, whereas high temperature causes low resistance. At normal running temperature, the resistance will be approx. 300-400 ohms." So, to answer Dieter it is a thermistor and you should be able to check it provided you have a thermometre that goes over 100C. Ted, I believe that the control units of the Elise and the F are identical so the fail safe should be the same for both. Spyros |
| Spyros Papageorghiou |
| Posted before reading Per's message, which I believe is a very accurate description of what happens when the sensor gets older and corrotion builds up. Spyros |
| Spyros Papageorghiou |
| Looks like we'll need to 'DIY' a reference range then! Thanks for looking Spyros :o) |
| Rob Bell |
From a general engine management book : Temp Resistance Voltage 0 4.8k-6.6k 4.00-4.50 10 4.0k 3.75-4.00 20 2.2k-2.8k 3.00-3.50 30 1.3k 3.25 40 1.0k-1.2k 2.50-3.00 50 1.0k 2.50 60 800 2.00-2.50 80 270-380 1.00-1.30 110 0.5 Open Circuit 5.0 +- 0.1 Short to Earth Zero Some comments to note : "Poor contact and corrosion are common reasons for inacurate signals from the CTS" Specific to Roven MEMS : "The ECM uses the CTS signal as a main correction factor when calculating ignition timing and injection duration" Per and Spyros are exactly right Rob. Go test the 'broken' sensor with your borrowed multimeter. Let us know how you get on. :-) Cheers, Paul. PS. You can test the new sensor in-situ by using pins 30 and 33 at the MEMS end. Be v.careful though. :-) |
| Paul Nothard |
| >>PS. You can test the new sensor in-situ by using pins 30 and 33 at the MEMS end. Be v.careful though. :-) << Or maybe not! ;o) LOL If broken circuit = infinitely high resistance, then this should mean that the fans kick in as part of the fail/safe? Okay, will try and check out the sensor readings and see if they correlate with the data you've found... Are you able to do the same Per? |
| Rob Bell |
If one or more sensors develop a serious fault then MEMS goes into LOS mode. LOS = Limited Operation Strategy, or Limp home mode. When the sensor is "detectebly broken", MEMS assumes that CTS (Water) is 60 degC, ATS (Air) is 35 degC and engine load is based on RPM. This MEMS book says nothing about fans. :-( I haven't got the MG tech manual handy, but I'm sure that if MEMS gets either 0v or 5v then the fan will kick in and a fault flagged. There is the complication that some relays are involved and faults can be there. Can't remember the details. :-( P. |
| Paul Nothard |
| Bloody hell - all that to switch a fan on! Come back constant V-belt driven fans and a pair of ladies tights in case the belt broke. Ted |
| Ted Newman |
Aha! Found the book. Says nothing about the cooling fans going on other than temperatures: On at 102 degC; Off at 96 degC. It would be stupid not to turn the fan on if the sensor failed. But it does go on when there is no voltage. I had the sensor replaced after my cooling fan was on all the time. Ted, leaving the subject of your tight aside... <hehe> You think the fans are tricky? You should read up on this MEMS m'larky. Really very interesting how it all works. :-) P. |
| Paul Nothard |
| Grumble, grumble ;-) OK, so a NTC- thermistor, but at last I would like to see one of both from inside... Hacksaw, hacksaw !! May be this data are of any help: GLASS ENCAPSULATED THERMISTORS (may be). The type in row 'G' looks nearly like ours. http://www.wuntronic.de/sensors/seriessd_e.htm |
| Dieter Koennecke |
| Okay, problem solved :o)) The rad fan now kicks in as designed, the idle rpm is spot on, and all horses have been safely returned to the stable. Fan-bloomin'-tastic! :o) The uneven idle and occasional miss was traced to those new sparkers I fitted- I shall inform K&N about this problem in due course. All that remains for me to do is to measure resistances in the 'failed' thermistor sensor to confirm component failure- or alternatively a poor electrical connection. Cheers all Rob |
| Rob Bell |
| Fantastic :))) Congratulations !! Got your bleed valve pictures also hosted to the bleed page. See some famous but dirty petrolhead fingers http://www.mgf-net.de/bleed/ Dieter |
| Dieter Koennecke |
| Rob, I'm glad you car is fine as well now! Unfortunately I don't have access to a multimeter but, as I said, you can have my old sensor if you like. Per |
| Per |
