OK, so in the current pair of defs, there's an extended parameter called Gear (Calculated)*. In the description of the value on openecu.org, it says that if one has non-standard gears or tires, the Gear Determination Thresholds need to be corrected for the relevant discrepancy from stock. Trouble is, I can't find a Gear Determination Threshold table anywhere. What gives?

I did the N/A defs first. The 'Gear Determination Threshold' was a late addition that I decided on while I was working on the turbo defs and I didn't go back and add it to the N/A defs. It will make the next release, however. This table is only needs to be changed if the transmission gear ratios, final drive, or overall tire/wheel diameter (that is, tire + wheel height) has changed from the factory specs.

Yup, I understand under what conditions it needs changed, I just couldn't find the table.

I just noticed this table and tried working out some changes in excel, but ran into some problems...

I'm using the formulas of:

(gear ratio*final drive ratio) = gear final drive ratio (engine revs per wheel rev)
and 1 Mile / Tire Circumference = tire rev / mile
then wheel r/m * gear fdr = engine rev / mile

Here's what I came up with for the OE tires and ratios...

10469.70973
5901.715357
4140.597421
2946.310902
2237.013833

Compared to the ECU values of:

7762.11
5003.14
3539.61
2615.4
402.34

I figured I might have an error in the tire circumference calculation, but it turns out that the correction between calculated and ECU is not constant. (compensation for tire expansion at speed perhaps?) Here are the corrections (convert calculated to ecu values):

0.741387316
0.847743359
0.854854901
0.88768636
0.179855839

So, I figured doing the calculations for different gears and tires, then applying the corrections would suffice. But I ended up with the following deltas in ECU values:

336.345362
9.98657423
-14.5358475
-40.74671466
-15.73077723

To me, this seems completely wrong.... Shouldn't they all be positive or negative? Also, 5th gear seems really off, even in the ECU. If the change was due to a factory compensation for tire expansion at speed, why is the 5th so different? I understand the force expanding the tire is proportional to the square of the wheel speed, but it still seems off.

Is it simply because of the way subaru determines the threshold from the actual?

I've changed my final drive (3.7->3.9), gear ratios (06 5mt -> PPG Helcal Syncro), and tires (215/45 -> 225/45), so I'm sure the ECU is all kinds of confused about the gear determination.

I assume the ECU simply does RPM/(60*MPH)?

Furthermore, what does the ECU use this information for?

Here's what I used to do my calculations:

20168/(Tire Diameter) * (Gear Ratio * Final Drive)
20168/(2 * (Tire Heigth * (Aspect Ratio/100) / 25.4)) + Wheel Height) * (Gear Ratio * FD)

So for my car, it was (RA gearing 1st)
20168/(2 * (225*0.45/25.4) + 17)*3.083*3.9 = 9710.46

These numbers seem to jive with yours (fairly close) so I don't know why the ECU is so far off. On the v7 ECU I am using the numbers I worked out were pretty close.

Hmm, was your 5th gear ECU number in line with the others or way off like mine?

Here's what I came up with for the OE tires and ratios...

10469.70973
5901.715357
4140.597421
2946.310902
2237.013833

Compared to the ECU values of:

7762.11
5003.14
3539.61
2615.4
402.34

You need to look at the labels for each cell. These are thresholds to determine the gear, so they are the closer to the midpoint between each gear.
7762.11 Min 1st Gear / Max 2nd Gear
5003.14 Min 2nd Gear / Max 3rd Gear
3539.61 Min 3rd Gear / Max 4th Gear
2615.4 Min 4th Gear / Max 5th Gear
402.34 Min 5th Gear / Max 6th Gear

So,

6th gear -> 0 - 402
5th gear -> 402 - 2615
4th gear -> 2615 - 3539
3rd gear -> 3539 - 5003
2nd gear -> 5003 - 7762
1st gear -> 7762+

So, compared to your calculations
10469.70973 = 1st
5901.715357 = 2nd
4140.597421 = 3rd
2946.310902 = 4th
2237.013833 = 5th

Even though this is an 06/07 WRX (I assume from the numbers), the logic to determine 6th gear is still there.

I figured it was something simple I overlooked :) :p

I added a column in the tables to calculate midpoints, and the numbers came out pretty close, except the 1/2 and 5/6, which makes sense.

Another thing I forgot, was that the ppg's change around all the ratios, so that explains some of the variations in deltas.

Thanks for clearing that up Teacups