My A2 doesn't like the cold

sco

A2OC Donor
I've noticed that when the outside temperature drops below about 6degC my A2 runs noticeably rough - doesn't pull as well and feels a bit like one cylinder is 'misfiring'. It starts easily whatever the temperature is and the idle is fine, and the roughness seems unrelated to coolant temperature so even when the engine is up to temp if the outside temp is low then the roughness sets in.

Any thoughts on this - I'm going to spray some egr cleaner into the intake pipe and run a tank of fuel system cleaner through to see if that helps but I was also wondering about doing some logging to see if one of the engine sensors was generating an erroneous value that was perhaps causing the problem.

Cheers,

Simon.
 
I've noticed that when the outside temperature drops below about 6degC my A2 runs noticeably rough - doesn't pull as well and feels a bit like one cylinder is 'misfiring'. It starts easily whatever the temperature is and the idle is fine, and the roughness seems unrelated to coolant temperature so even when the engine is up to temp if the outside temp is low then the roughness sets in.

Any thoughts on this - I'm going to spray some egr cleaner into the intake pipe and run a tank of fuel system cleaner through to see if that helps but I was also wondering about doing some logging to see if one of the engine sensors was generating an erroneous value that was perhaps causing the problem.

Cheers,

Simon.

Have you changed the Temperature sender? if not it could be worth it?

Steve B
 
Haven't changed anything (yet) - this temperature sender is that air or coolant?

Simon.
 
Haven't changed anything (yet) - this temperature sender is that air or coolant?

Simon.


The coolant temperature sender.

It "tells" the ECU what the engine temperature is (and a second circuit that "tells" the dash temp gauge what the engine temperature is)

They frequently fail and so the ECU is told that the engine is hot when it isn't (or cold when it isn't) and so you get different running depending on which way it has failed.
It is often undiagnosed because the temperature gauge can still read correctly even though the other circuit has failed.

Not expensive, not a major task and WELL worth it to eliminate the cheaper more common faults first.

Steve B
 
Ok thanks Steve - I'll see if I can log that channel with my obd scanner to see if it's reading odd values.

Simon.
 
Having looked through some of the old coolant temperature sender change threads they all seem to be associated with weird gauge behaviour which I don't have so I'm thinking maybe one of the other sensors - fuel temperature or MAF?

Simon.
 
Having looked through some of the old coolant temperature sender change threads they all seem to be associated with weird gauge behaviour which I don't have so I'm thinking maybe one of the other sensors - fuel temperature or MAF?

Simon.
Hi Simon
i don’t know which threads you looked through but the temperature readings are on a separate circuit and so it makes no difference if the temperature gauge is erratic or not.
The circuit that feeds the ECU often fails with no sign of it on the temperature gauge.

When it fails it dorsnt necessarily throw any error codes. The MAF would throw an error and so if you have no error then it is unlikely to be the MAF and that is WAY more eclxpensive than the temperature sender.

I think you will will find lots of people who have had an improvement in running by replacing the temp sender. They are so cheap and reasonably easy to change it is surely worth it when you have a clearly temperature related issue.

Steve B
 
Last edited:
As I said it is one sensor but two circuits (4 pins on the green sensor)
one circuit to the ECU and the other to the Temp gauge from the same sensor.

Steve B
 
Last edited:
yes you should be able to see a reading which the ECU gets in Deg C and this should correlate with the gauge, 60Deg C is the start of the gauge and each mark on the gauge is a further 10Deg C, mid gauge being 90 Deg C
it is normal for the gauge to out to the ECU feed by +/- 3 or 4 deg C any more than change the sender unit (only about £7 from Euro and appears to work OK and last OK)
check the correlation when the engine is very cold, if the engine as not been running all day then temp should be roughly the outside air temp, likely plus about 5 deg C

cheers,
 
Paul,

You were spot on - when the car was cold the coolant temp was reading 8degC and the outside temp was 3degC. The reading seemed fine to me - it warmed up to 20degC fairly quickly and then gradually climbed higher as I drove home, once it got to about 65degC it then tracked what the dash gauge was reading so I'm pretty sure it's working fine. I also monitored the charge air temp and that varied from about 5degCwith throttle closed to 12degC under boost which also seemed about right. The fuel temp just read as zero which probably means I picked the wrong channel to monitor so will have another go at that tomorrow.

Simon.
 
Paul,

You were spot on - when the car was cold the coolant temp was reading 8degC and the outside temp was 3degC. The reading seemed fine to me - it warmed up to 20degC fairly quickly and then gradually climbed higher as I drove home, once it got to about 65degC it then tracked what the dash gauge was reading so I'm pretty sure it's working fine. I also monitored the charge air temp and that varied from about 5degCwith throttle closed to 12degC under boost which also seemed about right. The fuel temp just read as zero which probably means I picked the wrong channel to monitor so will have another go at that tomorrow.

Simon.

Agreed this rules out the coolant temp sensor
The slight variation between ambient temp and coolant temp is simple the mass of the engine takes longer to equalise with the outside air temp


Sent from my iPhone using Tapatalk
 
Agreed this rules out the coolant temp sensor
The slight variation between ambient temp and coolant temp is simple the mass of the engine takes longer to equalise with the outside air temp


Sent from my iPhone using Tapatalk

Glad you were able to check this, it is always worth eliminating that as the cause and if you can scan it as done here it saves you swapping it.

Nice one
Steve b
 
More forum searching has revealed that the wiring to the fuel temperature sensor can get broken so I'm going to scan for codes and check the wiring at the weekend.

Simon.
 
More forum searching has revealed that the wiring to the fuel temperature sensor can get broken so I'm going to scan for codes and check the wiring at the weekend.

Simon.

Sounds like a good plan, the ECU certainly looks at the fuel temp and adjust injection fuel qty based on many factors including the fuel temp
How much difference it would make if the fuel temp was a long way out is anyones guess, but if VAGCOM is not monitoring a fuel temp then the reason why needs to be investigated. Which measuring block are you checking for fuel temp ? I will be hooking un vagcom to my a TDI tomorrow so can give you comparative readings

Cheers,
Paul
 
Hi Paul,

I was using an app on my phone that connects to an OBDlink reader via bluetooth - it doesn't have measuring blocks just channel names, I do have VCDS lite but not the full VagCom version.

Simon.
 
Hi Paul,

I was using an app on my phone that connects to an OBDlink reader via bluetooth - it doesn't have measuring blocks just channel names, I do have VCDS lite but not the full VagCom version.

Simon.

Use vcd light and let me have the measuring block number
I have had some dodgy results from the odb2 BT senders vagcom always provides the correct readings


Sent from my iPhone using Tapatalk
 
I scanned the car with vcds and it showed no engine faults, I've also got some data from group 000 which I think contains fuel temperature although I don't which channel yet.

Simon.

Capture2.PNG
Capture.PNG
 
Last edited:
that does't look like my VCDS display for block 000

That's possibly because it's an old version of VCDS (12.12), I did download the latest version of VCDS-Lite to try that but it won't work with my USB cable.

Simon.
 
Back
Top