Discussion and questions related to the course EFI Tuning Fundamentals
I am considering to re-calibrate my Micra 1.3lt rally car and I was looking into detailed injectors info.
Considering the power output of the car (currently ~120bhp, target ~135bhp) the fuel demand is low. For the moment it's still fine with the OEM injectors, which should be around 140cc/min, and i already have a set of 250cc/min available for future upgrades.
The problem is that I am not able to find any such small injectors being sold with characterization data, or anyone offering this as a service in a reasonable price.
Most of the cleaning facilities offer just a nominal flow value at 13.5V-ish and 3.0bar at best, but no dead times or extended flow information.
I am willing to either measure the injectors I am now using or buy a set with good data, even with slightly higher nominal flow.
What ECU are you using and how does it characterize injectors in the software?
my plan is to switch to a Haltech Elite 1500, so dead times, and steady state flow would be first data to look for. Considering that those small injectors will run at relatively large pulse widths, i suppose that the non-linear area will not be so much relevant.
One interesting technique to determine deadtimes is available if your ECU will allow both sequential and batch (once per engine revolution) injection. Tune the idle in sequential mode. Then switch to batch mode. If the mixture (Lambda value) stays the same then you have the correct dead times, if it is leaner in batch mode, then the dead-times need to be made longer, conversely if it's richer in batch mode then make the dead-times shorter. Change the deadtimes, re-tune in sequential mode, and repeat.
The reason this works is because in sequential there is one dead-time value per engine cycle, and in batch mode there are two deadtimes since the injectors fire twice per cycle. Both should deliver the same total amount of fuel, and this will only be true if the deadtime values are correct.
That's an interesting approach!
To take it one step further, by dropping the battery voltage, even more dead times could be possible to be defined.
I am not sure though if it makes so much sense to go into such depths, for an injection region that it will not be significantly relevant under load in this application.
It'a a nice exercise though, thanks for the input!
PS: What if: How about just re-tuning in batch mode BUT changing the dead time values instead of the main map values? Assuming that VE model is used, the ECU will define the Inj. Pulse between seq and batch.
Given that the AFR target stays the same, the total amount of fuel mass injected is the same between seq and batch.
The total pulse width is controlled by the ECU including the dead time (once for sec, twice for batch).
Then, by adjusting the dead time to hit the given AFR target in batch, will mean that the total amount of fuel will be correct for the seq also, right?
Or at least the repetition loops will be less...
I am thinking of just collecting a database of the available injector data (ID, DW, etc..) and defining "generic" values based on that. A bit of interpolation should bring it somehow close for the purpose.
Then, i will get a set of injectors serviced and rated and live with the consequences...
BTW -- you can power the injectors with a bench-top power supply, then changing the voltage is easy. You will find the VE you determined with in incorrect deadtime, will need adjusting, so yes you can tune in Batch and switch back to sequential -- bottom line is if there is a difference, the dead time is wrong.
One reason not to worry about any of this, is that you just bake any error into your VE table. If you change to larger injectors, you just revisit the table making slight adjustments. We did lots of that and engines ran just fine before there was any Injector Dynamics or other comprehensive characterization. If you changed injectors, you just went back to the dyno...