Testing for proper AE fuel

[dennisL]

My AE testing got slowed down after I decided to rewrite the AE portion of the program.  There was just too much useless stuff, like multiply by 1, runtime multiplier, and more.  I'm back into it now.


As to the minimum pulse width issue, I've learned that there is a min usable pw.  Here's a quote from something I found:
[size=0pt][/size]
[/color]Injectors should not be operated below 1.5-2.0msec PW’s due to the non-linear portion of the curve.  At extremely low PW’s, the injector’s coil does not have time to reach its magnetic potential, thus the opening and closing action of the injector becomes erratic.
[/color]
At a "too small" pw the amount of fuel is not predictable.  Each make and size injector has its own characteristic but the 1.5-2.0 msec min appears to be the accepted practice.  I've been using around 1 msec AE for my Lucas 009s but now I'll be testing 1.5 - 2.0 msec to see if I notice a difference.

[motorhead]

http://www.msextra.com/doc/ms3/injdeadtime.html

[$1987 GN$]

http://www.msextra.com/doc/ms3/injdeadtime.html

Thanks Mike, saved me the time of digging in my junk for something like that.


AJ___

Oh just thought of something else too. It would depend on what year you bought those things in . . . 
Need to find my farting graph on the new 42's.
Nonlinear behavior is not the same for all injectors.
I am pretty sure they worked fine ~1ms. 
Someone else has to have the data.
Where the heck is Chuck?

[dennisL]

Thanks Mike and AJ.  That Megasquirt stuff is very interesting.


Since I don't have a flowmeter or Megasquirt I'm using the engine and Power Logger to test.  I can't measure the dead time but the AFR curve shows the effect of low pw's.  I started with a very small pw (< 1ms) and am working my way up.  I can see the AFR actually go lean at the start of AE when the pulse is too small.  There is also a hesitation between gas pedal and engine response.  I'm up to 1.3ms and should be able to test some more today.


Dennis

[motorhead]

No problem.

No throttle cracker modifier in that chip? Might help with that tip in lean spot... but, only after you are happy with the injector tune. It has been a while since I looked at the stock binary.

[dennisL]

No throttle cracker modifier in that chip? Might help with that tip in lean spot...

I'm not sure what you're referring to (?)

The stock program triggers AE with a very small positive TPS increase (0.78%).
The AE pw = Coolant Temp. multiplier x delta TPS.
(There was another multiplier for TPS position, stock value of 1, which I removed).
Add pw to sum of all previous pw's.
[/size]The Max AE pw is selected by Coolant Temp.
[/size]If the sum of AE pw's is greater than Max, then limit pw to Max.  This is 99% of the time.
[/size]Add injector battery offset to pw and squirt!

[/size]I'm not sure what I could add to make throttle stabs more responsive.

[motorhead]

There should be something (a field or modifier) similar to an accelerator pump shot on a carb in the code that should fix things.

[dennisL]

Yea there is ... and that's what I'm trying to get straightened out.  Its called Acceleration Enrichment (AE) and I listed what the program does to create the "pump shot" in a previous post.


Here's what I've learned so far:  the minimum pw for my 009's is 1.358 ms @ 145°F.  The coolant temp. does make a difference, so I have to make the pw larger for colder temps.  I've got the compensation for injector dead-time spot on and the AE is now responsive.  A few more tweaks for very cold and it should be done.  Oh, and during a test drive today I thought my BLM was broke because all the cells were 127 and 128.  The log confirmed the ecm was learning to those values during closed loop.  This is the best I've seen for this car.


Guys, thanks again for getting me pointed in the right direction.

[motorhead]

Yea there is ... and that's what I'm trying to get straightened out.  Its called Acceleration Enrichment (AE) and I listed what the program does to create the "pump shot" in a previous post.


Here's what I've learned so far:  the minimum pw for my 009's is 1.358 ms @ 145°F.  The coolant temp. does make a difference, so I have to make the pw larger for colder temps.  I've got the compensation for injector dead-time spot on and the AE is now responsive.  A few more tweaks for very cold and it should be done.  Oh, and during a test drive today I thought my BLM was broke because all the cells were 127 and 128.  The log confirmed the ecm was learning to those values during closed loop.  This is the best I've seen for this car.


Guys, thanks again for getting me pointed in the right direction.

Sorry... the text in your quoted post was incredibly small.

1) Yup, now I see why you were interested in the "AE";
2) Yes, the AE will be temp dependent, and if we had an IAT input that was worth a damn it too would dictate fueling enrichment at various temps; and
3) At this time of the year (up North especially) there is going to be a fair bit of fuel added to compensate for air density. To this end I would revisit your figures in the dead of Summer to make sure your fuel trims (INT and BLM) don't swing way out of whack in the heat - that is assuming you do most of your driving in the Summer.