MAF and MS? part 2 with new 2.900 code

[fobar22]

<forum admin team: this thread continued from: viewtopic.php?f=38&p=225906#p225906>

Hello, I just got my mustang (NA 383) fired up for the first time ever yesterday. I planned on using MAP/MAF blend but I found that at idle adjusting the MAF correction curve did nothing, but changing the VE table actually affected the PW.

I am running MSII REV 2.8900, and TS.

I downloaded the 2.9000 code that was earlier in this thread.

Is my best bet to reprogram the MS with the newest code? Or should I just reverse the MAP/MAF blend option as I read about a few pages back?

Like I said I only have the car just started, and I didn't really understand what the problem was so I didn't run it too long or tune it too much.

first run.msq

On this run, about 7/8th the way through, you can see where I messed with the MAF table but it haveing no effect on the PW of the injectors.

2010-03-18_15.15.34.msl

Thanks for any help!

Dave

[grippo]

Your best bet is to load the attached code with a very minor change from the above code relating to injector test mode. It is labeled the same but has this minor change so you might as well use it. The main reason for switching is to use the 2.900 TS ini file that is attached. This should have the maf/map stuff correct. Note also that the 2.900 code requires you to add your injector flow rate under General in the fuel menus on the very left. This is a better way to calculate injector pulsewidth when using maf.

[fobar22]

Thanks, I will give it a try in the morning.

Dave

[merruse]

OK here is 2.900 code to test. It calculates PW direct from fuel injector flow rate, so this must be known and entered, under the Fuel/ General menu. A new TS ini file is also attached for this purpose. This fueling is only used for pure MAF or a new maf + map mode (maf for fuel, map for load table index (for spark, afr tables)).

Where can I download uncompiled C files?

[mrselfdestruct1994]

Your best bet is to load the attached code with a very minor change from the above code relating to injector test mode. It is labeled the same but has this minor change so you might as well use it. The main reason for switching is to use the 2.900 TS ini file that is attached. This should have the maf/map stuff correct. Note also that the 2.900 code requires you to add your injector flow rate under General in the fuel menus on the very left. This is a better way to calculate injector pulsewidth when using maf.

Hey mate, is there any chance we might be able to get our hands on Bruce's prototype maffactor.inc generator program that you mentioned in the last thread? I've just purchased an HPX MAF Sensor (05+ Ford Slot Style) and have asked for the transfer function data in the required 1024 point table format however I'm not sure if they will be able to supply such accurate data.

Thanks for all your hard work by the way, appreciate it.

[cmonref]

Look here for MAFFACTOR.inc

download/file.php?id=10335

[grippo]

Hey mate, is there any chance we might be able to get our hands on Bruce's prototype maffactor.inc generator program that you mentioned in the last thread? I've just purchased an HPX MAF Sensor (05+ Ford Slot Style) and have asked for the transfer function data in the required 1024 point table format however I'm not sure if they will be able to supply such accurate data.

Bruce hasn't had time to work on it, but he is going to send it to me so I can clean it up. What most people have done is get the Ford standard MAF curve that cmonref mentioned above and use the correction table to fine tune it.

[mrselfdestruct1994]

No problem mate, I may end up entering those values manually anyway as it seems prudent to me to try and use the transfer function supplied with the sensor. That's not too terrible though, maybe I can con a younger family member into doing it for a couple of bucks. In regards to the standard Ford maffactor.inc file posted above, should I assume that the values listed begin at 0.0 volts and progress in a linear fashion to 5.0 volts? For instance:

0.0000000V = 00mg/sec*10
0.0048828V = 03mg/sec*10
0.0097656V = 06mg/sec*10
0.0146484V = 09mg/sec*10
0.0195312V = 12mg/sec*10

etc...

4.9951172V = 27568mg/sec*10
5.0000000V = 27673mg/sec*10

Just want to make sure I enter the values correctly if I'm going to bother typing in all those entries.

