Results 1 to 2 of 2

Thread: Looking for clarification on wideband transfer function

  1. #1
    Tuner in Training
    Join Date
    Apr 2016
    Posts
    47

    Looking for clarification on wideband transfer function

    Car is a 94 Mazda RX7, LS1. Battery was relocated to the passenger storage bin behind the seat. Originally, the NGK AFX wideband was grounded on the battery (power is also ran to battery). The signal was pretty jagged, though, and moving the ground to the ebrake mount for a chassis ground smoothed it out a ton.

    So, assuming I want a 14.57 stoich reading from the Scanner, I used the built in transfer function. With the car warmed up, and the engine turned off, I was looking at a 14.20 reading on the gauge, and a 14.25 reading in the scanner. A few days later, I wanted to touch up VE a bit, so I adjusted it to .714xxx + 8.95. I recorded it again with the engine running, and noticed a big discrepancy. I ended up changing it to .714xxx + 9.25 to get it to match. Fast forward a bit, I turned closed loop back on and was driving home. At idle, the Scanner was reading 14.5x pretty steady, and the gauge was reading 14.1-14.5. I realize that the gauge is a slower readout, but the low swings on the gauge when the Scanner was so steady had me curious as to how accurate it all was.

    Tonight, I turned the key on (cold, engine off) and logged it to get the raw voltage readouts from the startup procedure. It went from ~.98, to ~3.94, to ~0.01, to ~4.83. These numbers would fluctuate +/- .01 from the stated values.

    So, based on the wideband transfer function spreadsheet found on this forum, I either get .6756 + 9 (based on .1 to 4.83 range) or .6900 + 9 (based on 0 to 4.83 range). Is this correct thinking? I can test it tomorrow, I just don't want to waste time chasing my tail trying to get this to be accurate. I don't want to focus on the gauge reading, but I want to be sure I'm not way off base. Everything is pretty close, but I want to dial it in and it's difficult to dial it in with questionable data

  2. #2
    Tuner in Training
    Join Date
    Apr 2016
    Posts
    47
    FWIW, I ended up using the .69 + 9 number, and it was a bit off. I settled on .69 + 8.9, and the numbers seem accurate. I'm going to take a look again later today to verify.