Originally Posted by cfg83
...everything I see in the better-mileage version looks like a close match to the EFIE... maybe the EFIE is based on the better-mileage version,
Other than basic concept, I doubt it -- they use two totally different approaches.
The better-mileage version uses a variable-threshold Shmitt trigger scheme. In non-geek-speak, they're choosing what they
want to call the middle of the sensor's range and sending the ECU a pulse train developed around that value. They use a LED dot/bar-graph driver IC to pick off different thresholds (unusual but clever), which means you get up to ten discrete adjustment settings.
The EFIE instead simply develops a constant (but adjustable) voltage and sums that to the signal already coming from the O2 sensor, almost as if you were to insert a battery in the signal line. Rather than discrete steps, it's infinitely adjustable.
I haven't studied either in depth, but at first glance both designs look like they'd get the job done just fine. Depending on the results you folks report, I may tinker with this in the future myself.
P.S. I read in some other threads that the wide-band (five wire?) sensors have a calibration resistor that tells the ECU how to interpret the reading; I wonder if on those you could just change the resistance to "remap" where it thinks 14.7:1 lives?