A few problems with Cobb's suggested maf cal procedure

[djuosnteisn]

So i did a re-cal this weekend, and finally figured out some of the oddities from the first time i cal'd my 3.25" maf. The first time i did it, i noticed that the wot afr's started going from lean to rich, and i also noticed that i was having to scale a significant % more than the cobb worksheet suggested.

Here are the reasons, and things to keep in mind when scaling for larger maf's.

1) The maf sensor characteristic curve is exponential by nature. You CANNOT multiply a non-linear curve by a single scalar without distorting it's natural curve. You can verify this yourself by simply multiplying the whole maf table by 2 or 3 and check out the new shape. It becomes more and more like a "right angle" (90* angle).

If you think about it, the shape should never change, period. Cause the amount of air flowing through the sensor will never change (the sensor always outputs anything from 0 - 5 volts), the only change is that the sensor is reading a "smaller percentage" of the overall air flow.

So if you use a single multiplier across the maf table, you'll end up with a "steeper" wot region, which will result in wot afr's going from lean to rich as you g/s increase. This is also a likely fix for those that experience lean afr's after a shift.

IMO the better way to scale the maf curve is to simply shift it to the left, and interpolate the curve further into the opened up region. But it's good enough to simply do the scaling as cobb suggests, and then smooth the hell out of the curve to re-create the stock maf curve (adding alot to the middle region to reduce steepness).

*edit* I just proved to myself that shifting the curve to the left is no different than multiplying the whole curve by a single scalar. I think my issue with having to add more "meat" to the middle section likely stems from the design of my particular intake, and probably doesn't apply to everyone. But it is still a good method for fixing a "lean after shift" situation IMO.

2) Cobb's calculator suggests a percentage increase based on the % change of the diameters of the maf housings. The correct percentage increase should be based on the % change in maf AREA. So say your going from stock 66.5mm to 88.7mm (3.25"), the calculator would suggest an increase of like 33%, when in actuality you need like 77% increase [pi*(66.5/2)^2] / [pi*(88.7/2)^2].


That is all.

[superskaterxes]

i can haz help with calibrations?

[djuosnteisn]

You can

*edit*

I just proved to myself that shifting the curve to the left and interpolating the new region is no different than simply multiplying by a single value.

I think my experience with having to add a bit more to the middle section stems from the intake design, and probably doesn't apply to everyone.

But i will say that increasing the middle part of the maf curve will still likely take care of any lean issues after shifting.

[Lex]

This is a throwback to nonlinear scaling using the SB. I remember that was a popular way of taking lean spots out of the curve.

[fjames]

55% maxes out at 500 if you care. The curve looks the same to me when scaling for normal numbers, but I don't have educated eyes. I've tried to smooth visual curves on other tables and it's a one thing leads to another nightmare - you just end up with a straight line, and I'd rather have the curve cause it's prettier.

[djuosnteisn]

it'll only max out if you achieve that voltage, which you won't on a bigger maf.