So I got to thinking about it yesterday... there is a program that can control A/F and timing (UTEC) and read A/F ratios from various widebands. How hard would it be to write a software program that automatically read the A/F ratio at any given RPM and add/remove fuel until it reached optimal and advance timing until knock? Doesn't seem like it would be that much more difficult, but then again I'm no software engineer...

ps I'd pay around $100 for this software...

 

there was some scuttlebut about this a while back re: the UTEC. a software guy with a Subaru UTEC wrote an interface program that was graphical instead of the HyperTerminal UTEC interface. it also had the ability to do a "Target AF Tune" or something like that to try optomizing the maps based on info from the wideband AFR,

if i remember correctly alot of people were uneasy about software making critical changes to fuel and timing, which could be potentially fatal if done wrong on a FI motor.

 

There are techniques to do this safely but the software in things like a UTEC just isn't that sophisticated. I've written industrial controls for boiler systems that do that exact same thing. The exhaust O2 is measured and is fed back as one of the tuning parameters. This keeps things tuned very nicely until a sensor fails. The problem is that if you get a signal from a sensor it's hard to tell if it's correct. The way around this is also predict from the systems history what the O2 should be for the present conditions. Then test the O2 sensors signal and only use it if falls within a reasonable range. You won't see this in ECUs for a while.

K

 

Ken, how is that different from the usual OBDII closed-loop operation in ECUs?

Genuinely interested. Thx!

 

It is closed loop. The difference is in how it handles the O2 sensors failure and allows it to keep running without causing damage.