Thanks champ,
Sam

[cmonref]

Why do all that typing when it is available in a post by CRenner "fixitagaintony" in the predecessor thread to this one? He posts two ZIP files. In "mafsensors.zip" is a file "maf_summary_March_09.xls" that has the full data for volts vs airflow for the ford default.

download/file.php?id=9805

[mrselfdestruct1994]

Why do all that typing when it is available in a post by CRenner "fixitagaintony" in the predecessor thread to this one? He posts two ZIP files. In "mafsensors.zip" is a file "maf_summary_March_09.xls" that has the full data for volts vs airflow for the ford default.

download/file.php?id=9805

Hey mate,

I'm going to be using a PMAS HPX MAF Sensor which comes with the attached tool to calculate the MAF Transfer Function. I'm guessing this isn't 100% accurate and obvious the install will effect the calibration but I figured it was worth a shot as it should be a lot closer than using a standard Ford Transfer Function. What I wanted to do was figure out how to turn the 30 point table the HPX tool provides into a 1024 point table for use in the maffactor.inc file.

That .zip you linked me to was extremely helpful and I can't believe I missed it as I've read over that last thread what seems like dozens of times! Using those files as a reference and following some tutorials on the internet I came up with the attached mafTransfer.xlsx file which I think generates a 1024 point table pretty well. I stuck with the polynomial trend at voltages below what was supplied to me which means I have a lot of zero's at the bottom of my table, however I noticed that CRenner (aka "fixitagaintony") used a linear interpolation from 0 to the first actual value. Does this actually matter, and if so, which method is going to be more accurate?

I have assumed for these values that they should be calculated from 0.0 volts to 5.0 in linear increments which I believe is the same way CRenner has done it. All I need to do now is figure out how to make it spit out the maffactor.inc file, however I could probably just copy and paste those values with minimal fuss.

Hopefully what I've done is fairly accurate and mathematically correct but if anyone sees any problems or inaccuracies with what I've done please let me know! I'm definitely no mathematical whiz so I could easily have screwed something up.

Cheers,
Sam

[grippo]

You are correct in that the processor assumes that the 1024 data points you enter correspond to 1024 evenly spaced voltages from 0 to 5 volts. The reason for the large number of points is not accuracy, but to avoid wasting computer time doing interpolation. The ADC on the processor has a resolution of 1 in 1024, that is it returns a number from 0 to 1023 corresponding to 0 - 5 Volts. If the processor can look at a 1024 point table, all it has to do to convert to a MAF reading is to use the ADC value as an index into the table to get the value, no interpolation needed and very fast.

[mrselfdestruct1994]

You are correct in that the processor assumes that the 1024 data points you enter correspond to 1024 evenly spaced voltages from 0 to 5 volts. The reason for the large number of points is not accuracy, but to avoid wasting computer time doing interpolation. The ADC on the processor has a resolution of 1 in 1024, that is it returns a number from 0 to 1023 corresponding to 0 - 5 Volts. If the processor can look at a 1024 point table, all it has to do to convert to a MAF reading is to use the ADC value as an index into the table to get the value, no interpolation needed and very fast.

Excellent! Thanks for that in-depth explanation. Do you have any comments about my question above? The transfer function I was supplied starts at 0.35 volts at 69mg/sec*10 and if I use the polynomial function I got from excel to calculate values below that it predicts I won't get any reading at all until 0.28 volts, which means my table actually starts with a whole bunch of zero's. Should I use that data or should I do as fixitagaintony has done and use linear airflow values from 0 volts to 0.35 volts? I'm thinking this might be a stupid question as it's possible the airflow may never dip below 69mg/sec*10 anyway but I'm not sure.

[grippo]

The shape of a MAF curve from 0 volts to .35 V is very linear, so that is what I would use (0 at 0 V. 69 at .35V and linear in between).

[mrselfdestruct1994]

The shape of a MAF curve from 0 volts to .35 V is very linear, so that is what I would use (0 at 0 V. 69 at .35V and linear in between).

Thanks for that mate, just to let you guys know if you are attempting to replace the default maffactor.inc file with your own figures and you are trying to get your values formatted correctly, I found the following procedure seemed to work well. Please note I haven't tested my file yet so make sure you save a copy of the default maffactor.inc in case this doesn't work!

