Dynamic Airflow - Revisiting Custom OS Question

[BigDaddyCool]

So I have wideband tuned the MAF and the VE pretty close, each seperately........... now throwing the maf/ve back together with the wideband still in, just in open loop etc to check they are working together i noticed that the dynamic airflow is dragging them both down for a better word........ now obviousily I need to probably just reign them both in a tad better BUT........ I never got an answer to my original question on another thread.

With the 'bluecat' VE tool.......... when you change the VE Table and spit it back out into HP Tuners it 'changes' the zone boundries and stuff on the dynamic airflow.......... ie. it uses that to reference what cells in the virtual ve tables, but it also then changes what cells it also references against in the dynamic airflow prediction tables..........

Now running custom os, people keep saying 'leave those tables alone' but I personally think we should be using bluecats tool, just to spit out the new dynamic airflow numbers....... why....... because it should essentially keep the calculations closer.

The dynamic airflow is clearly what decides what to lean on most, maf or ve while your driving in a mixed setup..... problem is, if you do mods etc.......... and you increase efficency at low rpm, it's still no difference if your up at 3-4000rpm on stock......... if your more efficient you should probably generally lean to whatever that section was meant to lean on..........

if you follow what i'm saying.......

I dont think there's been enough discussion really about the custom os, ve and dynamic airflow. I do see why people just run maf only, I mean at the end of the day it's only about getting the airflow/fueling as close to 'a target' all the time.......... and maf seems to be the easiest approach..... given the complexity of this stuff.

I'm just trying to wrap my head around it so i can run maf or mafless ........ without having to change the tune..... so for that to happen i need VE/MAF/Dynamic to all 'be right'.......

The more i dig, the more I'm considering just running maf despite wanting to go 'mafless'. Complicated.

I think mostly by changing the Dynamic Airflow from stock to what you've improved........ if you know your maf is stable in those rpm/map spots (for it to be beneficial), it will just force the car to lean more on MAF then VE.

[Chris@HPTuners]

when you are in SD there is no dynamic airflow prediction calculation. All the airmasses are set to the result of the MAP-based airmass (VE tables) calculation.

[BigDaddyCool]

when you are in SD there is no dynamic airflow prediction calculation. All the airmasses are set to the result of the MAP-based airmass (VE tables) calculation.

Yes hence why people use maf or mafless....... if you use one or the other you don't have the dynamic issues......... but i'm more referring to if you run both maf and mafless together like the stock setup is.

[HartRod]

Interesting discussion.

[Chris@HPTuners]

Yes hence why people use maf or mafless....... if you use one or the other you don't have the dynamic issues......... but i'm more referring to if you run both maf and mafless together like the stock setup is.

The VE airmass is used during transients and the MAF used during steady state operation. However, there is an important thing to note.

The dynamic airmass calculation is basically a prediction algorithm that tries to predict the airmass ahead of time to enable a more accurate fueling. During steady state it uses the MAF airmass as its base and predicts the future airmass using the base factor and the TPS & MAP factors. TPS and MAP change are the indicators of a tip-in or tip-out situation. If TPS increases and MAP decreases that's tip-in and the coefficients will tend to predict a higher airmass in future. The opposite for a tip-out.

During this time the ECM is constantly calculating a correction factor between the MAF and MAP-based (VE) airmass. During a larger transient situation the VE airmass is used but multiplied by the correction factor to remove any step offset. This airmass is then passes thru the normal dynamic airmass calculation and future airmass predicted normally. Once steady state returns, it swings back to the MAF. This is designed to model the manifold filling/emptying process that happens during throttle transients. You can see that for small transients it sticks with the MAF,TPS,MAP + prediction. For larger transients it uses VE*correctionfactor,TPS,MAP + prediction.

So you can see that the MAF is the main influence here (that's why the whole thing is disabled if the MAF fails) and by way of the calculated VE correction factor, the VE table is always offset up or down to wherever the MAF airmass is. In the editor you can see how the VE correction factor is initialized and at normal operating temps the factor is initialized to 1.0. Obviously the goal is to maintain the VE Correction factor close to 1.0.

So you can see here that GM tries to achieve the best of both worlds in their airmass calculation. MAF is the dominant factor, but when it is known to increase in error (small transients) the prediction (using TPS & MAP) compensates. When the error is known to be larger (large transients) the VE airmass multiplied by the correction factor (ideally which is 1.0) and the TPS & MAP prediction compensates.

If the MAF fails the prediction calculation is disabled completely and everything is set to the MAP-based (VE) airmass.

If the TPS or MAP fails the MAF is still trusted but some extra filtering occurs to dampen the response and some maximum limits are set based on the MAP-based airmass.

