MAF Smoothing Tool for most Gen IVs

[ttz06vette]

The lower range typically gets changed quite a bit. You can pick and choose which section of the new curve you want. The lower the number, typically the greater the error will be after fitting just because of resolution. Copy the new stuff from about 4600 on up. The low load area is also subject to turbulence which makes the MAF calibration want to be even farther from ideal.

You can also do it again and zero everything below idle and above 4600. It'll then only use that specific range to calculate a new curve. Then just copy the same range from the new table.

It's not meant as a magic fix or anything. The biggest benefit I've had is just filling in the mid section of the MAF where it's a little more difficult to get a steady stream of data.

I agree. When I just used the Lo values the new curve almost exactly matched my existing. As soon as I loaded the Hi values in, it distorted significantly. Excellent tool thanks!

[DSteck]

Realistically, the entire MAF range is probably better off broken up into two separate 3rd orders that intersect at a point (and most likely share the same slope at that point). This will do any section you want, based on any data set you punch in. As mentioned, zeros are ignored.

[Justinjor]

As an Excel junkie, I love the spreadsheet you made. It looks awesome and I love the simplicity of the formulas. Sometimes the most basic mathematics can provide powerful results.

Well done

[Wnts2Go10O]

gee, looks, ... familiar
hey DSteck, LARGE favor to ask... remember the dirty worksheet i posted? >.>

though from ur post in my thread: "The third order poly thing is nice to get you a rough starting point if you're way off base... but I'll never rely on it. If it results in my WOT AFR drifting away from commanded, you'd better believe I'm going to tweak the MAF table to get my desired result. Being a practicing engineer, I know all too well how "on paper" is rarely what actually happens."

somewhat goes against your OP here

[DSteck]

It doesn't go against it at all. It helps fill in the blanks... It's not meant to make a perfect MAF curve. I've repeatedly stated that.

I can almost guarantee the Excel sheet I posted doesn't use anything somebody's already posted. ANYBODY can plot some data, fit a trendline, and copy the coefficients out. My spreadsheet automates everything. Punch in numbers, get immediate results. I also did this completely on my own, have been doing it since well before your May 2011 post, and just decided to pretty up an existing spreadsheet I had. It also isn't back calculating on fuel delivery... It is smoothing an existing table. Why the hell am I even defending myself? I kind of read your post as saying I'm copying something you've done.

[Justinjor]

It doesn't go against it at all. It helps fill in the blanks... It's not meant to make a perfect MAF curve. I've repeatedly stated that.

I can almost guarantee the Excel sheet I posted doesn't use anything somebody's already posted. ANYBODY can plot some data, fit a trendline, and copy the coefficients out. My spreadsheet automates everything. Punch in numbers, get immediate results. I also did this completely on my own, have been doing it since well before your May 2011 post, and just decided to pretty up an existing spreadsheet I had. It also isn't back calculating on fuel delivery... It is smoothing an existing table. Why the hell am I even defending myself? I kind of read your post as saying I'm copying something you've done.

It looks good Dave, well done. I love it I'm going to use it as soon as I get home Thanks!

[Wnts2Go10O]

It doesn't go against it at all. It helps fill in the blanks... It's not meant to make a perfect MAF curve. I've repeatedly stated that.

I can almost guarantee the Excel sheet I posted doesn't use anything somebody's already posted. ANYBODY can plot some data, fit a trendline, and copy the coefficients out. My spreadsheet automates everything. Punch in numbers, get immediate results. I also did this completely on my own, have been doing it since well before your May 2011 post, and just decided to pretty up an existing spreadsheet I had. It also isn't back calculating on fuel delivery... It is smoothing an existing table. Why the hell am I even defending myself? I kind of read your post as saying I'm copying something you've done.

yea, the worksheet i made is "dirty" and even said that someone could take it and clean it up make it better, etc. in my post i even suggested that maybe you could go through and add a portion with the formula from mine.

i said it looked familiar, not a copy, just familiar.

sometimes coincidences are a bit eyebrow raising.

[jimmyblue]

hmmm



seeing that my car idles at 2000-2200 hz, and the new maf has a 69% increase in value, i think i might have an issue. lol

My 85mm MAFs read about 2500Hz in open air, no flow.
A bit of thermal conduction in the struts is my guess. Add
any intake reversion to that and you're looking at a hand-fit
exercise.

[DSteck]

yea, the worksheet i made is "dirty" and even said that someone could take it and clean it up make it better, etc. in my post i even suggested that maybe you could go through and add a portion with the formula from mine.

i said it looked familiar, not a copy, just familiar.

sometimes coincidences are a bit eyebrow raising.

Like I said, this is something I've had for a while. I just prettied it up and added some functionality.

[Sam88gta1]

This will work awesome for those killer KB 4.5" tubes.

[DSteck]

This will work awesome for those killer KB 4.5" tubes.

But the KB tube doesn't work!!!!!!!!!!!!!!!!!!111!!!one

[Sam88gta1]

But the KB tube doesn't work!!!!!!!!!!!!!!!!!!111!!!one

LOL!! He has NO data to back it up. You should see his mail order tuning.

He had a mod warn me to stop posting in that thread.

[IDRIVEAG8GT]

Well this is just nasty! Thanks Steck! I'm gonna go put it to work on a non-enhanced rescaled E67 and see what happens.

[DSteck]

nice may be you need to get bored more often.

I'm always bored.

[IDRIVEAG8GT]

I'm always bored.

Go fish?

[Alvin]

I'm really close to finishing up a spread sheet sorta similar to extrapolate unknown values in the MAF but more handy the VE table using only KNOWN good values. Basically what you'll do is come back after a long log or tuning session and find what you where able to hit driving the car then extrapolating the rest. I've taken it out to a 6th order poly just to try to fit it as best as I can

The key thing to remember though is this is extrapolating.. Its just a guess

[DSteck]

Well, to be technical, mine is mostly meant for interpolating (filling in a grey area between low load and WOT), although it can be used for extrapolating.

I was going to do something for the VE table, but I don't know how to do regression for a 3D surface.

[bluegoat06]

Wow, Cool tool Steck, i'll give it a try this afternoon

[redhardsupra]

@Alvin, don't get too carried away with high degree polynomials, at one moment you're overfitting. You gotta understand why the polynomials make a good approximation of the underlying physical model, and exactly what limits they come with.

[DSteck]

Clear your PM box, Marcin.