Re: Spark scatter at idle, possibly $8D specific.

If you have ever run a scanner on your car, you might see a phenomena which I would call spark scatter. Your engine speed is relatively constant at idle but your spark jumps all over the place...at regular intervals that seem to have a method behind the madness, but where is it coming from?

There are tables that are only used for ‘idle spark’ situations. Basically, they are correction tables used to trim engine RPM instead of using massive IAC movement. You are probably all familiar that if your car is running and you advance your timing, your engine speed can increase (until it gets normalized again acc’d to the desired RPM).

Looking at your data to justify this may be harder than you think, especially if your idle RPM only fluxuates from say 800-850 and your desired is in between those values. Also, the data rate of the ALDL is only going to show you about 7-8 frames/second, so more stuff is taking place behind the scenes than what your scanner is showing.

Idle spark advance retard vs RPM:

<BLOCKQUOTE>code:<HR><PRE>

From ANHT_HAC,
;-------------------------------------------
; IDLE SA RETARD vs RPM
;
; TBL = SA * (256/90)
;-------------------------------------------
; S.A. RPM
;---------------------------
L81EB: FCB 0 ; 0.0 0 RPM DIFF
FCB 14 ; 5.0 50
FCB 14 ; 5.0 100
FCB 14 ; 5.0 150
FCB 14 ; 5.0 200
;-------------------------------------------

and the table looks like this:

Idle SA Retard vs RPM Error

RPM Error Idle spark retard
0 0.0
50 5.0
100 5.0
150 5.0
200 5.0

</PRE>[/QUOTE]

There is enough information there to add a new table to Tuner Cat using his TDF editor (free for 30 days).

Same for Idle SA Spark Advance vs RPM

<BLOCKQUOTE>code:<HR><PRE>

; IDLE SA ADVANCE vs RPM ERROR
;
; TBL = 256/90
;-------------------------------------------
; S.A. RPM ERROR
;---------------------------
L81F0: FCB 14 ; 5.0 0
FCB 14 ; 5.0 50
FCB 14 ; 5.0 100
FCB 14 ; 5.0 150
FCB 14 ; 5.0 200
;-------------------------------------------

and the table looks like this:

Idle SA Spark Advance vs RPM Error

RPM Error Idle spark Advance
0 5.0
50 5.0
100 5.0
150 5.0
200 5.0

</PRE>[/QUOTE]

For testing, I basically set all values in SA retard to 0 and all values to SA advance to 9.8* except for the first entry, which I left at 0. If left me with the following data (at idle).
http://expert.ics.purdue.edu/~msandi...SA_TEST_01.xls

Since my spark changed ~9-10 degrees, it appears that I was actually in the Idle SA Spark Advance phase based on RPM error...didn’t have a couple spare EPROMs at the time to change a lot of stuff, so it was a slower process than I am used to.

I’m open to comments, suggestions, and questions. I’ll look through it some more tomorrow, hopefully getting into the MAP error scaling that it does (which the ECM leaves as the same at all RPMs).

Good Luck, -Matt

 

Good info. Thanks!

 

Above are the factory values in an AUJP BIN. Might need to click the pic to see full version.


[Modified by HighHopes85, 1:05 PM 5/8/2003]

 

Hey, thanks! You just answered another question of many that's been rattling around in my head!

 

HighHopes,

I have seen this and wondered the same thing.
Have you done more research into it?
I agree, it looks 'intelligent'....is it tied to the 02 crossings at idle somehow?

I don't have a clue but mine does the same thing...

BTW, is your system MAF or SD?

 

Hey Jim,

I think I finished up all my testing here...
http://expert.ics.purdue.edu/~msandi...SA_TEST_02.xls

(if you have Excel, you might have to right moiuse click and Save As)

For some reason unbeknownst to me, test C/bin C gave me problems with stalling when hot in gear and moving from park to drive.

Test D gave me a rich smell at idle even though the BLMs were still pretty close to 128.

BTW, I switched over to Speed Density because I could do it for 1/4th the cost of replacing my burnoff module or another failed MAF.

 

how do I print that in my edtr?

 

TunerCat's TDF editor. Free for 30 days, I believe.