Sudden high temperature without coolant loss, 1.4tdi

Robin_Cox

Member
I assumed it might be a water pump failure causing a lack of coolant flow, but reading threads here the symptoms don't seem to match, so I thought it makes sense to post.

Went to fill up with fuel, didn't notice anything out of the ordinary.

As I restarted I noticed the temperature gauge shot up to well beyond 90 degrees before dipping back to just below 90 where it was before we arrived at the petrol station. It never does this.

Headed back on to the local dual carriageway, and as we got up to speed I noticed that the temp gauge was now advancing above 90 - again something it never does. On Colour DIS I brought up the twin readings - engine was 96 and increasing slowly, dashboard display was listed as 106. I immediately turned the cabin heater onto High, full power, directed to vents and opened the windows. Temperature on DIS stabilised on both readings, then as we came off our slip road and down to back-roads speed the climatronic managed to get the temperatures under control and below 90 degrees again in both cases by the time we were home 5 minutes later.

Got the bonnet off - no sign of coolant leakage anywhere although it was at the lower edge of the expansion vessel range - but did note that the coolant lid was 3 turns off being locked home. I've not had it off in ages so I don't know how. It has quite a lot of friction but I wonder if the apparent mild overheat was due to the lack of pressurisation since the lid wasn't tightened down.

I do have a combined Gates belts and water pump set ready for when the engine build takes place - it had the belts done just before I bought it 30000 miles ago, but I have no evidence that the water pump was done at the same time.

Any thoughts appreciated.
 
This reads to me like the engine is producing more heat than usual - or at least, heating up the coolant more. My first thought is a pinhole through the cylinder head gasket into the water jacketing. Strange about the coolant cap though - has it been into a garage at all recently, who could have dagnosed the same and done that to keep the pressure out of the cooling system? This would stop coolant being pushed out or finding cracks to leak through, but wouldn't stop it boiling away due to the reduced pressure.
 
Temp on both should read the same - it's the same sensor in the same physical place, even if it two separate devices inside.
Perhaps the thermostat stuck closed, although I can't see why it would even close until the engine gets cold.
Perhaps something stopped the water pump impeller turning - is that even possible?
The cap being off/loose on the expansion tank won't affect anything until the temp exceeds 100 deg C - all it does is change the temp at which the water boils.
There is a lot of latent heat in an engine when it stops, as the fan stops at the same time it is conceivable that the engine temp could rise beyond what it was with the coolant circulating.
Does your electric fan work?
 
What's the strength of your anti-freeze? If too low, could your engine have frozen recently, resulting in the impeller blades shearing off? I'm assuming that they're plastic! If metal, seizure would probably cause the cambelt to fail under the same circumstances. A thermostat is more likely to fail open rather than shut, in my experience.

RAB
 
Gut feeling and easiest to check is the water pump, remove thermostat for quickest/easiest access to impeller
 
If the water pump is functioning then you'll be able to see coolant being returned to the expansion tank.
 
Lets hope it is simple as the cap has been left loose accidentally. Refill the coolant and look for external leaks. Carefully check the seal on the cap. Is there any coolant around the outside of the expansion tank? The tank can leak from the seam, the cap or the level sensor. As AndyP has said run the engine and even stone cold there should be a steady flow through the small return hose that enters the left of the tank in the upper half ( viewed from the front ). With the engine running and the cap on check for leaks especially around the expansion tank and coolant temperature sensor area.
If the relief valve has failed in the expansion tank then that would vent out the rectangular hole in the rear side of the tank. New cap and or tank not too expensive.

No return flow then good chance the water pump is broken.

What is the oil like? Is it slightly milky, not fearing the head but perhaps the oil cooler is leaking coolant into the oil. Again not too expensive to replace.

Have a sniff test done to verify the head gasket.
 
Edit : I thought I had posted the bulk of this last night, but it was sitting in the submission box ready to "reply".

Thank you all for your replies - it is really appreciated. I am still in a minor flap as it is an unexpected issue and is forcing me to change transport plans for Easter - I was due to do the Dundee-Plymouth drive this Saturday and back in a week (in this car on cruise, in 6th).. so now I am going to be taking the £1 project car Cosmo instead and driving myself in 5th for 9 and a bit hours ... first world problem.