Also, at high RPM the calculation is also disabled and a filtered MAF airmass is used. This is because the ECM doesn't have enough processing power to do the complex calculation at higher RPM, but also conveniently the filling/emptying effect is relatively small at higher airflows.

Note: the calibrations that decide the steady-state and transient condition are not in the editor.

Chris...

[kangsta]

The VE airmass is used during transients and the MAF used during steady state operation. However, there is an important thing to note.

The dynamic airmass calculation is basically a prediction algorithm that tries to predict the airmass ahead of time to enable a more accurate fueling. During steady state it uses the MAF airmass as its base and predicts the future airmass using the base factor and the TPS & MAP factors. TPS and MAP change are the indicators of a tip-in or tip-out situation. If TPS increases and MAP decreases that's tip-in and the coefficients will tend to predict a higher airmass in future. The opposite for a tip-out.

During this time the ECM is constantly calculating a correction factor between the MAF and MAP-based (VE) airmass. During a larger transient situation the VE airmass is used but multiplied by the correction factor to remove any step offset. This airmass is then passes thru the normal dynamic airmass calculation and future airmass predicted normally. Once steady state returns, it swings back to the MAF. This is designed to model the manifold filling/emptying process that happens during throttle transients. You can see that for small transients it sticks with the MAF,TPS,MAP + prediction. For larger transients it uses VE*correctionfactor,TPS,MAP + prediction.

So you can see that the MAF is the main influence here (that's why the whole thing is disabled if the MAF fails) and by way of the calculated VE correction factor, the VE table is always offset up or down to wherever the MAF airmass is. In the editor you can see how the VE correction factor is initialized and at normal operating temps the factor is initialized to 1.0. Obviously the goal is to maintain the VE Correction factor close to 1.0.

So you can see here that GM tries to achieve the best of both worlds in their airmass calculation. MAF is the dominant factor, but when it is known to increase in error (small transients) the prediction (using TPS & MAP) compensates. When the error is known to be larger (large transients) the VE airmass multiplied by the correction factor (ideally which is 1.0) and the TPS & MAP prediction compensates.

If the MAF fails the prediction calculation is disabled completely and everything is set to the MAP-based (VE) airmass.

If the TPS or MAP fails the MAF is still trusted but some extra filtering occurs to dampen the response and some maximum limits are set based on the MAP-based airmass.

Also, at high RPM the calculation is also disabled and a filtered MAF airmass is used. This is because the ECM doesn't have enough processing power to do the complex calculation at higher RPM, but also conveniently the filling/emptying effect is relatively small at higher airflows.

Note: the calibrations that decide the steady-state and transient condition are not in the editor.

Chris...

Thanks for that detailed description. Could we get the calibrations added in so we can see where the boundaries are?

[openwheel25]

This is very interesting. It is nice to find information on how the ecu works. My question is when you set the dynamic airflow to a low number to use maf only, does the dynamic airmass calculations completely quit working? Also when in maf only does the small transient corrections still work and is there a way to calibrate them? I know this is a old thread but it has great information. Thanks

[patooyee]

I'm sorry to pick an old post apart line by line but I am a new tuner and am enjoying learning more and more about the inner workings of an ECM, specifically the e67 since it is the one on the engine I am working on now. (Have owned an E40 LS2 in the past as well.) I'm hoping that dealing with this brain-dump of a post line-by-line I can learn a bit more than just what it says.

During this time the ECM is constantly calculating a correction factor between the MAF and MAP-based (VE) airmass. During a larger transient situation the VE airmass is used but multiplied by the correction factor to remove any step offset.

I understand this to mean that VE is weighted more heavily in the equation than MAF during large transients but MAF is still used? Why is MAF de-prioritized? And can you explain what you mean by "step offset?"

You can see that for small transients it sticks with the MAF,TPS,MAP + prediction. For larger transients it uses VE*correctionfactor,TPS,MAP + prediction.

Is this referring to the same correction factor found at Engine > Airflow > VE Correction Factor > Initial? Why would one want to change this factor? (I assume one might want to since it was put in the software.)

So you can see that the MAF is the main influence here (that's why the whole thing is disabled if the MAF fails) and by way of the calculated VE correction factor, the VE table is always offset up or down to wherever the MAF airmass is. In the editor you can see how the VE correction factor is initialized and at normal operating temps the factor is initialized to 1.0. Obviously the goal is to maintain the VE Correction factor close to 1.0.

I noticed the Initial VE Correction Factor table stops at 105kpa. How is boost handled in that case?

Correction factor at 1.0 = MAF tables and VE tables are in sync / good tune & low transients? If that is true, correction factor less than 1 =? And more than 1 = ?

