Forum » General Tuning Discussion » Charge Temp. Estimate

Charge Temp. Estimate

General Tuning Discussion

Discuss all things tuning in this section. News, products, problems and results. 


Page 1
Author
603 Views

With the G4+ and I series ECU's having a charge temp estimation table [like the Autronic has had for many years] , has anyone worked out a way to accurately set this table up.

Hi Grant,

This is a question that's cropped up a couple of times and unfortunately the answer right now that I have for you is no. I'm planning to do some testing shortly on our 86 with a temp sensor mounted down near the inlet valve so I can monitor and log the relationship between charge temp (air temp at the valve) IAT, ECT, and airflow. The 86 is a good candidate for this as I can run it on DI only so the cooling effect of the PI won't effect the sensor reading.

The aim of the testing will be to develop a step by step process to approach tuning the charge temp table so people can achieve consistent results - After all that's what the charge temp table is meant to help with.

Understand of course that if you set the entire table to a value of 0%, the Link G4+ will treat the fuel model in a conventional way where ECT is ignored in the calculation. This may be easier to tune if you're struggling to get results that are stable with the charge temp table referencing ECT too. A technique I'm currently using on our Z with the G4+ is to give a very small amount of weighting to the ECT sensor at idle/cruise. So perhaps numbers of 20-30% in the cruise and idle area, transitioning towards 5-15% at WOT/high rpm. This gives some consideration to the effect of ECT but won't cause the large swings in AFR that can occur with changing ECT when giving more weighting towards the ECT sensor.

Hi Andre,

thanks for you fast reply. From what I can see, when you put zero in the table the software when using modeled fuel, still uses Boyles ideal gas law to the iat/fuel calc. so you should have the iat table turned off? When you think about it [me anyway] at 0 or very low rpm the iat would almost be ect. It gets very difficult after that. This table, if you could get it right would give a very stable a/f under all conditions. Problem is it would be different on various engine models and every modified engine as well. I have been trying to think of a way of changing a parameter like you did on the fuel temp coefficient to do something similar with charge temp. Have looked at lots of Autronic charge temp tables and every one seems to have a different idea. Autronic also have there own air temp sensor which must be very fast acting, do you know anything about it?

Cheers,

Grant.

You're right, a value of zero will still take the IAT value into consideration in the ideal gas law calc that occurs in the background of the modelled fuel equation. Obviously we do want this happening though to ensure the fuel delivery accounts for changing air temperature. The issue is that the IAT sensor alone doesn't always relate directly to the air temp as it passes the intake valves.

The difficulty with the charge temp table is deciding how the actual charge temperature changes with relation to IAT, ECT and airflow. As you mention, you can't hope to apply the same charge temp estimate map to every engine combination and expect success. A plastic or composite intake for example will absorb less heat from the engine heads (ECT related) than an alloy intake manifold. This makes it more complex to develop an effective tuning technique for the charge temp estimate table.

What we normally see is that if we ignore ECT entirely and only use IAT for the input to the ideal gas law, the AFR will tend to be affected quite heavily by heat soak at low airflow. This makes it hard to achieve stable AFR with IAT alone. A correctly optimised charge temp table can correct this situation.

The Autronic IAT sensor is unique as far as I've seen to Autronic ECUs. I've got no data on how fast it responds though sorry.

Thanks Andre.

I just saw this after posting in the other thread. Thanks for asking the question :)

I have started to experiment with my table to see if I can get some more stable AFRS. I will show you next time we catch up Grant.

I will be following this very closely.

Thanks for your help Andre - keen to see your results.