Could be the temp sensor at the back of the engine block has or is going bad.

 

I am still on the learning curve with my 530i after years of working on older M50, M52, and S52 engines. Temp sensor on those engines was towards front of the engine block buried under the intake manifold. Very hard to get at them with the manifold installed. The last production versions of these older engines went to a single sensor with a dual element inside it. One sensing element was for instrument panel and the other was for the DME. Is the temp sensor on the back of the M54 engine a dual element like the M52/S52 engines? Is it under the manifold and impossible to get to without pulling the manifold?

 

I had thought about that temp sensor, but a few things confuse me:

a) Everyone I've seen who has replaced this sensor has reported far more erratic temp issues with the guage as a failure mode - ie it went to max instantly, or it stayed on cold all the time etc.
b) People seem to get a fault code logged mentioning this sensor - my code memory is clear
c) Mine appears to be uniformally 'about 15c too cold' which just seems a bit... convenient?

It's a $20 part so it'd be cool if thats all it was. But there isnt a huge amount of info on the net about replacing it and whether I can expect coolant to leak everywhere when doing it, etc etc. Or whether there is a useful test I can perform before I swap it out?

So I monitored the temps this morning. 15 minute drive from cold. Temperature slowly rose to about 72c in a uniform fashion and from then on dipped between 72 and 66ish. Even crawling through traffic or stationary at the lights didn't stop it cooling down,

I also watched the coolant outlet temperature whcih I beleive is the sensor on the outlet pipe for the radiator. This stayed very cold (15c) for about 5-10 minutes until it then steadily rose to more or less match the main coolant temperature. Surely this is the point at which the thermostat opened and allowed coolant through the radiator? This seems to indicate that whilst pehaps the functionality isn't correct, the thermostat itself is not stuck.

I'm really at a loss. Everything sensible points to thermstat but I'm just not sure thats the problem - given my findings above, the fact it was replaced yesterday AND the fact the one removed from the car is not stuck in the open position - its closed. If it was stuck closed it would overheat surely not overcool?

The other thing - would you expect the outlet temperature and the coolant temperature to equalise when the thermostat is open? I would - and thats what seems to be happening. Therefore if the main coolant temp sensor was faulty and giving me inaccurate low readings then why does the outlet temperature agree with it? Surely the outlet temperature would be recording a higher temperature than the faulty main sensor if that was the case.

Other than a faulty thermostat, what else could cause actual cool running rather than a faulty temp reading?

 

Fox, I may be stretching the analogy too far but here is my theory about what you are observing. I am a EE and spent the early years of my career working on gas turbine engine instrumentation for Navy aircraft. We would occasionally have problems with abnormally low temperature readings from the exhaust gas temp sensors on a particular GE engine that was caused by intermittent short circuits. The reason was the the engine controller was averaging the sensor readings so when the voltage dropped to zero during a very short duration short circuit, it lowered the average reading. The ECU did not detect the short circuit because it was not a hard, constant failure. When we measured the resistance of the circuit with the engine shut down, it measured normally. The short circuit was only happening with the vibration that occurred during normal engine operations. The fact that our BMW temp gauges are 'buffered' suggests to me that the gauge cluster computer and/or the DME are programmed to average the temperature readings. A repetitive, short duration short circuit under these conditions would give you abnormally low readings.

I just re-read your post above and noted that the sensors in two different locations show similar readings. That makes my theory above very remote possibility. Odds of both having short circuits is nil. Are you sure the other temp reading is from a sensor on the outlet pipe for the radiator? I thought that was just a thermo switch for the aux cooling fan. If the two temp readings you are getting are from the two elements in the dual temp sensor, then my thoughts about intermittent short are still valid.

My 530i has a failed thermostat right now according to the fault codes I can read with my Auto Enginuity OBD II reader. I have the Enhanced BMW package which allows me to read proprietary BMW fault codes that the generic code readers don't pick up. It is telling me that my thermostat is failed in the open position. My gauge still reads in the middle once the car is warmed up. I will have to check running temps while driving and report back to you.

What are you using to observe temperatures? The gauge cluster test mode or an OBD II code reader?

 

The engine temp sensor that sits in the back of the M54 engine is in fact very good and rarely goes bad.

Chances are:
1. Your new thermostat is going out of the door.
2. The wiring between the thermostat and the ECU is bad.
3. Engine Temp sensor (which is a nightmare to get to).

Search forum for M52, M54 cooling overhaul.

 

On the M54 engine the temp sensor near the back of the engine is a dual temp sensor--along with that on the m54 there is a oil temp sensor that I had to replace recently. The back sensor is under the intake manifold at the back--but it seems doable to replace by removing the cabin duct work and all--I don't know if there is another sensor--and if CNN90 says the back sensor rarely goes bad he's probably right on that point--mine has been good to me for 245,000 miles now

 

My money is on the t-stat. You might not be getting codes now but you likely will soon.

