Let's hope somebody could shed some light on this, especially if you know what multiplier is VCDS using (ME7 logger is using 3.05176e-005) as it would explain why the data is off if they are using wrong values
The data you see from ME7Logger are the most accurate values you will get in this case
.
Variables fra_w, frao_w, frau_w are stored as unsigned word (16bit) in the ecu.
Damos specifies the conversion from internal to physical value for these:
PHYS = INT / 32768, this returns values in range 0.0 to 1.99997 (a factor without unit).
The conversion factor which ME7L uses is therefore 1/32768, which is just 3.05176e-005.
Now, when you log with VCDS, the ECU processes each variable and performs an internal
conversion, then delivers for each variable three bytes (a, b, c) to VCDS.
Byte c is a conversion formula number
(see here:
http://nefariousmotorsports.com/forum/index.php/topic,22.msg22.html#msg22),
bytes a and b are data values used in the conversion formula.
VCDS takes a and b, then converts again using the formula and presents you the result.
In case of fra_w, frao_w, frau_w, the ecu always sets c = 20, so the formula for VCDS is:
a*(b-128)/128 [in %].
The ecu sets a=50 and b to result from a conversion of the original 16bit value.
You can see, that VCDS can compute the final percentage result with a resolution
of 50/128 = 0.390625% in a range of -50% .. +50% (b can be 0 ..255).
Be aware, when you log via KWP1285 (as VCDS does), the ecu alway processes variables
internally before delivering data and you probably loose accuracy.