Wasted -> Seq & Injector Opening Times

[TheMonkey]

trying to sort my base tune today. found that it was idling really nice when in wasted mode, but as soon as i plug in cam sync, motor went super lean and it didn't like it.

i had measured my opening on a scope last year, and found that the kink in current was at 1.1ms.

but, today, i find that i have a nice wasted idle at 1.8ms, and seq at 3.2ms. this gets the same AFR and similar behavior. little algebra suggests that the difference should have opening time set to 0.4ms.

ideally, i'd like to have a similar tune whether in wasted or seq.

i'm going to adjust opening time according to this method, but i find it interesting that this opening time is so much less than measured on the scope.

[Joe]

This subject interests me as well. Have you tried adjusting the injector timing to see if that will bring things inline??

Just thinking out loud but I am wondering how much of a difference is actually the injector opening time and how much is the timing of when the fuel is injected. I've found a huge difference in idle quality and stability by playing with the injector timing settings on a regular MS. Just curious how much effect this will have on true sequential.

Joe

[grippo]

Lokk at what you get on the bench when you go from wasted 1.8ms to sequential. With a 1.1 ms open time, assuming that is what is in your msq, you should get 2.4 ms in sequential. That is in line with what I have seen on the bench with similar PWs and open time. I believe something else is going on when you switch to sequential, but check it on the bench. You can adjust rpm or req fuel or whatever to get a 1.8 ms pw in wasted, then plug in cam sync and see what it goes to.

[TheMonkey]

Hey Joe-

yes, i did fool with timing.

on the bench, i set my cylinder offset & timing delay so that my seq squirt starts directly after my exhaust valve closes.

my settings for that are 2 cylinders offset, and 45% delay. iirc, this is 30 degrees after the beginning of intake stroke, but i would have to go back and confirm.

today, i found that in sequential, the timing did not make a difference whether i kept that setting, or went back to the the default of 0 offset, 0 delay - which pretty much squirts right after the TDC. but... keep in mind, my base tune is still very roughed in.

i did find, that little tweaks in opening time made fa very big difference. for instance..... i tried bumping it down just a little from 1.1ms to 0.9ms.... in wasted it was idling at 11.5 AFR, then i plug in cam sync and sequential it would idle at 14.8 AFR. i had to go all the way down to 0.4ms.

my point is.... opening time has a very big difference on AFR conversion from wasted->seq, but the fuel timing really didn't change AFR at all (only 1 data point, but it was an extreme change in fuel timing). i suspect that when things start to get dialed, proper fuel timing may smooth idle some, but not because it changes AFR.

Scott.

[TheMonkey]

Lokk at what you get on the bench when you go from wasted 1.8ms to sequential. With a 1.1 ms open time, assuming that is what is in your msq, you should get 2.4 ms in sequential. That is in line with what I have seen on the bench with similar PWs and open time. I believe something else is going on when you switch to sequential, but check it on the bench. You can adjust rpm or req fuel or whatever to get a 1.8 ms pw in wasted, then plug in cam sync and see what it goes to.

Al-

it seems to be doing just that.

the test i mentioned to Joe.... where i was in wasted, then plugged in cam sync was with opening time set to 0.9.

from the datalog, it appears to have gone from 1.9ms to 3.0ms. that calculates to within 1ms of expected.

so.... you can see with opening time set to 0.9 was still too much opening time based on how much the AFR moved. i'll attach it. you can see IGN variable swap from 0 to 1 when cam sync comes in.

i was able to smooth out idle much more since this datalog.

[racingmini_mtl]

Mike,

The opening time is kind of a misnomer because it also includes the closing time (actually opening time minus closing time). It would be better called something like dead time since it is the time during which fuel is not actually being injected compared to the commanded pulse width. The physical opening time will contribute a positive value to this because there's no fuel being injected and the closing time will contribute a negative value because fuel continues to be injected.

I must say that I'm surprised that you get what seems like a long closing time of about 0.7ms if you scope an actual opening time of 1.1ms and see that the dead time is 0.4ms from watching the AFRs. But that is a possible value which will depend on the injector drivers and your injector characteristics. If that value is what is needed to have correct results then you should use it.

Jean

[Joe]

TheMonkey,

I've tried both methods for injector "Dead time" Scope vs injector squirts. And actually had good results with both. I have yet to be able to measure injector "dead time" , compare to scope calculation and be able to change between the number or injections and get the same AFR. For my sequencer project I sent my injectors out and had "Dead time' measured and voltage correction measured. One of the things I was hoping to test with my injector specs was the AFR difference between bank and sequential.

I hope the data I recieved on my injectors is valid. If there is a difference it would lead me to believe that the number of injections and timing are the variable. Anyway this has always been interesting to me. Injector "Dead time" to me seems difficult to calulate without a bench setup (my opinion). Mine were measured by yaw performance. Not sure how Paul actually does this.

Anyway congrats on getting it this far. I look forward to finally getting my hands on a sequencer.

Joe

[TheMonkey]

maybe today, i'll get the idle dialed for smooth conversion, i'll share my results.

here is the scope shot from last year that i came up with 1.1ms. this was in batch mode from one bank of 4 (blue scale is on the right):

[Bruce Bowling]

I assume that in wasted mode you are only squirting the port injector once per 720 degrees crank - is this true? If you are to squirt multiple times, then any error in open/close time fuel will be magnified.

In fact it may be possible to hone in on the open/close fuel time in wasted just by switching the number of squirts and seeing if there is a trend in the change in mixture - the more times you squirt the more dominant the open/close time fuel addition becomes.

But there is more to this...

- Bruce

[Bruce Bowling]

on the bench, i set my cylinder offset & timing delay so that my seq squirt starts directly after my exhaust valve closes.

There is something else to consider on the inconsistent mixture readings b/w wasted and sequential....

There was a SAE paper published way back in the 70's from Bosch where they were running sequential injection and they moved the injection timing around a full 720 degrees engine cycle. They monitored the emission output from the exhaust while doing this. What they found was a spike in HC output whenever the injection was near the valve overlap time. Makes total sense, it is possible for the mixture to get pulled from the intake to the combustion chamber to the exhaust during this time.

So, when in batch mode (I was using wasted term above, we are talking batch injection) many injectors fire at one time. Some may be during valve overlap period while others are on a closed valve. I know from fact that any UEGO (wideband) sensor is extremely sensitive to any unburned hydrocarbons (HC). This will screw up readings galore, and since there are unburned HCs this means that the mixture changes and there is also a change in CO and H2 constitutes. The UEGO sensor is sensitive to HC, CO, and H2.

What you need for this test is access to a 5-gas analyzer (a 4-gas will also work). Monitor the HC and CO output in batch (one injection) and sequential modes. You need the separate gas quantities measured to gain any insight, a UEGO reacts to all of them and one cannot quantify the effect.

Also interesting, most OEM controls use an injection strategy known as closed-valve injection. This is where they inject fuel always when the intake valve is closed (or as much as they can). They predict when the intake valve will open, back off from this point and come up with a number that keeps the squirt within the valve closed time (as much as possible). Sounds silly, but there are reasons for this:

1) Injecting on the valve when it is closed will help vaporize the fuel. The intake valve is hot and the fuel hitting it will become a vapor and mix with the air in intake tract.

2) If you squirt when the valve is open, the stream of fuel can hit the cylinder wall and this washes off the lubricating oil. This is supposedly why engines can now get 200,000+ miles (Km) without any bore wear.

3) The emissions are reduced in this mode, and there is lessened chance of reversion effects.

Just an interesting tidbit...

- Bruce