| Okay, no worries Per. Could you drop it in the post, and I'll check it out? Drop me an e-mail, and I'll give you my address. Cheers Rob PS I've knocked up an experimental heat shield for the 57i K&N- and I can't believe how well it is working- the improvement in mid range torque is terrific! |
| Rob Bell |
| Paul I as an ex motor mechanic like to see what is happening, that is to say things going up and down or round and round - makes life nice and easy - today everything seems to be controlled by solid state computers and car computers seem to be like all other computers, when they go wrong they ask you 'are you sure?' and then go into a sulk! Ted |
| Ted Newman |
| > PS I've knocked up an experimental heat shield for the 57i K&N- and I can't believe how well it is working- the improvement in mid range torque is terrific! C'mon then Rob, details please :) BTW, I saw pictures of the new K&N induction filter in the new Demon Tweaks Catalogue (came free with "Car" Mag), incase noone has seen it, it appears to just add a sealed box around the cone, so that only air from the induction pipes actually goes into the filter. I might knock something together myself to do this. I'll make a sacrifice of a lunch box for it I think ;) (And wrap it in a few layers of ali foil to try and make a heat shield.) Steve. |
| Steve Childs |
| Paul found, Temp Resistance Voltage 0 4.8k-6.6k 4.00-4.50 10 4.0k 3.75-4.00 20 2.2k-2.8k 3.00-3.50 30 1.3k 3.25 40 1.0k-1.2k 2.50-3.00 50 1.0k 2.50 60 800 2.00-2.50 80 270-380 1.00-1.30 110 0.5 Open Circuit 5.0 +- 0.1 Short to Earth Zero fan On at 102 degC; Off at 96 degC. ---------------------------------- Had a discussion with a Rover garage this morning. Was told that it could be an advantage to let the frontcooler go ON at lower temperatures. Worth to add a resistor or dampening circuit ? to decrease that range for -20° ? Any suggestions ? Or a new thread ? Dieter |
| Dieter Koennecke |
| But Ted, Can you imagine the convoluted route a fan belt would have to take on the F for the engine to drive the fan directly (or were you thinking of it coming straight through the cabin :-) ) Wasn't it the introduction of the transverse engine that "killed off" direct drive fans (I can't think how a mini fan was driven - I presume they were electric...) Neil. |
| Neil |
| Checked the resistance range of the sensor I removed from my car last night. At 90+ Celcius got a reading of 300 Ohms. At 20 Celcius got a reading of 2k Ohms. Looks as though the sensor was okay after all, and the problem was a corroded electrical contact. No problem- at least I have a spare sensor now! Interesting question Dieter- possibly a subject for a new thread? |
| Rob Bell |
| Rob, were there any "deposits" on the probe? Just a thought, Calcium (whatever) Carbonate etc (from hard water) doesn't conduct heat terribly well. I wonder if cleaning the bit that's in the water would change the readings? Neil. |
| Neil |
| Yes Neil, there are. I haven't removed them when I tested the probe though, so this is not likely to be the explanation in my case, but a good point duely noted for a write-up I have planned. :o) |
| Rob Bell |
| Dieter, To achieve what you are referring to, you would need to LOWER the resistance! But I wouldn't do that, since the sensor's temperature reading not only controls the fan but also plays a part in ignition timing and fuel injection settings. Per |
| Per |
| Per, thanks for the reminder on related controls !! Lower resistance is clear, but in any way not for the high OFF resistance. So I left out in my mind just from the beginning a parallel simple resistor. I fear this problem is to big for us :( Or any electronics specialist can find a solution. Hmm, what can get influenced on ignition time and fuel injection related to the - coolant temperature... That crap sensor can give only 'true' output between 0 and app 100 degress, cause this seams to be sensor type related. See the resistance curve, there is nearly no important change at temperatures below 20°. Is that the only sensor which has influence on - cold start - warm start - idle in cold condition - idle in warm condition - engine in hot condition .. and what else ? |
| Dieter Koennecke |
This thread was discussed between 18/04/2001 and 04/05/2001
MG MGF Technical index
This thread is from the archive. The Live MG MGF Technical BBS is active now.