- Try Unplugging your IAT sensor momentarily to see if your fuel trims come down. (or the AFR comes down Richer as it was normally) If they do use the custom table noobtuner explained with detail above…
If unplugging the IAT sensor doesnt do much… Check also these: - make sure your Fuelling → Corrections → charge temp map → is all 0%.
- If you have a small fuel surge tank make sure your fuel temperatures are not skyrocketing.
- Check for manifold or exhaust leaks (put a paper cardboard and seal lightly the exhuast end pipe at idle)
- Also check if your automatic idle settings so your ignition advance doesnt pull too much timing from the base ignition.
What you’re describing is the same problem we are all having when we have an IAT sensor that is vulnerable to heat soak.
I suggest you make your VE adjustments for 20 to 25 degree’s (or whatever you think is ideal for your setup) and then tune the custom table as the sensor heat soaks.
I have mine working really well after just two test drives so far.
Tomorrow I will narrow the MAP window to increase resolution between 20kpa and 30kpa and I’ll post my settings for illustration purposes.
Thanks for the list
- IAT I’ll give a try next time I drive the car. (I’ve already constructed a correction table sort of like Noobtuner that I can enable soon)
- Charge temp map is set to 0%
- Fuel temperatures are monitored with my flexfuel sensor, there is a small enrichment when the fuel heats up
- I’ll check for leaks in exhaust system (has to be removed soon so good way to check everything)
- ignition advance is heavily limited (only does ±2 degrees currently)
Weird thing is that the IAT doesn’t really heat up a lot, 35°C after 30 minutes idling isn’t too bad. Unless this is enough to offset the calibration. I’ll play a bit with it and see what it does
Thanks!
No you and I think many others are not fully understanding what is happening. These problems are related to the critical charge temp table.
Understand this important point: Once air temp matches the CLT the fuel model will always be at its most correct regardless of whats on the charge temp table. This is because the air temp entering the cylinder becomes an empirical measurement not an estimation any more, agree?
Air gets denser when its colder, requiring more fuel to maintain lambda, right?
Charge temp is defined as the temperature of the air/fuel mixture at the exact point it passes into the combustion chamber. Its continuously used to calculate the air density in the fueling model. Get this value wrong and fueling is also wrong.
So lets work backwards if you have charge temp set to 100% and air is colder than CLT then it will be lean but ONLY when air temp is below CLT temp because the ECU is calculating charge temp that is higher and less dense temp the whole time. The ecu will be delivering the same fuel regardless of real air temp. Fueling only becomes correct when its heatsoaked where IAT and CLT are matching because this is what it was calculating all along. This is the opposite problem your having right?
So lets look at the opposite situation. You have the charge temp on 0%. Now once again at full heatsoak condition when the IAT and CLT match, lambda will remain correct, life is good. But now if the air temp goes lower the ECU will start aggressively adding more fuel because it calculates the air entering the cylinder to be colder and more dense. But as this cold air travels down the runners and through the cylinder head it will not stay at the same cold temperature that was measured, it will have warmed up considerably. So the air is less dense than the ECU has calculated and the result will be running rich. So the result is too rich with colder air and as always correct when heatsoaked. NOW! if you have skewed the VE so that it appears correct with cold air then it will appear to lean out once its heatsoaked. This matches your problem right?
SO! the correct thing to do is use the charge temp table to estimate a split between the 2 sensors to make an estimate of the actual temp of the air as it enters the cylinder. Lets put some numbers in the above example. CLT 80, air 20. So lets say at a particular rpm and load the air passing through the intake and cylinder head warms it up to a real temp of 50 degrees so this is the true “charge temp” Now it becomes obvious that if we tell the ecu the charge temp is still 20 degrees (0% on the table) the ECU will supply too much fuel because it thinks the air is more dense (rich). Inversely if we tell the ecu the charge temp is 80 degrees (100% on the table) then the ecu will not supply enough fuel because it calculates the air to be less dense than it actually is (lean). In both cases as the air temp reaches heatsoak condition lambda will come back to be correct again.