 

The funny part is the O.P. installed a new tstat and the problem persisted.

 

Ok guys - update.

Set off for a 35 mile drive. First 10 miles. same problem. Then the guage settled on centre. Where it stayed for the rest of the journey. Came back home a few hours later. Guage rose to the middle in normal time and remained there throughout. Plugged laptop back in when I got back and went for another drive - temps in the 90's for the first time since this problem started - fluctuating between 89 and 96 - higher in traffic, lower when driving at speed, each time watching as the temperature headed towards 96ish the coolant outlet temp sensor recording a rise in temp from about 20-30 right up to near the same level - ie, the thermostat was opening and closing to moderate temperature. Exactly as you'd expect - working what appears to be normally.

It remains to be seen how it behaves after a cold start tommorrow (I'm expecting it to play up again, how can it possibly fix itself) but... what on earth? It's working normally now and has been all afternoon. I say normal - it's about 2-3c above its usual operating range (Seems to run 93-96 whereas before it was 90-93) but could this be a natural variation due to a different thermostat?

I'm still pretty confused.

 

If it starts playing up I'd consider changing the sensor to be sure. I'd expect the Radiator output temp to be noticeably lower than the engine temp, especially in England at this time of year. (The point of the radiator is to provide cooler water to cool the engine.)

Are you sure you just didn't have a bit of air trapped in the cooling system in the location of the sensor, this would cause a cooler reading. It could be this air now made it's way out and you are now getting accurate readings.

 

Are you sure you just didn't have a bit of air trapped in the cooling system in the location of the sensor, this would cause a cooler reading. It could be this air now made it's way out and you are now getting accurate readings.

When I had trapped air in my old 535i with an M30 engine, the local temperature would go up because the coolant in the vicinity of the air would boil and the steam was hotter than the water. I used to see this when I did not bleed all the air out of the system.

 

Personally, I'd leave the engine temp sensor alone. That thing lasts some 300K miles.

 

Hi guys,

Another update.

Yesterday I did about 100 miles. The car performed perfectly throughout. It heated it up normally to correct temperature and remained there at all times.

This morning it began to heat up but as it got half way between the quarter and centre mark it once again dipped back down towards a quarter. But then it resumed it's travel back to center and stayed there for the rest of the journey with a coolant temp in the normal 93c range after warmup.

This is very confusing. I almost thought it had fixed itself before its little blip this morning...

 

Rebleed your cooling system--me thinks you have some air in there

 

And another update.

Drove home. Monitored temps. Everything working fine - thermostat opening only once the car had reached full operating temp, you could see it working as the outlet temp value would rise/fall. Stopped at a shop. Went in for 5-10 minutes. Got back in car - problem returned, engine running cool. Engine was running at between 70-80 and the outlet temp was roughly the same indicating the thermostat was open. Parked car on drive. Let it idle. No change.

Turned engine off. Turned it back on. Coolant temp rose to 93ish and stayed there, outlet temp fell indicating closed thermostat. Engine began to heat up, thermostat opened, etc etc. Ie, working normally.

So, it seems I can 'reset' the problem by turning the engine on and off. How bizarre is that? I'm pretty sure it's neither the thermostat nor the temp sensor. I think it's something to do with whatever supplies signal to the MAP controlled thermostat. Sometimes it seems to send the signal to open the 'stat early when its not appropriate for it to do so.

Quite how I get to the bottom of that I've no idea!

People keep mentioning air in the system. I guess I've nothing to lose by trying to bleed it, but if there was air in the system (I am not saying you guys are wrong, I just want to understand whats happening here), how come:

a) It's overcooling not overheating
b) It sometimes works fine and can be triggered by cycling the engine?

 

Have heard of problems similar to this corrected by removing the electric clip and cleaning with cleaner then putting the clip back on--might be that simple--if it wasn't for the clip on the sensor under the intake on the back of the engine.

 

Which electric clip was he referring to?
Is there a wire near the thermostat housing that reads temp. data?

 

I think Poolman is onto something. Maybe there's an electrical contact that gets wore down, and when the engine heats up it deforms having intermittent contact. The DME will take average readings between contact (normal operating temp) and non-contact (maybe 0°C?) and you will see the temp "cool" on the OBC. Maybe I'm wrong?

 

It is certainly sounding electrical to me, bad contact or internal electric breakdown of a sensor. My experience of temp sensors is they should last for ever but every so often they fail, probably with internal intermittent bad contact.

I'd recommend:
1) Bleed the system and clean all the related electrical contacts, if that fails
2) Change the sensor to be sure, it's not an expensive part

 

I'll give that a go then - nothingt o lose!

