Older Bmw: very strange Temp Gauge problem

[InYaMooey]

Can anyone help in this one ?

BMW E39 M62 Engine (not that I think that is really relevant)

Problem is the Dash Temperature gauge goes to maximum, this is after the needle gets to about 10.30 position then straight to the red
Now the strange thing, it only happens when the ambient temperature is down to single figures !! Works perfectly all summer!!

New sender, plug contacts cleaned and cable looks ok before disappearing into the loom ..

Any ideas on this one? The ambient temperature part is really frustrating me.

Plus I’m st with electrickery

Thanks in advance

Edited by InYaMooey on Friday 25th September 18:26

[GreenV8S]

InYaMooey said:

this is after the needle gets to about 10.30 position then straight to the red

What temperature does 'about 10:30 position' represent?

[stevieturbo]

Then test the wiring under these conditions by varying a resistance to see if it behaves as it should during a fault condition.

If it does it isnt the sender.

Then unplug the wiring at the dash and test the dash itself in a similar manner.

[njw1]

I had exactly the same problem after doing some work on the cooling system on my 540, the temp gauge would go straight into the red shortly after starting the engine and the only way I could get it to reset was to disconnect the battery and after doing this three or four times it strangely seemed to cure it. Have you had the codes read?

Sorry I can't be more helpful!

[Huff]

Until you can trace which sender/bit of wire is borked, do buy a cheap obd2 reader (EML327 clone) that'll let you monitor the coolant temp in real time on your phone. NB these engines run at about 105degC nominal +/-3degrees (V8s) by design/fitted thermostat. AND - at that - the temp gauge reading is heavily damped/modulated - it should be 'straight-up' /'12o'clcok', from about 85degC, until about 112- 113 - when the aux electric fan will kick inon high (ie noisy) even when the AC is off.

Sound like you've something wrong with a sender, or wiring.

The temp gauge first quarter mark is about 60dgC , BTW, whether e39 6 or 8 cyl. It's the point at which the interior vent /demist fan system turns-on properly ( -it is programmed not to before that, to ensure a speedy warm-up!)

Edited by Huff on Friday 25th September 22:29

[E-bmw]

You say "works perfectly all summer" is this the same as worked fine until I put the heater on?

[InYaMooey]

Thanks all,

GreenV8S said:

What temperature does 'about 10:30 position' represent?

Hi Peter, normally anywhere between here (see pic) and 12:00 is when it decides it likes the red zone sharpish (as in straight over) Huff has put some numbers up.

stevieturbo said:

Said lots about electrickery .

Will have to get a mate on this as im good with Oily,Rusty or wet bits not so much the devils invisible magic

I FORGOT TO SAY The other problem is it's intermittent but only happens at lower ambient temps sometimes <12 but more so <10c

It does occasionally revert to 12:00 when driving but not very often (no noticeable road bumps or thumping the dash )

And ignition on/off doesn't solve the problem.

njw1 said:

I had exactly the same problem.... disconnect the battery ...

Did this a few times last year and didn't help now have years worth of 25mpg average that i'd rather not scrub

Huff said:

Sound like you've something wrong with a sender, or wiring.

The temp gauge first quarter mark is about 60dgC , BTW, whether e39 6 or 8 cyl. It's the point at which the interior vent /demist fan system turns-on properly ( -it is programmed not to before that, to ensure a speedy warm-up!)

New sender, i have a fear that it is wiring

Doesn't coincide with fans as far as i can tell.

E-bmw said:

You say "works perfectly all summer" is this the same as worked fine until I put the heater on?

Very much not ..

Edited Yes i have new pads ready to go on

Edited by InYaMooey on Saturday 26th September 18:46

[Penelope Stopit]

Have a hunch it's a fault within the instrument cluster, but.....

What does the gauge show when the engine is running and the temperature sensor is disconnected, if the gauge swings over to hot there is a possibility that a cable is going open circuit between sender and gauge.