Also, at high RPM the calculation is also disabled and a filtered MAF airmass is used. This is because the ECM doesn't have enough processing power to do the complex calculation at higher RPM

This statement is confusing to me because I have heard from some Ecotec guys that are boosting the e67 NA applications that they are leaving their MAF active all the way through their rpm range. So unless the ECM deactivates it despite them thinking it is active isn't the ECM operating fast enough? I think they basically tune via MAF and not VE because, until recently, the lack of a true VE table has confounded them. I think they are only just now catching onto the separate VVE tools that have been around for a little while and even more recently the one built into VCM 2.25.

Which brings me to my next question. I know that you can extend the axis in the VVE tool in 2.25 to accommodate not only higher rpm ranges than stock but also 2-3 bar MAP sensors. Once you do it kind of screws up the coefficient boundaries which then need to be adjusted. Once you export the VVE table to equations in the ECM though what happens with all the other tables in the ECM that still have the 105kpa limit? (Such as initial correction factor above.)

Lastly, I have seen you and Bill refer to a custom PID to scan a 2+ bar MAP sensor but have yet to find a link describing that pid. Can you please direct me to where you are referring?

Thank you for all you have done in the past, are doing, and will continue doing. I've never regretted my purchase of HPTuners and look forward to learning more about it daily.

[patooyee]

Did I say something wrong?

[BigDaddyCool]

Coming back to this topic......... I would like to know and or see if there are these hidden options for steady state show up in the editor Chris mentioned.

I have zeroed out all my dynamic and ve based tables completely............ in the hopes it would run maf...... but still dynamic logs on the scanner the same as if the data was all in the tables still....... which I find odd.

So essentially from what i have read since my first post, the REAL benefit of going mafless despite it being harder to tune........... is that it completely removes this BS from the equation.

It to me is frustrating that there is no simple way to choose MAF Only, MAFLESS Only or Mixed Mode and define in mixed mode if chosen how it works together. I would think HP Tuners would have this as a priority. We as tuners all know with experience what to change and disable to archive our goals with mafless and what not......... but what we cant do is see the hidden things missing possibly from the editor.

Food for thought.

[GXPLS4]

Annual bump for this perhaps with the new software advancements we could get this??

[kingtal0n]

I have a SD question. If I am using SD, and install new injectors (double original size), and ECU injector size is half of what it really is (scaled down 50% in the software), and then I go into my VE table and half everything there to compensate,

will that 'halve' my cylinder airmass table as well? (more scaled for higher pressure?)
And basically everything else (need to fix idle airflow tables, etc...)

Isn't there a table that pulls fuel from the VE table that might be used instead, so the VE table can remain the same, and just using coefficients of 0.50 for example be adjusted externally? I think I saw a table that lets you decide where to put "1.0 stoich" that could be changed, or something like that. The reason I don't just change the numbers on my engine to test them is because I do not have real time tuning.

[roastin300]

This guy lol. If you arent exceeding the airflow limitation then there is no reason to halve the injector AND the airflow tables. If you do then yes you have to adjust your spark timing because it is based on g/s and would now be way off. If your not exceeding max airflow (depends on what PCM and year you have) then you can do the simple Stoich and IFR trick to scale the injectors. Do a search and you will find those discussions. Its much easier. If you need more airflow for your spark tables, or your MAF, then the method your trying to acheive is needed. Or you install a bigger aftermarket MAF or install the MAF card in a larger tube, and do the Stoich/IFR trick to get the injector flow rate scaled.

[kingtal0n]

This guy lol. If you arent exceeding the airflow limitation then there is no reason to halve the injector AND the airflow tables. If you do then yes you have to adjust your spark timing because it is based on g/s and would now be way off. If your not exceeding max airflow (depends on what PCM and year you have) then you can do the simple Stoich and IFR trick to scale the injectors. Do a search and you will find those discussions. Its much easier. If you need more airflow for your spark tables, or your MAF, then the method your trying to acheive is needed. Or you install a bigger aftermarket MAF or install the MAF card in a larger tube, and do the Stoich/IFR trick to get the injector flow rate scaled.

thanks,
I am using 411 PCM with 125OS, tahoe 2002 flex fuel

I believe it came with 50~lb/hr injectors.
I want to run 80lb/hr. I will try to make it work first without scaling. But I wanted to prepare in case I need to scale, there are alot of airflow numbers changing at the same time.

[rbiron2]

I would like to bump this thread back up as I have noticed a lot of parameters available under dynamic airflow now (p59 - 2bar OS). I would really like to understand how the blending areas work as I have dialed in my MAF and my VE. When I turn them both back on the truck runs really great unless I command and hold 3rd gear (heavy load) and do a WOT pull. It starts running lean every time at 3100 rpm. I have the High RPM Disable set to 4000, it I set it lower at 2000 then the issue goes away, however, I believe the issue to be an adjustment needed to the way the ECU predicts the transition. Any insight would be greatly appreciated!

Log - Blended Lean Condition.jpg