In reading through the mathematics behind the MAF calculations, I've decided to see how it all works in APPLICATION. I don't even have a chip burner yet, but I have played with a hex editor, some .bin editing programs, and some data acquisition software. I apologize for any mistakes that I make because of that. My reason for posting this is because I want to see how this fits together in the real world, and I'm hoping that you guys could verify this for me. If this goes according to my plan, there should be a few "major" sections that I'd like to discuss with specific questions in each section, hence the "Part I" you see above. The following math is based upon formulas and techniques that I've learned here on the DIY PROM board. A lot of the underlying stuff is from RBob; his understanding of the code and ease of explanation really helped me understand what's going on. All that I've done is merely rearrange some things -- I don't want to take credit for what I didn't do. Here goes nothing.

OK, in the "real world" situation, you want to see how rich or how lean you are by logging data from the car's ECM. The BLM that shows up at a given point is specified along with an airflow reading, which is measured in grams per second. Say, for example, that the data acquisition software indicates that at a given point, there are 38 gm/sec of air being read by the MAF sensor. Looking at the appropriate MAF table (#2, in this case), 38 gm/sec doesn't fall nicely on a row. What I want to know is "what voltage from the MAF sensor indicates 38 gm/sec?" This is what I've come up with.

MAF TABLE #2 SCALAR = 48

tbl gm/sec volts row
119 ; 22.3 ; 1.46 ; 0
133 ; 25.0 ; 1.55 ; 1
147 ; 27.6 ; 1.65 ; 2
163 ; 30.6 ; 1.74 ; 3
180 ; 34.1 ; 1.83 ; 4
198 ; 37.1 ; 1.92 ; 5
217 ; 40.7 ; 2.01 ; 6
237 ; 44.5 ; 2.10 ; 7
254 ; 47.7 ; 2.19 ; 8

1. First, find the TBL for the measured airflow of 38 gm/sec.
TBL = gm/sec * 256 / scalar
TBL = 38 * 256 / 48
TBL = 202.66, when rounded becomes 203

2. Find which row the calculated TBL starts at. In this case, 203 starts at 198 (row 5). Find the difference between rows. This will be used to calculate the "depth" of 203 into row 5.
(Calculated TBL) - (starting TBL) = (TBL difference)
203 - 198 = 5

3. Since the MAF table only contains nine total rows, the interpolation is doubled. Therefore, there are 32 "spots" between each row for a value to lie. To find the depth into the row:
(TBL difference) / 32 = (depth into row)
5 / 32 = 0.15625

4. We know that the TBL value of 203 is somewhere along row 5.x, x being the depth into that row. Add.
(depth into row) + (starting row) = (full row determinant)
0.15625 + 5 = 5.15625

5. Find the lookup argument that was initially used by the ECM in the first place.
(full row determinant) * 32 = (lookup argument)
5.15625 * 32 = 165

6. The lookup argument is part of a hex number that is calculated from the MAF voltage. This number contains two parts; one that determines the MAF table used and one that contains the lookup argument. See RBob's thread on table lookups.

6a. We know that MAF table #2 is being used, so the first part of the number is $02.

6b. The second part is the lookup argument converted to decimal, which is $A5.

6c. Put the two numbers together and you have $02A5.

7. Convert that hex number to decimal: 677.

8. The ECM's internal value is the last number that needs to be computed to determine MAF voltage.
(MAF decimal number) / 7 = (ECM's internal value)
677 / 7 = 96.71, or 97.

9. Finally, the MAF voltage can be determined.
(ECM's internal value) * 0.02 = (MAF voltage)
97 * 0.02 = 1.94

10. This yields an answer of 1.94 volts.

Does this method of determining MAF voltage from airflow reading show validity? 1.94 volts falls appropriately between rows 5 and 6, but I'm not sure of how accurate all of this is since the ECM rounds numbers.

As of right now, I'm not concerned with how this relates to tuning MAF specifically. I'd like to take it one step at a time and see if my logic is correct. Again, I don't take credit for doing this. It's merely a rearranging of formulas and equations that have been discussed on the DIY PROM board before.

 

looks good. now the next step is to take those airflow readings and show how the ECM process's that into a PW at the injector.
if youve read my other posts id like to actually use larger and More maf tables and take full advanatge of the ecm's 8 bit a/d converter. right now its handicapped it would be nice to use it to its full advanatage. care to help. ?

 

What all is involved with this? I know practically nothing about programming, let alone programming in assembly language. Though I probably wouldn't need extended MAF tables yet on my car, it'd be an interesting thing to undertake and make work.


Incidentally, I think I found an error in my logic. In step 3, the difference between rows is divided by 32. However, this isn't correct. Note the difference between rows 5 and 6 -- 19 units. Therefore, in step 2 it's actually a fraction more like 5/19 than 5/32. To fix this, I set up a quick proportion and set 5/19 equal to x/32. X turns out to be 8. The final answer still comes to the same thing (1.94 volts), but for other TBL numbers it's more accurate. For example, assume that the TBL was 216. That's a difference of 18 from 198 (row 5). In my old method, that would be 18/32. Hardly; it's 18/19. The similar fraction in 32th's is 30/32, which is much more accurate. That yields a voltage of 2.0, which is just shy of 2.01, the voltage for row 6 .