1. Copy and paste your 1024 values from Excel to Notepad
2. Copy and paste the files from Notepad to Microsoft Word
3. Open the Find and Replace tool
4. Enter ^p in the Find field and enter ^tDW^t in the Replace field (^p means new paragraph and ^t means tab)
5. Run the Find and Replace function (you will notice that it will not work for the first and last letters)
6. Make sure you enter the first DW and tab manually and remove the extra tab and DW from the end of the file
7. Open your original maffactor.inc file and copy and paste in your new values

I'm expecting my new sensor in the mail with a week or two so I will let you guys know how it goes.

[FixItAgainTony]

mrselfdestruct1994,
Glad the spreadsheet data was useful to you. As Al pointed out, in the low range, the exponential nature of the MAF curve can be approximated by a linear function. When using polynomial fitting, the resulting function can do odd things depending on the power used, amount of data, etc. I think Bruce pointed this out in the previous thread. On the data regressed, I saw a problem close to the lower points of the data set. The linear function at the low end of the data set forced the MAF curve to be well defined and always increasing to get around some of the limitations of using a polynomial fit. It seemed to work well.

- Charles.

[mrselfdestruct1994]

mrselfdestruct1994,
Glad the spreadsheet data was useful to you. As Al pointed out, in the low range, the exponential nature of the MAF curve can be approximated by a linear function. When using polynomial fitting, the resulting function can do odd things depending on the power used, amount of data, etc. I think Bruce pointed this out in the previous thread. On the data regressed, I saw a problem close to the lower points of the data set. The linear function at the low end of the data set forced the MAF curve to be well defined and always increasing to get around some of the limitations of using a polynomial fit. It seemed to work well.

- Charles.

Hey mate, I found the same thing in my own crude testing. If I tried to include the value of 0mg/sec*10 at 0 volts in my graph dataset it threw a lot of my low end data out of whack. As you and Bruce have suggested I am now using my original polynomial fit data with linear values from 0.35v to 0v. Thanks for all your help.

[durson1]

I can't seem to get this to work for me. With the new 2.9 code. When I try to write to controller, The error message I get is:
error loading inc mapping file:maffactor.inc

I've tried with the default ford maffactor.inc file, and one I offset with the excel file someone posted a while back.
What am I doing wrong? Probably something simple.

MS2 V3 2.900 code.

[durson1]

I'm following these directions from grippo on the previous MAF and MS topic.

"I remembered that the real problem with not being able to download the maffactor.inc table is that the code to do this wasn't in the ini file. So I made a new (attached) ini for 2.891 code - the latest code. Using this ini and TS (NOT MT) this appears to work fine. Here is what is necessary to get this to work:

1. Put the default Ford maffactor.inc file (also attached) in the same directory as you use for the mainController.ini file, but rename it maffactor_default.inc. Copy the file you want to use in its place into maffactor.inc. This file must be of the same format as the one that is attached. It is the file that will be downloaded by TS.

2. Install the new attached ini file, bring up MS2/MT and select the new ini.

3. Go under Tools. There will be an entry: "Calibrate MAF". Select this and you will get a small dialog that says "Write to Controller". Push the button and it will download. Give it some time (watch the blue wavy lines to go away). Then you are done. ONLY DO THIS WITH THE ENGINE **NOT** RUNNING - just with the key on accessory.

4. If it doesn't work for you let me know the error message. Also note that you only get one chance to burn one table. After that TS caches the table and will only reburn the cached table until you close TS, then bring it back up. At that point, copy whatever file you want into maffactor.inc and TS will re-read the maffactor.inc."

I realize these directions are for the v2.891 code. Do these not appy to the v2.900 code? What am I missing?

I had abandoned my MAF dreams, but I feel like I'm tuning in circles in speed density. I'd like to give it another try. Any help is appreciated. Thanks

[grippo]

The 2.900 ini file also has the mod to allow downloading the maf file. One of the instructions I posted referred to MT, but this mod will only work in TS, as it says elsewhere in the instructions. Second, the maffactor.inc file must be in the same directory as the ini file is in.

The error message sounds like there is something in your maffactor file that the table generator doesn't like. If you post your file I will take a look at it, but I won't be able to test it until next Wed as I will be away from home until then.

[durson1]

I'm using tunerstudio. I put the file in my project folder, in the project config folder which is where my ini is. Here are the 2 I've tried, renamed to maffactor_default.inc. My modified one didn't work so I tried the original unmodified one, that didn't work either.