Is the idea that you want to log the MAF in g/s instead of just volts?
Just a wild idea, but here's what I see in iVTS - maybe you could set up your MAF like in pic 1 below, and then make a "Fuel Table 2 inactive" that uses the MAF, like in pic 2. I don't know. Seems like that would spit out g/s somewhere.

 

I will have to look into that. Problem #1 I'm not using an oem sized maf housing. Number two is a cal table has so few data points (16?) I don't know how quality that reading can be.

This is the housing I purchased. Should fit in the factory maf location perfect.

This is to validate compressor performance and hopefully be used to regulate engine bypass air when I get that far. ( advanced turbo lag strategy)

I really should have gotten a i88!!!!

 

This is my rough new maf scale for that denso sensor. How would you alter the maf tube size to an existing calibration if possible? Im sure the preprogrammed MAF scale for the Subaru sensor is much more precise than a cal table with 16 slots.

 

Cool! Did you make that table in Excel or something?
I see that the mass points get closer together % wise as it goes higher in flow.

You know what I would do first from where you are there is plot all 46 or whatever of those points out in some graphic software like Catia or AutoCad if you have anything like that, or any Office software that can make curves, lines, points, all on a to-scale X - Y plot. Because I would want to see what that curve looks like. Mass flow vs voltage. Then you could play around graphically with how to position your 16 points that you have to work with, so as to minimize error.

I imagine that the vts software is going to go straight-line from one point to the next, whereas the real function would be a smooth continuous curve, which you have somehow picked those 46 points off of. So you'd be trying to position your 16 points so as to minimize the "chord height tolerance" which would be the worst (most) distance between your 15 lines and the actual curve.

That is where something like Catia would be great, because you could put a spline through the 46 points, or a "3D Curve" (in Freestyle or ISD AeroExpert) and you'd have something real close to the real curve. The 3D Curve could be "near points" instead of "through points" if there is trouble with the points being a little rough.
Anyway that would be your "reference" curve, from which to judge the quality of what you can get from your 16 points.

The basic idea of multiplying the OE flow #'s by the area ratio 1.51 seems like reasonable. I suppose the actual ratio will be a hair greater than that due to less relative skin drag effect in the larger tube. But I don't know how much.

 

Open calc

Looks like I'm only good with a 80mm tube for this turbo. Any more flow will be close to its limit although I'm sure it can handle a little more..

Anyways I love to tinker and this is why I got an aftermarket ecu

 

Did some looking for you.

You could use a evo 7-9 MAF, and wire it into a DI input then set the DI as MAF sensor, then under the cal set it as Evo7-9. It also has the CAL for STiv7-9 and the DBW STi MAF CAL. That would keep it easy. Or you can wire it in to the vipec and use a cal table if its not one of those mafs

 

Gotcha and that wouldn't change fuel calcs etc because the tables would still say kpa vs rpm.

I do see your point with the Evo maf being simpler .

At this point I will run analog with the Subaru 07 wrx maf I have, I crashed mine in 2009 and kept a bunch of engine pieces incl map sensor and maf sensor. Soon to add a MAP to the UICP too. I will deal with the % error I am stuck with in certain ranges. If more accuracy is desired Ill travel down the alternative paths.

I did a practice entry into the vipec cal table and I did 25g/s increments and I started at the 1.3 my table begins at.\

becoming more flexible than ever is what I see when I peek around. Note that you can have dual tables of all the important stuff with completely different axis for each one if you so desire.

 

No you would only be using it to log data. If you changed the axis to G/S then it would want to start using the MAF.

 

Well darn it now that I go through this there are some things about the calibration tables that look unnecessarily crude.
For one thing when I input a decimal g/s like 1.3 for my first value, it won't take the decimal, it just puts in a 1. So I started at 2 which was close to something on your chart. Hopefully the fricking thing can extrapolate itself.
Then probably worse, the increment requirement! I thought you'd be able to key in every g/s value individually so each one would be exactly whatever you want. But no you have to give it a start point and an increment. Am I missing something? Is there a way to put in the values you really want? What if you don't want equal increments?

Here's my rough run through this MAF thing:

 

You can have overlays also. So instead of just using a dual table you could use fuel map 2 as a overlay of map 1

 

Definitely crude. More resolution is needed for a maf entry IMHO atleast to be able to make your own scale increments individually. Like I said I wonder if we could edit the existing maf scale which likely has better resolution than a cal table.

I'm gonna ask at vi-pec

 

Aha, yes, now I get it. Good idea.
Tell them if they could hire Tom Collins to improve this type of stuff they'd be way ahead.
That isn't me by the way.

 

Well who is it

 

Bang!

 

Tom Collins was the maker of ecuplus. Not a fair comparison really because ecuplus was a lot less comprehensive, but it just always seemed to me like the way Tom did everything was so clear and sensible.

Is the Dave Kriedeman post from the Vipec forum?