Is anyone else here running a BARO referenced fuel pressure regulator, with fairly large injectors for your engine?
I have a 1368cc 4cyl with 650cc DW injectors, and am finding that at ultra low load (sub 30kpa) and near the beginning of the injectors nonlinear region, there’s a large deviation in the required VE values - and overall, the VE table doesn’t trend how one would expect.
I suspect this is because the injector cal. / injector deadtimes are set for only the base fuel pressure delta, which cannot account for the pretty large effect that fuel pressure delta has on deadtimes.
Is there currently a way to account for this in the EMU V3, without having to use the VE table as a workaround (it’s not very consistent)? I’m sure the team is very busy with all of the existing work being done, but if there could be fuel pressure delta based deadtime tables that would be incredible - or perhaps there is a straightforward to achieve what I need, that I am not aware of.
My current plan is to try using a MAP vs Inj PW table, calculating effective pressure delta based on map, and then applying a PW offset based on those delta values. Does that sound like a workable solution?
EDIT: It appears the above plan won’t work, as I can only make a correction table based on Injector DC and MAP, not Injector PW and MAP - which makes it more complex as now there are additional factors that affect the axis.
I expirienced something similar like u. The deadtime is not a function of delta fuelpressure in this pressure range. If u are sure that ur deadtimes are correct then ur injectors are simply not suitable for such low injection times with the given pressure.
U simply can’t reach ur target lambda und u won’t reach it even with complex maps.
The only solution is to increase injection times to get out if this critical zone by lowering lambda target in low loads until fueling is stable. Then ur VE will also be normal again.
i am still dont know what fuel pressure regulator selection i have to select. i run an external fpu that has a vacuum connection so normaly if there is lets say 2bar of boost and i run 3bar fpressure, the fpu raises fuel pressure to 5bar , so i have constant 3bar on the injectors.
do i have to select map refference or do i have to select map refference + fuel pressure sensor?
You should be using MAP Referenced (as your FPR is connected to your inlet reference via the vacuum line).
Your base pressure is whatever your regulator is, without the engine running. So in your case if its a 3 bar regulator, your base reference is 300kpa. The ECU handles the rest based on your MAP reference.
Do you have a fuel pressure sensor connected to the ECU and calibrated? If not, then dont choose “+ Fuel Pressure Sensor”, just “MAP Referenced”.
For sure, the injectors are on what I’d call the edge of being acceptable in this application ( I sure do wish they were a little smaller, however). The charecterization data I have is from Deatschwerks directly - I’d hope it’s somewhat close to accurate.
I agree it may not be the biggest factor, but deadtime certainly is a function of pressure at this operating state, with a sensitivity of around 0.12ms per bar, around the 470KPa pressure delta that I idle at. I suppose the real question is if its enough of a factor to warrant an additional axis. While its not huge, it ends up being an error in the deadtime of around 13%. Curve-fitting the deadtime data yields an actual deadtime of 0.934ms at idle, however the ECU uses a value of 0.858ms in that case.
The actual time value isn’t particularly large, but a 13% error in deadtime, when the injector is only energized for 1.5-1.7ms is certainly notable, with that error making up 8% of the injector energized time, and almost 19% of the actual injector open time (the error that ultimately shows up as a…curious…VE table). (Unless EMU reports PW with deadtime subtracted, which would make the actual error smaller). In the ultra low load cruise state, the error becomes monstrous in comparison to open time, which is where a lot of my issues are, as this error introduces its own nonlinearity in terms of its impact on VE required.
I’d be perfectly content with just using a correction table - I just can’t since there is no injector pulse width axis option. It also makes sense why this may not be the easiest to implement, as you’d need two parameters for injector pulse width in order to have a table with a given axis that is modifying the parameter that is, itself, the axis. You’d need Inj PW before modification, and INJ PW (corrected). It’s also completely valid if this is low on the priority list for the ECUMaster devs, given how much is going into V3 at the moment.
Update: as of 3.038, the injector deadtime table has been given another axis (effective fuel pressure) which resolves this issue entirely, without any workarounds.