Let's discuss VE Tables.
#41
So, adding resolution to the bottom end of the VE table is only useful for auto-tuning?
In my car I only get below 20kPa during decel/overrun, and sometimes it will even come off of overrun fuel cutoff on it's own (I have it set at 20kPa threshold) before the RPM setpoint. I have noticed that both VEA and VEAL have a hard time tuning the lower end of the table (back when decel cutoff was disabled).
But I never saw any advantage to having it tuned with greater resolution as long as the car overruns smoothly and idles well? Why do it?
In my car I only get below 20kPa during decel/overrun, and sometimes it will even come off of overrun fuel cutoff on it's own (I have it set at 20kPa threshold) before the RPM setpoint. I have noticed that both VEA and VEAL have a hard time tuning the lower end of the table (back when decel cutoff was disabled).
But I never saw any advantage to having it tuned with greater resolution as long as the car overruns smoothly and idles well? Why do it?
#42
Boost Czar
iTrader: (62)
Join Date: May 2005
Location: Chantilly, VA
Posts: 79,501
Total Cats: 4,079
I dont let my ECU autotune that low, otherwise, no matter what I do it tries to pull too much fuel from that row and then it will run shitty on throttle off to throttle on transitions and be too lean. I just manually adjust them to be similar to the rows above albiet a little less fuel considering the load and deal with a slightly rich condition when I lift for a split second.
#43
The drop in "VE" with MAP, that is consistent across RPMs, looks like an error in injector dead-time. Dead-time too small.
Also note that injector flow rates go nonlinear at very short injection pulse widths. The injector curve can be described by 2 numbers, at a given system voltage. So full characterization can be done with 2 sets of numbers at different voltages. The OEs go through a lot of trouble characterizing injector flow curves and coding for them, in order to get excellent light-load drivability.
The above correction can be "baked in" with a 3D VE table as Joe has done, resulting in funky VE tables at very low MAP, but this will only be truly accurate at a single system voltage.
Also note that injector flow rates go nonlinear at very short injection pulse widths. The injector curve can be described by 2 numbers, at a given system voltage. So full characterization can be done with 2 sets of numbers at different voltages. The OEs go through a lot of trouble characterizing injector flow curves and coding for them, in order to get excellent light-load drivability.
The above correction can be "baked in" with a 3D VE table as Joe has done, resulting in funky VE tables at very low MAP, but this will only be truly accurate at a single system voltage.
#45
I dont let my ECU autotune that low, otherwise, no matter what I do it tries to pull too much fuel from that row and then it will run shitty on throttle off to throttle on transitions and be too lean. I just manually adjust them to be similar to the rows above albiet a little less fuel considering the load and deal with a slightly rich condition when I lift for a split second.
Joe, what are you using for injectors? Despite having the (now dinosaur) MSPNP gen1, I find that the speed and accuracy of the EV14 injectors compensate for the lack of resolution in the controller.
#47
My TPS arrived today, I have been running mapDOT since forever and I'm pretty happy with the results (after massive headaches tuning the accel enrichment).
Not even sure it's worth the headache now
Anyway, the last row on my map, I also tuned it to replicate the one above (if I remember correctly it's set at 28kPa) and made it just a bit leaner. Works great for me!
Not even sure it's worth the headache now
Anyway, the last row on my map, I also tuned it to replicate the one above (if I remember correctly it's set at 28kPa) and made it just a bit leaner. Works great for me!
Thread
Thread Starter
Forum
Replies
Last Post