I need clarification on the different modes. Closed loop referes to the high.lo octane fule and timing map correct and open uses O2 sensorfeedback right? and how do i tell when the car is running the open and closed loop mode. I also have another question, in the injector scaling thread, there was one guy who posed injector latency values of several injectors and specifically PTE880cc which is what i run, when i uses those values, i had to revert my injector scaling in ECUflash from 748 to 560 to get them to run correctly with the injector latency values posted up, so my question is: were those latency values i used meant to scale the injectors via the latency table instead of the injector scaling value? Also, my LTFT mid are between 2-3 but my STFT dive in and out usually from -15 to 25 max when i'm using the valves provided to me for the latency table with my 880 scaled to 560.

 

Thanks for posting this here.. I'll answer the questions in the next post..

 

Open loop is what runs off the Fuel maps
Closed loop relies on O2 feedback

This is the simple answer, it turns out theres a little more to it, in that the fuel trims are what adjust, the percentage of difference in the fuel trims is what the ECU determines from the Open loop map. In other words, the closer your open loop map is to the closed loop AFR target at idle and cruise, the closer to zero your fuel trims are.

Welp, there is a flag in MUT that tells you specifically, however its not documented and nobody knows which one it is. The best way to infer it is to look at the O2 feedback value, if its 0 and stays there, its probably open loop. A more complicated way to know is to look at the TPS Voltage, and ECULoad and compare it to the Open Loop TPS and Open Loop Load table in the ECU, if the value is higher than that RPM point, or the load is higher than the other load table, then its in open loop.

OBVIOUSLY, using O2 Feedback is the easiest way to tell for most people.

Injector scaling is a trial end error endeavor considering how new this is to everyone. This is the reason why Injector scale and latency values aren't frequently posted.

1) Injector Scale sets the rough size, you can use this to get the short term fuel trims close to where you want it.

2) If fuel trim low (Idle) is richer than fuel-trim Middle (Cruise loads) (regardless of if its -10% and -3%, just that there's a difference) then you can adjust the latency until both trims are roughly the same (-10% and -10% for example)

3) Finally you can increase the injector size slightly (if the value is negative, meaning trims are leaning things out), or decrease it (meaning the value is positive meaning you need to add fuel), until the trims hover within +/- 10% or as close to 0 as possible. The closer to 0 you are, the more likely the AFR values in your fuel maps will come close to what the actual AFR's are in your wideband, its never 100% accurate, but it will let you get your mind around the values easier. Keep in mind this will vary due to Intake mods and other things that can alter/throw off the MAF sensor.

Ultimately, you'll probably rough in the settings, go for a drive doing pretty normal things, and come back and adjust it again until your happy..

LTFT's are the critical value, as they are the average running condition and shouldbe as close to zero as possible. STFT's are the immediate adjustment required to offset LTFT's for immediate conditions, if they are diving around at idle, it can mean a few things, but typically its load related, the better the injectors are scaled, and if you know there are no issues such as leaks, or aftermarket MAF pipes, filters, partially open DV's and whatnot, the less it will shift like that.

 

Oh, let me point out that the trim values are assuming 0 = perfect trim, in Evoscan the scaling is different, and the value would be 100 = perfect trim.. In OBD-II scanners, the MUT tool, and Mitsulogger, its 0.

 

Off topic question, will evoscan .98 now work with DLL without a conversion tool? And does mitsulogger work with DLL without conversion?

 

As far as I know, Evoscan "Should" work. However I never updated the conversion tool to support any newer versions.

Mitsulogger was designed specifically to be compatible with Data Log Lab, and all you need to do is use the provided Definition file, any additional items added to the logs can be added to the definition so they show up also.