Tutorial: How to scale your MAF [pics]
Evolved Member
Joined: Jan 2006
Posts: 953
Likes: 2
From: Oxfordshire
Please excuse my question as a neophyte, but the MAF scaling value is an 8 bit number, so 255 is largest number that can be entered. Aren't we going to max out at 255 by ~125-150Hz and just end up filling most of the table with the maximum value for the data type (assuming ~20% correction)?
Thanks,
Mike
Thanks,
Mike
MB
Well, we changed the scale to a 16 bit value, but isn't the right column still 8 bit (uint8)? So, we can represent Hz to 65535, but the data values in the right column only to 255. Or am I understanding this incorrectly?
Any thought on this Eric?
My question is this:
I just tried this yesterday on a friend's IX after we installed his K&N intake (I know ...). It seems obvious that one correction % will not work across the entire map. My initial thought was to take a few samples in closed loop, such as idle and a few cruising RPMs, and attempt to create a sample curve to base the rest of the values on. Has anyone tried this method? If so, was it successful?
I just tried this yesterday on a friend's IX after we installed his K&N intake (I know ...). It seems obvious that one correction % will not work across the entire map. My initial thought was to take a few samples in closed loop, such as idle and a few cruising RPMs, and attempt to create a sample curve to base the rest of the values on. Has anyone tried this method? If so, was it successful?
Well, there are many ways to approach rescaling your MAF. There is no right and wrong way, unless you want to make sure that you have true airflow readings.
Your suggestion sounds fine to me. Basically you just have to get your curves to be somewhat the same shape, with no big dips or spikes too far away from a stock log.
A quick and dirty way can be as follows:
1. See if there are any LTFT changes for both Lo and Mid. If so, use the appropriate Hz row to adjust.
2. Then, do a WOT pull from low RPMs to redline (assuming this is safe on your tune and new intake) and note the Hz values where any dips or spikes from stock were. Then, you can use your MAF scaling table to smooth these out.
Another option to #2 (if you don't really care about true airflow numbers so much) is to just use your open loop fuel maps to retune your AFR curve. Just use the maf scaling to get your trims back in check, if need be.
Eric
Your suggestion sounds fine to me. Basically you just have to get your curves to be somewhat the same shape, with no big dips or spikes too far away from a stock log.
A quick and dirty way can be as follows:
1. See if there are any LTFT changes for both Lo and Mid. If so, use the appropriate Hz row to adjust.
2. Then, do a WOT pull from low RPMs to redline (assuming this is safe on your tune and new intake) and note the Hz values where any dips or spikes from stock were. Then, you can use your MAF scaling table to smooth these out.
Another option to #2 (if you don't really care about true airflow numbers so much) is to just use your open loop fuel maps to retune your AFR curve. Just use the maf scaling to get your trims back in check, if need be.
Eric
Last edited by l2r99gst; Jul 16, 2007 at 12:30 PM.
Evolved Member
Joined: Jan 2006
Posts: 953
Likes: 2
From: Oxfordshire
The g/s or L/Hz is an 8 bit number as mentioned, the values in ECUflash go to 321( in original format) as the data is not in it's raw format. Re reading the post I can see what you are saying, but I got a little confused about the mention of Hz there, sorry for that.
Worst case correction is a 10% increase, t5his would be if you were seeing 1966 Hz ( 12288 g/s) on the MAF. A more reasonable 1311 Hz ( 8192 g/s) sees 12% or so increase, so defiantly a restriction on the upward changes. Any one have any data to show if these correction % are a problem in re scaling the MAF?
MB
Worst case correction is a 10% increase, t5his would be if you were seeing 1966 Hz ( 12288 g/s) on the MAF. A more reasonable 1311 Hz ( 8192 g/s) sees 12% or so increase, so defiantly a restriction on the upward changes. Any one have any data to show if these correction % are a problem in re scaling the MAF?
MB
The g/s or L/Hz is an 8 bit number as mentioned, the values in ECUflash go to 321( in original format) as the data is not in it's raw format. Re reading the post I can see what you are saying, but I got a little confused about the mention of Hz there, sorry for that.
Worst case correction is a 10% increase, t5his would be if you were seeing 1966 Hz ( 12288 g/s) on the MAF. A more reasonable 1311 Hz ( 8192 g/s) sees 12% or so increase, so defiantly a restriction on the upward changes. Any one have any data to show if these correction % are a problem in re scaling the MAF?
MB
Worst case correction is a 10% increase, t5his would be if you were seeing 1966 Hz ( 12288 g/s) on the MAF. A more reasonable 1311 Hz ( 8192 g/s) sees 12% or so increase, so defiantly a restriction on the upward changes. Any one have any data to show if these correction % are a problem in re scaling the MAF?
MB
I'm not quite sure where you are getting your g/s values, but they are way off.
A mass airflow can be found from the hz values, but it involves baro and temperature, too. So, there isn't a direct correlation between Hz and mass airflow (g/s). If you want a rough estimation, you can look at my spreadsheet, but 1500Hz is roughly 240 g/s.
Eric
Evolved Member
Joined: Jan 2006
Posts: 953
Likes: 2
From: Oxfordshire
Hope the pic show what I'm on about, relating the Hz( left column) to the data ( right column) Not saying Hz=g/s but they are read next to each other in the MAF scale map.
Just check out your Evo MAF spread sheet, did you up date the correction factor of /10.25 on the earlier on for a 6.25? could be where my figures are getting screwed up.
MB
Just check out your Evo MAF spread sheet, did you up date the correction factor of /10.25 on the earlier on for a 6.25? could be where my figures are getting screwed up.
MB
Last edited by burgers22; Jul 16, 2007 at 01:37 PM.
Anyway, the second column insn't g/s anyway. The second column is like a L/Hz value so that the MAF can be calibrated to give a volumetric airflow per Hz.
Eric
Evolved Member
Joined: Jan 2006
Posts: 953
Likes: 2
From: Oxfordshire
MB
a quick question here!
I have installed my Apexi intake and all is well!
the car drive fine and the intake def holds more boost towards redlin.
now the prob comes where i was logging and before at 1.55 bar i would hit Loads of 270 since then i did the 2 byte mod and now i hit just over 300 on the same boost!!
do i need to do this mod or should i just leave it alone which i prefer to do!
in ECU flash i don't have load zones to adjust over 300.
should i just add more load zones in ECU flash.
now another thing can the ECU work with a greater than 300 load???
i know when i had my scooby the MAF used to give probs at over 300g/s
thanks
I have installed my Apexi intake and all is well!
the car drive fine and the intake def holds more boost towards redlin.
now the prob comes where i was logging and before at 1.55 bar i would hit Loads of 270 since then i did the 2 byte mod and now i hit just over 300 on the same boost!!
do i need to do this mod or should i just leave it alone which i prefer to do!
in ECU flash i don't have load zones to adjust over 300.
should i just add more load zones in ECU flash.
now another thing can the ECU work with a greater than 300 load???
i know when i had my scooby the MAF used to give probs at over 300g/s
thanks
I also have a quick question. i have the M1 reflash from jestr with an Apex'i cone filter. I noticed that my maf is not rescaled on my flash. Was this left out by accident or does he tune around it?