Very much doubt that the above is the fault as a cable will go open or short circuit with vibration more so than it would with a variation in temperature (summer/winter)

Instrument cluster fault is the favourite, there could be a dry joint at the cluster circuit board that is breaking down when a certain ammount of current is flowing through it during colder outside temperatures

There could be an internal fault in the cluster that can't be seen (electronic or connection)

[Megaflow]

Gauges like that typically work on resistance. They have a power feed, a normal earth, and the sensor earth. The garage works out the value by comparing the resistance of the two earths. If it is going to max, it will be losing one of the earths to make a comparison with, most probably the standard earth.

[buggalugs]

Get some diagnostics on it, what does the ECU think the temperature is when the fault is present. If it sees a high nonsense value look again at the wiring to the sensor etc. Corrosion in connectors, broken wires, bad grounds. If the ECU sees the right temp then look at the cluster and wiring.

[GreenV8S]

Do you know whether the gauge is driven from the ECU or has a separate sender on that car?

[buggalugs]

Would be extremely surprised if the gauge on an E39 had a separate sender, my E30 did but they came out in the early 80’s. E39’s are mid 90’s. Coil on plug, traction control, CAN, digital dash etc

[stevieturbo]

buggalugs said:

Would be extremely surprised if the gauge on an E39 had a separate sender, my E30 did but they came out in the early 80’s. E39’s are mid 90’s. Coil on plug, traction control, CAN, digital dash etc

Mid 90's is still pretty old.

I would not expect the dash to be fed from the ecu in that manner....it probably is a simple sensor.

[InYaMooey]

Penelope Stopit said:

Have a hunch it's a fault within the instrument cluster, but.....

What does the gauge show when the engine is running and the temperature sensor is disconnected, if the gauge swings over to hot there is a possibility that a cable is going open circuit between sender and gauge.

Very much doubt that the above is the fault as a cable will go open or short circuit with vibration more so than it would with a variation in temperature (summer/winter)

Instrument cluster fault is the favourite, there could be a dry joint at the cluster circuit board that is breaking down when a certain ammount of current is flowing through it during colder outside temperatures

There could be an internal fault in the cluster that can't be seen (electronic or connection)

Ignition OFF needle hard left
Ignition ON Cold engine needle primes to first white bar this is the same with the sensor plugged in or not !
Warm engine sensor unplugged needle returns to first white bar.

I also think it might be a cluster fault so will pull it and re seat the plugs/wiggle the wires (I can manage that much electrickery )


Megaflow/Buggalugs
4 pins in the sensor/plug all cleaned and even when the plug is not fully seated giving a reading back to the gauge.

I have INPA (used for sorting the rear SLS and a few other bits) but every time i threaten the car with it typically the gauge isn't/stops playing up. As i said it is intermittent and not appearing so often at the current temp 13deg

Although it did hard over this morning but by the time id got home, you guessed it, vertical needle

Pete/bugga
Yep Stevie, 2 sensors, this one, near the Stat at the top front of engine, the ECU/Fan etc sensor is on the lower rad outlet.

Thanks again all.

[InYaMooey]

InYaMooey said:

Ignition OFF needle hard left
Ignition ON Cold engine needle primes to first white bar this is the same with the sensor plugged in or not !
Warm engine sensor unplugged needle returns to first white bar.

i will try and catch it when needle is far right and if any difference or back to first white bar or hard left (this will be slightly difficult to judge as its only a few mm)

Edited to add: Pulled sensor plug when needle hard right/in red and stayed in that position

Plugged back in and sill red but returned central a few minutes later (not surprised by this with it being intermittent and borderline ambient 12c not lower)

Edited by InYaMooey on Sunday 27th September 17:51

[Penelope Stopit]

InYaMooey said:

InYaMooey said:

Ignition OFF needle hard left
Ignition ON Cold engine needle primes to first white bar this is the same with the sensor plugged in or not !
Warm engine sensor unplugged needle returns to first white bar.

