Inferred MAP

[Higgs Boson]

My MAP "sensor" is showing 108 kpa at WOT (2016 Mustang) and that is the max map value in the tune....

I know it's not a real measurement but I also don't think that is accurate, either.

What are we doing about MAP (and does it even matter) values and also max air load values?

What are the implications of leaving the WOT air load table stock with modified cars? Too high or too low values?

Why not set the WOT Air Load table to 1.0 all the way across? What is the point of it going up and down as it moves across the RPM range?


etc, open discussion

[murfie]

You can disable the cylair anticipation and it will run completely off the Maf values. In a FI application this is the easy route. I guess if you do some extensive NA modifications this would be the easy way as well. You just lose the fall back for If and when the maf fails. Of course if your modifications throw the speed density values very far off and you don't correct them you don't want the computer falling back on them or cross checking the maf values with them.


If you want to retune the speed density tables. First you start by disabling them and logging what values come from the MAF.

[Higgs Boson]

So are you saying the Cyl Anticipation and Desired Airmass are VE/MAF blending criteria? If I disable Anticipation and Desired Airmass then I will run purely off MAF only?

[murfie]

I want to say desired air mass needs to remain enabled. The MAF still uses it to infer MAP.
Clyair anticipation is the switch that says to compare the throttle position calculated values and sensor calculated values for MAP. Both of these values go through The manifold filling model(how air enters and exits the stock TB and manifold). Because the throttle position calculated map and manifold filling model are based on the throttle body, modifications(FI, different size TB, different manifold)can cause these values to have a large amount of error in them. To correct this error in the Throttle position calculated MAP tables you need to disable the modification done by the compare and recalculate off the values purely from the MAF as it will have less error in the final calculation because it will measure the new volume of air going into the manifold. Both values will have error in them but the cylair filter and a high enough multiplier values should be enough to compensate at least for the Value being calculated from the MAF. For a more accurate manifold filling model you would need to copy the proper tables/values from the factory calibration of the car the manifold and throttle body came from. If it's after market I don't know what to do but rely on the values the MAF gives you. I guess you could measure and do some really complicated math but that's beyond the realm of most ordinary people.

[Higgs Boson]

I noticed a rousch tune has desired airmass initialization off but idle is on. I turned off Desired Airmass Init. and also CylAir Anticipation (left on MAF filter) and the car runs SOOOOOO much better now. No more jerky part throttle, etc. This may be what I needed to get more aggressive on cam timing as well.

I'll know more tomorrow after I drive it a little bit.

[murfie]

Both calculations use a manifold filling model. If the car runs with out the desired airmass it appears we don't have access to modify this. I was hoping we did. We have to rely on the maf by turning off desired air mass and cylair anticipation.
From the calculated maf values we can get new speed density values but there will always be a margin of error. The cylair filter multipliers will hopefully but enough to always take care of this error.


Reading more about manifold filling models they are designed to be dynamic and compensate for a very large number of engines. So what I'm thinking of as error is just the nature of intake manifold filling models. Calculating new speed density values from the MAF will give optimal tables for when the MAF fails. And if you can get the new values in controlled conditions you can turn the cross checking back on for extra environmental variable safety.

[AKDMB]

What part modifications require you to modify the inferred MAP? How do you tune it?

[murfie]

What part modifications require you to modify the inferred MAP? How do you tune it?

It looks like Even slight VCT modification to a completely stock engine can causes the speed density to wrong. So any modification to improve airflow can cause it to be wrong. You can disable it and leave it as it is with higher filter multipliers and not need to modify the speed density and rely completely on the MAF sensor. To recreate speed density would take a lot of time, math, and logging MP% and MAP value from the MAF sensor. The MAP value from the MAF is still a predicted value, But it comes from measurement and not a Throttle body flow model. See Attached picture. figure 2 shows the compare between MAF value and Inferred MAP. figure 2a is a more detailed path of inferred MAP. It shows the parameters that go into calculating inferred MAP.Amodel.PNG

[Higgs Boson]

If your car isn't stock, I would disable it. You literally can't do anything, from VCT changes to airflow modifications without having some negative side effect in driveability at this point. The MAF can easily absorb these changes (especially when you correct for it in the maf curve).

I am confident that HPT will eventually give us a way to properly tune the VE on these cars like we have for GMVE. IMO, a car runs a LOT better with speed density or a hybrid, but if all we have right now is MAF, then that's what I will use. Having a touch less transient response is better to me than a surging/bucking part throttle power delivery.

[murfie]

Coming back to this the IPC under airflow-general shows up to values of almost a two bar MAP sensor at 1.1 load. Some points a little higher some a little lower. This to me means with a high enough cylair multipliers(2 is highest) the MAF calculated values should be enough to handle up to ~15 psi of boost. That would be two bar and make sense with the multiplier max being 2. See my speed density calculation correction method in the mapped points isolation thread.

[murfie]

http://www.google.com/patents/US6363316

I think this applies and has very useful info.
If anyone is interested in the cylair equations and how the multipliers work.