So in this simplified example setting the charge temp table to 50% will cure the problem. 50% split between 20 and 80 is 50 degrees which matches the actual charge temp of 50 degrees so lambda remains correct. As the air temp rises so too will the charge temp calculation as it maintains the 50% split between CLT and IAT, emulating the real charge temp rise. This will continue all the way up to full heatsoak condition when the IAT and CLT matches each other but same as before fueling remains correct.
Of course this example is only at one given theoretical point of load and rpm. Different conditions will require different split to estimate the real charge temp, hence the 3d table.
Last but not least placement of the IAT is relevant, if its tucked in a corner out of direct airflow or getting influenced by the engine temperature too much then it becomes more difficult. Some compensation can be made by raising the values in the charge temp table but ultimately moving it may be the best solution.
in theory yes, in practice no. i just got the same problem just this weekend. night tuning 8c IAT, i set my VE fuel at 90C coolant ok ? Everything good lambda 1.000. Removing the IAT plug and going to failsafe value of 25c it become lean, normal behaviour right ?
Next day midday, REAL IAT temperature at 32-35C, it required 5-10% more fuel on my VE table. So IAT may have been truly higher, fuel didnt want leaning out. Because it was the EMU classic i started doing my thing trying removing fuel when IAT was low, and adding fuel over 25C to fight back the fuel model… I had the same problem in megasquirt but tunerstudio allows you to zero out the Intake Air gas law table… I think that was a decade ago i discovered it because i didnt have problem with older ECUs that didnt do automatic fuel adding removing with IATs as part of the “Gas law”
So in this example when IAT at 8c it was ok but IAT at 35c it was lean, what % split did you have on your charge temp table at this running/idling condition? Is your IAT actually reading IAT or is it being influenced by the temperature of the engine and intake manifold?
Values on the charge temp table will directly affect how much the lambda changes when the IAT is at different temperatures. For it to work correctly all other IAT corrections need to be 100 because they are only 2D tables and will not dynamically adjust the correction for different running conditions.
If you want to “ZERO” out the IAT in the gas-law calculation its easy! simply put 100(%) on the charge temp table.
Sounds like you still dont fully understand how this table works and interacts with the tune. Maybe read my previous post again and try and fully understand the whole explanation step by step.
Thank you for taking the time to contribute to this discussion. I think I failed to understand that the Charge Temp Table isn’t really optional, but rather, its built into the fuel strategy as much as the VE table itself. Setting it to zero is not really ideal, (unless maybe you have ITB trumpets out the hood of your snow drift car.)
I will spend some time getting well acquainted with what you have written and start re-tuning with a new appreciation of the Charge Temp Table.
Thanks.
Mal.
You can Read at the help menu what Charge temp does.
"Charge defines the contents of the cylinder after closing of all valves, the charge consists of the fresh mixture and the residual gases. The simplest way is to use IAT to estimate air density. This approach doesn’t consider the effect of air heating via intake manifold and engine elements. The lower air flow, the air heating is more significant. To estimate this phenomena the Charge temperature estimation table was introduced. This table defines how much IAT and CLT influence the charge temperature. 0% means the charge temperature is equal to IAT sensor value, 100% means that charge temp. is equal to CLT sensor value. "
It assumes that Intake Air Temp is even higher that the heat soaked Intake Maniifold because the ports of the engine are deeper in the engine core than the IAT sensor is. So it assumes that the charge AIR will be even hotter than what the IAT sensor displays. Super Logical and correct… IN THEORY…
In practice even in scorching hot IATs without any charge temp corrections, the model already falsley assumes way lower Air Density because of the high IAT’s in a heat soaked traffic laden commute. A Charge Temp table, of lets say, 100% would assume IAT temperature will be equal to Coolant, so your calculations would assume e.x. 90Celsius of Air Temps and NOT the IAT display of 50-60 or 70Celsius.
As it stands, the charge temp table is basically useless. Whenever I’m stuck in traffic and the IATs go up, the mixture leans out, so I have to use a custom enrichment correction to compensate. Does anyone actually use this table for anything?
Yes, I use it minimally
This is tough one because I see the point of it but maybe it doesn’t work the way as intended or it’s counter intuitive. I wonder how the OEs deal with this besides having an infinite number of comp tables.
