Intake Air Temperature Sensor - do we need one?

evolve automotive · 18th July 2011, 13:44

Hi Guys,

I am very hesitantly posting this as it could potentially cause some ill feelings but this is not my intention so please read carefully.

Over the last few days I have received emails about the intake air temperature sensor and our Alpha-N software.

The questions have arisen due to the release of another companies' MAFLESS tune where they do not use the IAT sensor.

Some of the emails I have received are suggesting that we are for no reason keeping the sensor and one even suggested that we do it so we can make additional sales from selling a kit!

So, I have no choice but to cover this topic and hopefully it will serve to educate those that are interested.

Intake temperature sensors have been on BMW's since the E30/E34 era. They are there for a reason and are an incredibly important sensor.
ECU's have intake air temperature compensation maps build in where the general strategy is to offset the ignition and fuel maps.

As the intake temperature rises the ignition timing is retarded and as the temperature lowers the ignition timing is advanced. At a certain intake temperature the ignition timing is neither advanced or retarded.

This is true for the MSS52 DME fitted to our E39 M5's also.

If you unplug the MAF's you will see that the default value for this sensor is 60 degC.

Below is a test we carried out on a car where we disconnected the IAT sensor.

Power vs Air Fuel Ratio

Power vs Torque

Power vs Actual intake temperature inside airbox (using dyno probe)

You can clearly see that disconnecting the inlet air temperature probe adds quite a bit of fuel and retards at the very least in this situation upto 5 deg ignition timing (which is alot!). A full 20hp peak comes down so we can easily deduce the IAT sensor compensation maps work and quite aggressively at that....... obviously for good reason!

Now, for us to create a map that does not take the IAT sensor into consideration we can follow many strategies. Not one of them is something we would ever go ahead with because we understand the importance of this vital sensor.

We could simply switch off the fault code for the IAT sensor when disconnected and map the fueling and ignition with the default value left at 60degC or change it to any value we want. However the mapping would only be optimal at the inlet temperature at which we mapped for at the time. As soon as the temperatures change both the fuel and ignition applied would be not be optimal. How far away from optimal would depend on how much difference there is between the temperature at which we mapped at and the actual inlet temperature. The bigger the difference the less optimal the applied fueling and ignition.

With this shown you can now ask the question of how would this effect the running of the engine. Well, suppose we have an actual outside temperature of 40 degC (Some parts of USA and definately the middle east) and the default setting of the sensor was set to say 25 degC. You would have too less fuel and way too much ignition timing. This would lead to detonation and lean running and therefore adaptive values being either very negative of positive.

Now, there is another aspect to this and this is how the actual temperatures change whilst during any normal driving cycle. If you care to log your inlet temperatures at say 25 degC ambient you will see how the temperature swings so massively especially in town driving. Even at a drag strip the inlet temps would rise as the car is stationary to something like 60+degC even with 15degC outside and by the time your at the line ready to go you have completely the wrong fuel and ignition for full throttle driving.

The situation worsens with a power tune because all power tunes will run more ignition timing. You can see where that can lead!

The reverse of this would be very low outside and therefore inlet temps (very cold countries such as Sweden etc) and a default which is much higher. The you have too much fuel and not enough ignition timing.

The fuel for part throttle would eventually work itself out because of the lambda sensors but for full throttle you are in the situation above. Ignition timing would always be offset though.

Basically there is no assumed preset value for the IAT sensor you can set to work optimally under all conditions.

The sensor is critical and must be used!

There are various ways to stop the fault code appearing but all lead to the same situation and as someone who has spent a huge amount of time developing the calibration for the S62 with at least 120 cars worldwide being run on AlphaN.... IAT MUST BE USED!!

If you have the above strategy used in the base mapping then even if you do fit a temperature sensor it's a waste of time because your base calibration is set for which ever temperature the mapping was done for.

If you think it's not important then you might aswell delete the coolant temperature sensors too!!!!!!!

Most people on this board are very intelligent and I am rather shocked to see the correspondence from members here even asking me about this and even more shocked at the ones who have accused our AlphaN setup using sensors which are not actually required and it's all part of a plan to earn a few dollars from a IAT relocation kit!

I hope the above brief explanation puts people's minds at ease and shows that we develop our engine calibrations with the correct approach from all angles and not just from the point of view to 'run' the engine and 'to make power'.

Again, I will mention that the above is no way intended to harm another company and their strategy but is a reply to those that doubt our methods because of how another company has decided to calibrate their version of an AlphaN tune.

All car manufacturers use IAT sensors in their cars and it's not just about emissions as you can clearly see from the above.

Bestia · 18th July 2011, 14:44

Thanks for the info Sal. Those of us who have personally talked to you understand your honesty and love for our cars. You have a lot of prominent board members vouching for you and EVOLVE. Keep up the good work.

rao · 18th July 2011, 14:54

What is being said here applies to any car with fuel injection, especially one that does NOT use a MAF sensor.

Not using intake air temperature as part of the calculations is a bad idea; it is really that simple.

Moving that IAT around to "trick" the computer is also not a good idea.

vantaam5 · 18th July 2011, 15:00

Thanks for taking the time to post this explanation up,mate!

Got tired of replying to emails members sent me about that matter lately.

Hope the question is now clearly answered with real world data to proof it.

evolve automotive · 18th July 2011, 15:07

Quote:

Originally Posted by rao

What is being said here applies to any car with fuel injection, especially one that does NOT use a MAF sensor.

Not using intake air temperature as part of the calculations is a bad idea; it is really that simple.

Moving that IAT around to "trick" the computer is also not a good idea.

Exactly!

Tricking the ECU is actually alot worse than many people think. Always in the inlet tract. Many companies put them in the bumper or fog light area.

evolve automotive · 18th July 2011, 15:07

Quote:

Originally Posted by vantaam5

Thanks for taking the time to post this explanation up,mate!

Got tired of replying to emails members sent me about that matter lately.

Hope the question is now clearly answered with real world data to proof it.

No worries. Please reply to any further enquiries by just linking them to this thread.

Mike S. · 18th July 2011, 15:28

Sal - thanks for posting this. I thought it was a very clear explanation as to why IATS are needed.

cab · 18th July 2011, 15:50

Sal - thanks very much for the detailed post and numbers. The deltas aren't huge of course in terms of power (well, OK, on these motors 5% is a huge gain w/o a forced O2 ingestion setup of some kind), between a default setting and the "correct" setting, but as I assumed, the only "safe" way to go is to enrichen the mixture which then leaves that power on the table (which defeats at least part of the purpose of these tunes).

evolve automotive · 18th July 2011, 16:35

Quote:

Originally Posted by cab

Sal - thanks very much for the detailed post and numbers. The deltas aren't huge of course in terms of power (well, OK, on these motors 5% is a huge gain w/o a forced O2 ingestion setup of some kind), between a default setting and the "correct" setting, but as I assumed, the only "safe" way to go is to enrichen the mixture which then leaves that power on the table (which defeats at least part of the purpose of these tunes).

No worries.

You might find that a REAL 20hp is actually quite big and more than enough for you to very much notice it.

I did experiments like this all of the time at the beginning of the programme and I wish I could get another 20hp out of mine with just tuning because that is enough to make it push you back in the seat that bit more!

HaydenM5 · 18th July 2011, 18:22

I am by no means an M5 ecu tuner (all my tuning experience is rotaries/Rx7s etc), but the idea of not having an IAT sensor is frightening. Your ECU needs this as a vital input. Engineers don't put sensors in "just for fun" so that everyone can delete them. Thanks for your informative post Evolve.