i will try and catch it when needle is far right and if any difference or back to first white bar or hard left (this will be slightly difficult to judge as its only a few mm)

Edited to add: Pulled sensor plug when needle hard right/in red and stayed in that position

Plugged back in and sill red but returned central a few minutes later (not surprised by this with it being intermittent and borderline ambient 12c not lower)

Edited by InYaMooey on Sunday 27th September 17:51

Nicely proven then

If a cable or connection between sensor and gauge was going open circuit the gauge would swing down to cold

The only other thing that could cause the problem other than something being wrong with the cluster would be the sensor cable shorting to something

What are the chances of a cable shorting to something at a given temperature?

Extremely small chance of the above happening

Cluster fault is now the favourite after your diagnostics unless the temperature sensor sends a signal to the DME ECU which then sends a signal to the cluster, doubtful but can't get hold of a diagram to be sure

Could unplug the ECU to prove if it does send a signal to the gauge

[GreenV8S]

InYaMooey said:

2 sensors, this one, near the Stat at the top front of engine, the ECU/Fan etc sensor is on the lower rad outlet.

Coolant temperature sensors are almost always thermisters. If the gauge is driven from a separate sensor, I think it is likely to be a simple analogue gauge. In that case there would be a resistance inside the gauge which is in series with the thermister. The two resistors act as a voltage divider and the voltage is what drives the needle.

Thermisters are typically high resistance when cold. In order to produce a 'hot' signal either there is an electrical fault within the gauge itself, or the sensor is showing a very low resistance. It is unusual ime for these sensors to short out (open circuit failure is more likely) but a wiring short to ground would produce exactly the symptoms you're seeing.

It's also possible that the supply voltage is way too high or there is a fault increasing the resistance of the internal resistor but that seems less likely. I wouldn't be surprised if you found a wire with damaged insulation.

[buggalugs]

According to Figure 14 here, if you jump pins 1 and 2 together the guage should go max hot, which gives you somewhere to start from in testing.

https://www.pelicanparts.com/techarticles/BMW-X5/1...

[InYaMooey]

Penelope Stopit said:

Cluster fault is now the favourite after your diagnostics unless the temperature sensor sends a signal to the DME ECU which then sends a signal to the cluster, doubtful but can't get hold of a diagram to be sure

Could unplug the ECU to prove if it does send a signal to the gauge

Thank you for going above and beyond

My favorite was also cluster (but see follow up below)

Really don't/didn't (see below) want to start messing with ECU

GreenV8S said:

I wouldn't be surprised if you found a wire with damaged insulation.

Thanks Peter (We've been in discussion before on high power V8 TVRs under a different name )

I think this might be nail head situation.

again see below

buggalugs said:

According to Figure 14 here, if you jump pins 1 and 2 together the guage should go max hot, which gives you somewhere to start from in testing.

https://www.pelicanparts.com/techarticles/BMW-X5/1...

This was all very strange !

Jumped the pins and nothing, so had another go after moving the plug (180deg ish) and the temp gauge went mad, full over then back to zero and back up again to max. This was dependent on ignition on full or engine running.

After this, plugging the sensor back in, fired up but then immediately stalled and wouldn't start without a bit of coaxing (throttle) so i suspect it has nudged the ECU in some way (Pete, 4 wires in the plug not 2?? Penelope, ECU )

So i have a feeling its cable/plug connection based hopefully before they disappear into the main loom, that's only a few inches..


Might not know much more for a few weeks so will revisit when the easterlies arrive


On the upside i drove out for 10 Mins just to get up to normal temps ...



ste phone pics but a small thank you

[buggalugs]

Looks like it was worth the drive out!

Everything from Fig 12 onward is about testing the sensor so worth going through those steps. But if bending the cable (which I guess is what you meant?) makes the car go nuts then that's got to be suspect.