racingmini_mtl wrote: but remember that what is good for idle and driveability may not be what's needed for maximum power at WOT and high RPM. There has been some discussions on this topic and some interesting technical information but I can't find the posts I have in mind at this time. But a search on sequential and injection timing should give some results.
Jean
Jean, this whole thing is your fault. I needed to dig into this since you brought it up...
Now that I have at least part of a handle on my idle tuning I started digging into the pile of contradictory information on the 'net regarding tuning injection timing. After several hours of research I could find a source to prove just about any strategy you could want. At the end of this short odyssey I decided to try an approach that was advocated by some of the more reputable looking sources I could find.
The general consensus is that properly timed injection timing helps noticeably at low load and short injection pulse widths, there is almost no advantage at all between batch fire and timed injection at WOT.
The tuning approach I am experimenting with uses the theory that you want to center the fuel injection event around the point of maximum intake air velocity. Now I don't have access to any expensive engine analysis software or a dyno so I am using a rough approximation for the timing of this event. Maximum intake velocity will occur somewhere around the maximum lift on the intake lobe of the cam. In general, maximum velocity will be slight before maximum lift at low RPMs and slightly after maximum lift at high RPMs. On my cam the intake lobe center is at 252 degrees BTDC.
I looked through a couple data logs and found that on my engine my injector pulse widths were between 4ms and 6ms in the low-load operating area between 1500 and 3500 RPM. To keep me from losing my mind with the math I picked the middle duration of 5ms for my additional calculations. I little more math showed me that a 5ms injection pulse takes 30 degrees of crank rotation at 1000 RPM. Applying this 30 degrees per 1000 RPM I get the following:
1000 RPM - 30 degrees
2000 RPM - 60 degrees
3000 RPM - 90 degrees
I did all this to make sure that if I centered my injection event around maximum intake lift that the injection event would end early enough in the intake cycle to give me a feeling that all the fuel might make it into the cylinder. This last part is definitely a guess.
For my first attempt I decided I wanted my injection events to end no later than 240 degrees BTDC. This means that my idea of keeping the injection event centered on the cam starts to fall apart between 1000 and 2000 RPM but I was hoping it would still be close enough to see a benefit from all this math.
I set up my injection timing using these rough numbers for all the bins outside of my idle range. I ended up with injection end timing between 260 and 240 degrees BTDC, maxing out on the timing around 3000 RPM where they are all set at 240. I wanted to stick with injection end time and not use injection centered timing because I still felt that I wanted to make sure the injection event finished within the desired intake stroke. I felt is was better to spray gas on the hot, closed intake valve prior to it opening rather than letting the end of the injection event cause gas to pool in the runner waiting another 540 degrees.
I plugged these numbers into my MSQ this morning and drove to work with VEAL adjusting my VE table which was within a couple percent of ideal on my previous injection timing. At the end of the drive, VEAL had to reduce almost all bins on my VE table by between 3% and 7%.
As expected, the biggest decrease in required fuel was at low-load with only a couple percent change at WOT.
So, from my point of view it looks like the theory of injecting fuel close to max intake air velocity yields a more efficient mixture on my engine than spraying fuel on the closed valve when not idling. For idle, it still looks like the fuel on the valve timing works better for me. The engine *did* feel smoother at cruise with the new timing, my butt dyno is not sensitive enough to know if there was any increase in part-throttle power. I'm guessing there was a little increase just because the mixture was more efficient and the engine felt smoother.
I think I will go through a series of experiments now where I get me VE table dialed in well with a given timing, then change one or two columns and allow VEAL to tell me if the mixture is more or less efficient. It's not as good as a gas analyzer or dyno but it appears to work.
The "two times around the circle" diagram in the other post did not help. Either I'm missing something that is stupidly obvious (entirely possible), or this only makes sense if you have worked in the code.
Thanks.
