In a previous post, I promised to test the effects of the injector timing on a sequential system.

I tested at 2500 rpm, with no load, and with the EGA hooked up at the exhuast manifold, before the cat. This is on a 1L, 3 cylinder. I tested 60 and 20 degrees before Intake valve open, 40 degrees after, and 100 degrees after open. This angle is the injection end angle, meaning that's when they stop.
In line with theory, the best emissions were with the injections finished before the intake valve opened (20 degrees before opening). The HC's were 138 ppm, CO .67%
There was no significant difference between -60, -20, and +40.

The worst emissions were with the injections finishing at 100 after opening. The HC's were 340 ppm, CO .88%. When this returned to idle, there was a definite misfire, with terrible HC's. I repeated this by going back to 20 before open and saw good emissions, and then going back to 100 after, and got bad emissions a second time to rule out some other mechanical failure. The second time also resulted in an idle misfire.

I didn't have time to do it at load, but previously (same type of engine), I tested at WOT and 5700 RPM (on an engine dyno) and didn't see any power difference with injection timing. I wasn't measuring emissions then.

Just some evidence for the skeptics.

 

Yup, emissions only feature, not power.
At peak power the injectors are (should be) open 80% of the time or more meaning injector timing would do diddly squat. With the short pulse widths at idle the DC drops drastically and the sequential would be great.
I wonder if the newer SEFI pcm's adjust the firing angle based on pulse width duration... I'd think they would.
BTW, is there any chance you could run this test on a larger engine closer to a sbc?

 

Most sequential systems use larger injectors to avoid going over 60% duty cycle. This always keeps the injection on the closed valve. I'll be on the engine dyno again today, I'll try some really bad injection timing on there at high loads, see if it's noticeable.

Slower processors may set the beginning angle vs MAP and RPM. We set the end angle, based on the calculated PW adjust the beginning angle accordingly, and once the injection starts, we update the PW and thus change the end angle.

I'm in China, working on an emissions calibration / debug our new ECU, so not much for sbc's here. Sorry.

 

Sounds like fun... except for the lack of sbc's

 

"The worst emissions were with the injections finishing at 100 after opening. The HC's were 340 ppm, CO .88%."

i wonder if you leaned it out some if they would come down, & also if the motor would run without the misfire that way.

i know the CO would come up, but what about the HC if it were richened up some?

 

But why? It's obvious to me that it's better the other way. The control difference in the tests were the injector end angle. It was left in closed loop (which is a requirement), and the closed loop at no load is a little bit lean biased in the calibration.

I have an SAE paper from ford that they did in the late 80's (880691) that shows the same results. They were also measuring the cylinder pressure, and were seeing terrible consistency with fuel shot onto an open valve. Cold engine was REALLY bad.

BUT!! They also tested with Air Assisted injectors and saw really GOOD emissions with shooting on an open valve and closed. Likely due to not wetting the wall (and cylinder wall) so much and the better atomization in general.

 

While talk about emissions, and WOT HP is nice, they're not the only concerns.

 

System current load less transient (single injectors firing in sequence rather than groups of 8).

PW increases.

Combustion stabilizes. Cylinder wall wetting decreases. Plug wetting decreases. Cold starting improves.

Cylinder to cylinder variation decreases.

Less variation allows more precise timing.

gotta go eat. what else?

 

You caught the cold start part....

How about, just overall drivibility?.
What's nice for cruise, makes for a lousy idle, and what works on a warm engine ain't right for a cold one. And what works for cruise is way off for going towards WOT/PE mode.

I'm beginning to believe that plays a large part in injector sizing. The smaller the injector the longer, it's on, and that increases the likelyhood, of some of the injectors, being right once inna while (timing wise).