MAF Scaling - Open/Closed Loop Fueling info

[JimmyMac]

Originally Posted by Turdinator

I noticed a while ago that the car felt the nicest to drive when the outside temps were around 20c. If we could tweak the table for more performance when cold and a tad safer when warm I'd be very happy.

I notice this a lot too. A certain times of the day the car feels very strong, and at other times it seems lacking. I was thinking it was the dense air, but it seems to also happen at times when the air is a little hotter than 20c or 68F. It doesn't pull harder at 10c than 20c. So I was thinking it might have been running on a different table.

[Turdinator]

Originally Posted by JimmyMac

I notice this a lot too. A certain times of the day the car feels very strong, and at other times it seems lacking. I was thinking it was the dense air, but it seems to also happen at times when the air is a little hotter than 20c or 68F. It doesn't pull harder at 10c than 20c. So I was thinking it might have been running on a different table.

It was the way performance dropped off below 20c that surprised me. I wondered if it was to do with the direct injection since this is my first car with that. If it is the tune I would love to fix it.

[JimmyMac]

Originally Posted by Turdinator

It was the way performance dropped off below 20c that surprised me. I wondered if it was to do with the direct injection since this is my first car with that. If it is the tune I would love to fix it.

I'm glad you mentioned it because I do notice that is a sweet spot (temp) for me. I think it was low 70's to high 60's Fahrenheit that the car felt great. I've been trying to keep an eye on the outside temp when the car was running good and it aways seems to be around this range. Wonder if we can "intercool" our NA cars hehe. I think I'll try and pay even more attention to it now. Maybe log it if I can remember to.

[Kodename47]

Originally Posted by solidONE

Really? So it's the load that gets adjusted... that would explain the relatively low % of adjustment in the table, I guess. I've decreased the entire IAT compensation table by .4%. I'll see if it has the desired result.

I probably would leave the OEM values TBH, they will have been tested far more thoroughly than we can do.

What sort of temperature ranges are you seeing? What you could do is widen the x axis rather than change the values in the table.

[solidONE]

Originally Posted by Kodename47

I probably would leave the OEM values TBH, they will have been tested far more thoroughly than we can do.

What sort of temperature ranges are you seeing? What you could do is widen the x axis rather than change the values in the table.

I could try that. My recorded IATs have ranged from 60-105 under normal driving conditions, so I've been trying to get the MAF scale dial in for 80 degrees. With the .4% decrease in the temp compensation while driving with 90-100 degrees IAT's the fuel trim has gone more negative for the most part compared previous logs at about the same temperature.

[solidONE]

Originally Posted by Turdinator

It was the way performance dropped off below 20c that surprised me. I wondered if it was to do with the direct injection since this is my first car with that. If it is the tune I would love to fix it.

That is surprising...

[ML]

I just feel like I'm doing something wrong here.

[vgi]

Originally Posted by ML

I just feel like I'm doing something wrong here.

that doesn't look right. i suggest to post your log file.

[ML]

Originally Posted by vgi

that doesn't look right. i suggest to post your log file.

When doing the OL on your tool, does it have to be in a sequence? I had a lot of small log files and combined them into one, and then filtered out all the CL from the csv. I think this might be the issue, that or I'm doing my AFR error wrong.

[Wepeel]

Are there really only 4 LTFT ranges for this car? I know on my STI there were LTFT A, B, C, D, each with a progressively higher g/s range (I *think* D was >40 g/s), and then D would be what was applied during open loop WOT, so although the fuel trims weren't linear with mass flow rate it still attempted to account for that.

Disclaimer here - I'm only monitoring fuel trims from the Torque app, which is reported as "LTFT1", so not sure if that has anything to do with it. I feel like I see a lot more than 4 ranges of fuel trims. It almost seems that there's a crossover around 3k rpm... if I'm cruising at under 3k rpm at a given highway load, and then speed up to over 3k rpm, the fuel trim changes although the flow range is constant. It goes from postive something to negative something. Maybe port/DI transitions?

[vgi]

Originally Posted by ML

When doing the OL on your tool, does it have to be in a sequence? I had a lot of small log files and combined them into one, and then filtered out all the CL from the csv. I think this might be the issue, that or I'm doing my AFR error wrong.

Please explain this "I'm doing my AFR error wrong"? You're not calculating AFR error on your own, are you?

I suggest you do not mess with your log files at all. Just click on 'Load Log' when it promts you if you would 'like to reset column name or filters' click yes and makes sure all is set correctly: select your columns, set filters to filter out non-OL data, click ok. When you select Throttle Open Angle depending on what tuning solution you use you can use Trottle Position or Accelerator Pedal Angle. It's preset at 85% but I STRINGLY suggest you should take a look at the log curve and determine when transition happens from OpenLoop to CloseLoop and vice-versa for the parameter you decided to use.
Eg.


After you click OK the columns for different WOT runs populated. If you see too few data rows - that means at a brief point you had full throttle but you probably didn't do WOT run to redline. I would suggest to copy last column wot run data over the column with few data rows. If you still have WOT run columns empty and have more logs - just repeat the process except this time you can skip setting column names (unless you changed column names ) and filters. so, click on 'Load Log', click 'no' to the prompt to reset columns/filters. You should see more WOT columns populated. you can have up to 12 WOT runs loaded.
if you have pasted current MAF scaling already then just click 'go'

PS you can play with finetuning the filters or ask Kodename47 for best settings - he has spent a lot time on that.

PPS @everyone Please, if you have gone through Usage tab and if something is not clear or is not covered - send me revised Usage text once you've learned whatever was not clear or not covered. you're getting helped, so pls help others.

[ML]

@vgi thanks, I was using a separate spreadsheet to find my AFR error. Doing it your way looks a lot better.

edit: @steve99 , might want to add the above comment to the OP.

[vgi]

Originally Posted by ML

@vgi thanks, I was using a separate spreadsheet to find my AFR error. Doing it your way looks a lot better.

that's the idea - automate as much as possible so their is no need to f.. around with spreadsheet(s)

[Kodename47]

@ML if you're trying to do logging to make changes ensure you stick to 1 main principle and you can use 1 log for both OL and CL: Only go WOT when logging a complete pull. You only really need 2/3 for the tool to work properly. For CL changes you only need partial throttle and I tend to try and vary pedal movement to change the throttle plate angle but rarely go above 3.5k. I find using different gears and slowly going up the range works quite well.


I think the setting you really need to look at is the AFR Error +/-%. If you don't change that, your tip in throttle will show massive positive trims and artificially raise your curve. If your NA I would have it set exactly as @vgi posted above. What I would do is load in your runs with the filter set to default (200) and then look at the individual runs and see what the maximum genuine error is throughout the curve. If you're NA it should be no more than 5-10%. If that's the case, clear the run data, reload the log but change the filter to just above the max genuine error and the resultant curve will fall back into place.