How to code for larger injectors

New tuner here. When installing larger injectors what is the proper way to tell the ECM how much fuel is now being injected for a given pressure and pulse width? Would that be done in the pulse width table? According to the injector flow sheet, there is a lot more fuel going in that my current table shows. Next, there are only a few data points on the flow sheet. I could put those into a spreadsheet and fit a regression equation to fill out the entire table. Or am I way off base?

Brandon,

It sounds like you're on the right track with the regression equation between the tested data in the Pulsewidth table. Check this thread out for more details. Feel free to post back up if you get stuck.

https://www.hpacademy.com/forum/practical-diesel-tuning/show/tuning-larger-injectors/#post69786

Nick

I was able to get a good plot of mm3/uS at 160MPa since the flow sheet has three data points for 160MPa. It came out very linear. I have one data point at 25MPa and one at 120MPa. Without a second one at each pressure I can't plot them to see if the slope of the line is the same. Has anyone here done this before? I could make a pretty big assumption and say the slope of the lines at all pressures are the same, just shifted, and it would probably be closer to correct than my current table.

Since I have 60% over injectors, I wondered about just multiplying my whole table by 1.6. When I compared the flow sheet to my PW table I found that they deliver 45% more fuel at the low end and as much as 113% more on the high end. Guess that shortcut is out.

Starting to wonder if this is even a worthwhile exercise on a non-emissions engine.

Brandon,

Without a fairly granular calibration sheet from the supplier it is absolutely an artful exercise.

With 60% over injectors you'd want to divide by 1.6 - not multiply (unless you're talking about the mm3 axis breakpoints. You may also consider subtracting out injector energize time before doing your multiplication/division and then add it back in when you're done. You can see this time by looking at the .1mm3 cell for each pressure.

-Nick

Nick,

Yes, multiplying the fuel qty axis is what I meant but worded poorly. This IS probably more of a learning exercise than anything else. If I ever am successful at building out a new PW table, I guess the next thing would be to go back and command more fuel instead of simply tricking the ECM with larger injectors.

My latest idea was to multiply pressure times pw for all 5 data points on the flow sheet and plot the product. The graph for the 60% injectors is below. The data lined up really well and the equation fit well too. I can use the equation to calculate an expected fuel qty for a given pressure and PW with the thought of using that to rebuild the mm3 axis.

I repeated the process with the same manufacturer's stock replacement injector sheet and the line looked identical just with lower fuel volume (which is what one would expect). As a test I took the stock PW table and applied the regression equation from the previous step hoping to answers would match the fuel quantities on the axis for each table column. (i.e. all the pulse widths in the 60mm3 column when cranked through the equation with their corresponding MPa would yield roughly 60mm3)

Results were less than exciting but now wondering if the injector energize time you mention plays a role. The general trend was for the equation to over predict at very small mm3s then moving right across the table it would under predict at low pressure and eventually over predict at high pressure. In the example of 60mm3, it predicts 39mm3 at 30MPa, and 72mm3 at 180MPa. Almost dead on at 100MPa.

Not sure how to factor that energize time in. 5mm3 is the first column that has a meaningful pw for all rail pressures. My 1mm3 column runs up against the min pw limiter at 40MPa which renders it almost useless since the desired rail pressure in never under 35MPa except in hot restart cranking.

Hey Brandon,

Nice job on the line fit. That's pretty clever!

With regard to electro/mechanical latency calculation, if you take the pw from the 10mm3 cell and subtract the 5mm3 cell the difference should represent only the time needed for additional fuel flow (call this 'f'). On the other hand, the pw listed in the 5mm3 cell represents the fuel flow + latency (f + l). Subtracting your value 'f' from (f+l) should give you a pretty accurate latency number. Keep in mind latency is effected significantly by pressure so you'll want a good value for each pressure column.

Hope this helps,

Nick