I have done the rear O2 hack to allow EvoScan to log my LC-1 wideband via ECU ADC input. However, I notice there is a difference between the data logged via ECU ADC and the COM Port method.
Yes, there is a difference even if I were to log just these 2 parameters alone without the rest which may slow down EvoScan.

My observation shows most of the time, the COM port data is leaner than the ECU ADC's data,
Any idea which would be the more accurate data to base my tuning on?

Edit: I forgot to add that the rear O2 ADC data is matching the figures on my LC-1 gauge.

 

use the COM data to get the ADC math right
you will need to change the formula till they match up

 

Are you saying the COM (Serial) port data is more accurate compared to the rear O2 ADC data?

My current formula is x*0.0147+7.35 which is for the default 0-5v (7.35 AFR-22.39 AFR) and it matches the output on my LC-1 guage.

 

that formula should be correct assuming the defaults in lmprogrammer

what I do to verify the ECU/LC1 is not being affected by ground offsets, is in LMProgrammer set the LC1 to output 0v then log the voltage in the ECU/ADC

then set it to 1v, and relog
2v, log
3v, log

etc

if your logged voltages match the settings you used in LMProgrammer then you have a good connection...