Coolant is correct concentration using the G13 or G13+ concentrate diluted with de-ionised water - I did the thermostat and temp sensor a couple of years ago and keep the levels topped up with freshly-diluted concentrate.

The car has not been into a garage since the MOT in mid-October and 312mm brake installation a week later, and mileage since the MOT has been relatively low as I mostly use my project car (the £1 A2) for commuting and everyday use all through the winter, especially as this car is planned for the engine build.

I don't know what the water pump is, but everything indicates it and the thermostat have been working fine generally recently - it was just in the last 20 minutes driving home that something odd happened and I've not yet got my finger on it.

I don't know about the electric fan, yet - I was able to get the temperatures down using the heater core method as I moved onto backroads, but that is not viable for a 540 mile motorway journey!

Second edit :

I was thinking about this before I got up this morning - thank you all for the additional thoughts and suggestions overnight. I am going to look for return flow in the expansion tank after topping the level back to the middle of the normal range - it is only fractionally below at the moment. Will also look at the oil ; the coolant itself looks very clean.

Third edit :

The one thing I forgot to mention is that there is an in-line DEFA electric coolant heater plumbed in next to the oil cooler. However, this has never given any issues in the few thousand miles since it was fitted, although I've not used it as a heater yet as my other car with Webasto was used in preference over winter. The last time the coolant lid was off would have been when the circuit was topped up after the fitting session, so whether it wasn't fully tightened down is open to question although I've had the car fully up to temperature several times over the winter without losing coolant.

Fourth edit :

regarding fan(s) ; on a 1.4tdi with air-conditioning, what is the normal status for the fans when the car is started (below temperature) and so on - I seem to remember reading somewhere that turning the a/c on switches on one of the fans but I have drawn a blank looking for threads describing what is running and under which circumstances.

Fifth edit : managed to find fan threads using Google, now trying to assimilate a lot of data without confusing myself.
 
Last edited:
Yes coolant flow back to the bottle is indeed the smoking gun relating to water pump and coolant flow.
Run at fast idle 1500 rpm once at 90 good flow should be present
 
Last edited:
So. Coolant fan runs when a/c is on, car is well below operating temperature so when a/c is turned to 'off', the fan stops. Obviously enough. That at least eliminates fan failure as a possibility.

Stone cold, there is no visible return to the coolant tank - but then again there isn't in my other Tdi either (have been watching that to see if there were any obvious differences).

Going to carry on now at a fast idle to see if coolant return commences once it has warmed up a bit. I did wonder this morning if the fact that I could use the cabin heater to cool the coolant down effectively in real time indicated that there was circulation of the coolant yesterday as I assume that if the circuit is static it wouldn't be possible to take heat out of the coolant as effectively.

Doesn't help that it is absolutely baltic outside - drizzle and around 4 centigrade aren't the most pleasant conditions.
 
Is it possible that it’s simply a faulty temperature sender? You have confirmed that you saw different readings from the sensor for ecu and dashboard, so what if the dashboard was simply wrong indicating an ‘overheat’ condition when the temperature was actually correct?
Perhaps you could check readings again as it warms up using VCDS to see if there’s still a discrepancy and if so then replace the sender unit.
 
Will have a look. Weather is beginning to clear up - it got progressively worse earlier so I had to retreat inside.

The odd thing was seeing the ECU coolant temperature heading north of 90 degrees - I think it got to 98 just as I was cranking the heater up and then dropped away again, let alone the dashboard value being 106. There seems to be a constant differential of around 6 degrees right now - I don't remember seeing this on one occasion before Christmas when I was warming the car up.
 
The odd thing was seeing the ECU coolant temperature heading north of 90 degrees - I think it got to 98 just as I was cranking the heater up and then dropped away again, let alone the dashboard value being 106. There seems to be a constant differential of around 6 degrees right now - I don't remember seeing this on one occasion before Christmas when I was warming the car up.

I would assume that they should both read identically, however if that is the case, why on earth did Audi design a system that effectively has two sensors, essentially used for the same purpose? The only outcome of which is confusion for the driver when either one fails ...
 
I would assume that they should both read identically, however if that is the case, why on earth did Audi design a system that effectively has two sensors, essentially used for the same purpose? The only outcome of which is confusion for the driver when either one fails ...
They are not used for the same purpose. One is used for calculating fuel injection, the other for fuel consumption. If they had used just one sensor, it would be difficult to know when it had failed. They should both read the same temperature.

