TdracerTd wrote:If the 0v wide band is set to 0 and the 5v wideband is set to 25.5 then when the wideband signal Is 2.5v (stoich) x will be 12.7.
X at this point will be 127. The 256 comes from the 8bit analog to digital conversion, there are 256 steps (0-255 counts) over 0-5V.
TdracerTd wrote:((1.52-0.5)*(12.7/256))+0.5 = 0.1 then if stoich value = 14.7 so 147 * 0.1= 14.7. If stoich was set to 1(lambda) then 0.1 * 10= 1????
With the X being out by a factor of 10, it would be .. ((1.52-0.5)*(127/256))+0.5 = 1.00
So at stoich of 12.7 AFR lambda is 1.00, as you would expect.
TdracerTd wrote:Just not sure about the bit where the stoich value is * 10. why is that?
The Stoich * 10 is for the XDF tables. As an example, the target AFR table will currently have X / 10 or X * 0.1 as the calculation (gives exactly the same result). With say 12.7 AFR displayed by TunerPro the raw data stored in the ECU will be 127.
So to display lambda, the raw data in the table will still be 127 and we want that to display lambda of 1.00 with a stoich if 12.7. So the calc is X / 127. With 127 being Stoich * 10. The ECU uses AFR * 10 for all its internal calcs.
TdracerTd wrote:I see by setting the 0v and 5v wideband at their limits it gives us maximum resolution on the sensor input to the ecu. This was baking my noodle for a bit.

Correct, the ECU code is setup to work with a larger range than lambda so its not ideal. The settings of 0 and 25.5 so that the ECU does not modify the raw analog value when trying to work out AFR. Instead it lets TunerPro do the calculation. Note, this means that VE learn will not work!