Further testing has basically summed up for me that the problem is that sometimes the thermostat opens at about 70ish rather than 90ish. When this happens, obviously the engine runs cool. If its happening and I shut off the engine and restart, it 'fixes' itself and then operates correctly for the rest of the journey.

So, figuring out why sometimes the thermostat opens early is the task - I'm figuring it must be electrical if I can 'reset' the problem with a simple power off/on. I like your idea of cleaning the contacts on the power connector. I'm less convinced with the temp sensor (But obviously I'll try that if nothing else works) given that nothing is leading me to beleive the temperatures it is reporting are incorrect and all failures I've read about make it read nothing or ultra-wacky values rather than just 20c under where it should be.

I'm guessing it COULD be the new thermostat thats bad but that'd be a bit of a co-incidence, right? I regret not doing the same detailed level of testing on the old one before I had it swapped out. I just saw low temps on the highway, thought thermostat, and that was that. So truthfully I don't really know if the problem I have now is the same one I had before I replaced the stat or simply a similar problem thats not related to the original one.

 

If you didn't have a code to begin with you didn't have a problem with your thermostat, M54 engine will throw a code whenever thermostat is OOS.

Try to re-seat your dual temp sensor connector is easy said then done. To disconnect it you need to pry both sides of the tabs to pull out, whata****ingpainintheass are you kidding me!!!! :yikes::yikes::yikes::yikes:

 

At no point have I had a fault code, so I suspect you may be right but reading around produced tales of people with broken thermostats that didn't throw codes so I figured mine was like that.

That said it's not stuck open - it operates - just, well, sometimes at the wrong time..

 

I hear you sound and clear. I'd have done the same:thumbup:

Try to re-seat the dual temp sensor connector, I have replaced it before, it's doable.

But I think something else has triggered the DME to tell the thermostat to open wide.

 

What else would do that? It seems there are a myriad of things it can use to determinate when to open it, including engine load. How I am supposed to find out which signal is being misinterpreted and why I've no idea. Infact I'm actually seeing a scenario whereby this might not even be fixable?

 

I'm thinking the same, that load is what the DME detects and then send signal down to heat up the wax in thermostat. But with all that said, the question still remain how does the DME sense LOAD? is it by vacuum? Or RPM? Or whatever else? Honestly, I don't know.

Oh BTW my car is OFF by 10 degree C.

 

Off by 10c? How do you mean? Your temps are 10c lower than they should be?

 

No, it reads 10C lower than the actual temp. The reason I noticed this because I can hear the aux fan kicks in while my OBC test number 7 shows 80-82C, but we all know the aux fan kicks in low speed at 91C by the sensor switch on the lower radiator.

I don't know why this happens but I think it has something to do with the S/C software. Not 100% sure though. However, mine doesn't show any kind of erratic movement. Before S/C operating temp was around 94-96C, now with S/C 84-89C hardly ever seen over 90C anymore, and BTW no thermostat code.

 

The late 2000 (2000 1/2) M52 TU does have the heater in the thermostat. The thermostat is identical to those used in the M54 engine.


Sent from BimmerApp mobile app

 

OP, did you ever fix this problem? It seems that I have the same issues you described and are breaking my head on finding a solution. I know it's a long shot cause it's a post from 2012 but who knows

 

The DME indirectly determines engine load by measuring the outlet temp of the rad. That gives it the clue to open the t'stat more.

I would suspect the dual temp sensor, esp if the temp gauge fluctuates rapidly; the actual temp simply does not change that rapidly.

One side feeds the DME, and the other side feeds the gauge.

 

The temperature does not fluctuate rapidly, It has quite the steady build up (even slower than normal I would say) to that 73/74 degrees.

I will use my infrared thermometer tonight to see how hot the outside of the coolant hoses are to just get a better indication of temperature.

 

You'll get false low readings from points on the hoses near the radiator; there is a fair bit of cooling effect down the hose from the t-stat. I've found differences of 10+ degrees C between the t-stat housing and only a few inches down the upper rad hose.

For accurate readings you need to be aiming at the thermostat housing or at least the housing's nipple (that the upper rad hose attaches to.) And as close as you're able with the IR thermometer without getting clipped by the fan. The viewing cone on IR thermometers is fairly wide and if you're not quite close, the reading will be an average of the t-stat and surrounding engine bits, which will be quite a bit lower than the t-stat itself.

EDIT: Once you've taken some readings, you may find this post helpful in interpreting the results.
https://www.bimmerfest.com/forums/showpost.php?p=7721699&postcount=10

BTW, it would help going forward if you were to list the model and year of your E39. That will determine which engine you have and thus what temperature reading is normal.