RAB
 
Although the fan is running at the low setting for the climate, you still need to prove that it works for the coolant at the higher speeds. Not got the temperatures to hand that enable the 2 fan speeds for the coolant to hand , but have previously posted them. At a guess around 95 for the medium and 103 for the high speed. All the lowest speed does is ensure airflow over the aircon radiator.

You have checked the coolant temp sensor for any signs of leaks?
 
They are not used for the same purpose. One is used for calculating fuel injection, the other for fuel consumption. If they had used just one sensor, it would be difficult to know when it had failed. They should both read the same temperature.

RAB
By the same purpose I mean for providing a temperature signal, the fact the two signals are provided and that there isn't a warning that the two differ is confusing. If there was only one and sensor and it was also used for the dashboard display then the driver would have a clue if it goes wonky. The issue is that the two systems are not linked in such a way that an obvious warning is displayed if they are giving inconsistent readings.
 
By the same purpose I mean for providing a temperature signal, the fact the two signals are provided and that there isn't a warning that the two differ is confusing.
That's because Audi probably view failure as abnormal.
If there was only one and sensor and it was also used for the dashboard display then the driver would have a clue if it goes wonky.
Why would he? There's only one signal with nothing to compare it with.

Sensor quality is another matter. The conditions are not particularly arduous; with a decent quality sensor, they should almost never fail.

RAB
 
Although the fan is running at the low setting for the climate, you still need to prove that it works for the coolant at the higher speeds. Not got the temperatures to hand that enable the 2 fan speeds for the coolant to hand , but have previously posted them. At a guess around 95 for the medium and 103 for the high speed. All the lowest speed does is ensure airflow over the aircon radiator.

You have checked the coolant temp sensor for any signs of leaks?
Hi All,

sorry, no, the rain came back in earnest and then work requirements got in the way. I'm settled now on leaving this for a week as I have to go to Plymouth using the project car .. which will be interesting, first long trip in this one.

Thank you all again, I will come back to this when I'm back up north.
 
So.

It's not the water pump. With the return pipe to the expansion tank disconnected and held over a collection vessel, coolant flows when the engine is running, and the nature of the flow is that it is being pumped rather than something happening by chance.

It's not the fan. That runs.

Watching the coolant sensor feeds on colour DIS, the first few seconds after the car is started there is no data on the "dashboard" channel - it is just a "-", whereas there is a temperature listed on the "ecu" channel. Then after a couple of minutes, the dashboard channel flickers from a dash to a reading and back again before stabilising a few degrees above the ecu reading, although as the temperatures approach 60-70 degrees they tend to get closer together. Possibly a dodgy sensor? I replaced this in summer 2018, but it is possible that something has started to misbehave. Can anyone else with a colour DIS see if theirs also gives similarly weird readings (temperature (e) vs temperature (d) when stone cold and starting up? I've never watched mine like this before so I don't know what is normal behaviour. Even more bizarrely, even when the (d) channel was reading 96-97 degrees, the needle wasn't going beyond the normal position, although 2 nights back the needle was beyond 90 when this all started.

The more observant will notice that I wasn't planning to use this car yesterday. However, this afternoon, while preparing my £1 car for the journey to Plymouth - the last job on the list was checking and topping up the oil. It was about half-full (intact Tdi dipstick), so I chucked in another 250mL or so, checked the dipstick again and it was between half and full so I decided that was enough and put the dipstick back. On returning from the bin after chucking the waste paper away I noticed the foot diameter pool of oil dripping out from beneath the oil filter housing. It would appear that in removing and replacing the dipstick that I've cracked the dipstick tube or something else down there catastrophically, but I cannot deal with this now...

Edit - I have a suspicion it might be :


that's a job for 2 weeks time....
 
Last edited:
Oh no, sorry to hear this, you're not having much luck at the moment :( Fingers crossed the Plymouth run goes OK.

I've noticed on my CDIS that the ECU temperature lags the dash temperature on warmup, but I've not really looked beyond the ECU getting to 90°. I'm going to get plenty of miles in on that A2 this weekend, so I'll keep more of any eye on what it's doing.
 
Back
Top