The problem remains that at the same CLT and MAT (AIT), the injectors respond differently after a hot restart than under continuous conditions. Hence another variable must come into play to be used in adjusting the dead time change due to fuel / injector winding temperatures.

For now, I will continue with tuning steady state with rich idle and let EGO fix hot restarts. MAT correction will be used to keep high kPa AFR consistent over varying AIT as measured at the cold side of the FMIC. If we get some data points on the same set-up with return vs returnless, then I may make the change

One further question: is the problem solved with low impedance injectors driven by proper circuits? Just looking for more learning.

Is the output of the flex fuel sensor zero to five volts? If so, we could replace the flex fuel sensor with a temperature sensor down in the fuel rail area or a surface mounted fuel rail temperature sensor as an input for fuel enrichment on hot starts?

Hell, if that isn't an option how about a simple map switch based on this temperature input? Have your alternate fuel map 5% richer than your standard map and have either a manual switch in the drivers control area to activate it on hot starts, or have a temperature sensor in the fuel rail area to do automatic switching based on a break over temperature.

How does the OEM ECU handle this issue? We all badmouth the Band-Aid piggyback controllers... but they don't have issues like this :(

Keith

you could easily tableswitch fuel maps (or just the req_fuel) based on the fuel temp. that's incredibly simple.

the mazda ecu probably uses the pressure sensor in tank after starts to control the pump in a better fashion, not just on constantly. maybe pwm in certain situations. i dunno. returnless never made sense to me.

Returnless is used to prevent heating of gas in tank and therefore decreases the evaporative emissions.