I think of the charge temp as a “quick” calculation adjustment until the real IAT catches up. Maybe that’s the misunderstanding here.
perfect gas law quirk aside, lets break down the timing of what’s going on
- cruising with throttle partially opened
-
real IAT, calculations good, afr tuned as desired
-
IAT = 40c | CLT = 90c | BLEND 0% = 40c
- as soon as stopped and throttle closed idling
-
??? IAT, calculations ???, afr tuned as desired
-
IAT = 40c | CLT = 90c | BLEND 0% = 40c
- stopped for 30 seconds or whatever “heat soak” takes
-
real IAT, calculations good but afr tuned for pre-heat soak
-
IAT = 60c | CLT = 90c | BLEND 0% = 60c
It’s worth considering that 2 is where things are going wrong. Things are tuned too lean there because the intake temp becomes hotter than you realize quicker than than the IAT shows. As the IAT actually starts to show warmer, it leans it out even further. Consider this scenario.
- cruising with throttle partially opened
-
real IAT, calculations good, afr tuned as desired
-
IAT = 40c | CLT = 90c | BLEND 0% = 40c
- as soon as stopped and throttle closed idling
-
??? IAT, calculations 50%, afr tuned as desired
-
IAT = 40c | CLT = 90c | BLEND 50% = 65c
- stopped for 30 seconds or whatever “heat soak” takes
-
real IAT, calculations better, afr tuned for pre-heat soak
-
IAT = 60c | CLT = 90c | BLEND 50% = 75c
@ngt
I’m not saying ever make that table 100% but it’s pretty easy to see how some correction could help. In the end it really depends on the timing of the scenario here (how long until things start to drift from the desired values). I don’t disagree with the decision to make the IAT comp table also correct AFR target value in V3. The IAT comp table isn’t there to “correct” the fuel mix to the normal values, it’s there to enrich or enlean the mix and target. If you need to make corrections, that is what these other tables are for.
- The wording is pretty specific in the IAT correction table help menu incase you don’t already know…
used to implement engine cooling strategies
This is just from a random tune I pulled up
This car runs and idles fantastic no matter how long it sits in traffic. Very minimal drifting of afr at idle even on hot days. And even if it does drift around a little bit, who really cares. At idle it starts to run like crap way before any possible damage can occur. As it stays smooth and drivability is still OE or better, I forget about it mostly.
The fun part is I just tuned another car which the iat’s don’t really spread with sitting in traffic. But in the night it got 8celsius to 12max. Well the whole car had richer afrs. Next day, midday. Outside temps 22-23 with sun hitting the hood iat’s went 30-34. Full lean… 16-17 afr.
I had another car with a super sensitive IAT sensor it would go 60celsius from 40 during a shift change within 0.5 seconds and then back to 40c within 1 second… after reopening throttle… Wild 
For me IATs at this point is only good as data points for intercooler intake efficiency.
For me IATs at this point is only good as data points for intercooler intake efficiency.
Maybe for you, but for the ECU it’s doing all it’s calculations for the entire fuel map with it, whether you like it or not. 
Yeah thanks for reminding me thats why im protesting over here and on facebook for 1-2 years 
.
That charge temp thing i was looking desperately 3-4 years ago why when i sit on idle seeing AFR change, i would literally see the Injection ms, go down. And i had found out that charge table which i made all 0% it was improved… The old default maps had this charge table to 70-80% on idle, newer ones had this everything on 0%
But also after a longer idling heat soak made it go Leaner again less agressively, same thing again but only slower, at first i thought it might be the fuel getting hot, so i did a facebook post 1-2 years ago because i found out about the gas law model..
I was suspicious about it when i had the same problem with megasquirt it has a 2D table which calls it MAT Density Table.
Now my problem is not the charge temp table which is useless for me (have NOT seen a car needing a Coolant temp IAT equivelant YET)
Its not the Gas Law Model.
Its that the IAT Table now also modifies Lambda Targets. And its not a 3D table so why would anyone want a e.x. +10% enrichment when you have a heat soak on idle ? ANd if you needed component protection like OEM cars have you would use EGT…
Genuinely curious what OE cars have EGTs
if they dont have an egt sensor, they usually have a calculation model from how much heating energy their wideband sensor requires to work properly…
Basically all porsche turbo models have egt sensors