Battling AE again.

bluetrepidation · Post by **bluetrepidation** » Sat Jun 17, 2006 4:26 pm

Here is the situation.....

With my car idling I can blip the throttle. Now RPM comes up slightly and the engine begins to slow down. The throttle is completely closed at this point causing my MAP vacuum to be very high. Higher than at idle. Now I can blip the throttle again and this time the engine goes dead lean for a second and sputters.

Now Acceleration Enrichment. When this happens I can see me MAP accel slam to 300 kPa/s. MAP vs TPS is set to %50. My top bin is set at 20 ms @ 240 kPa/s. MAPdot is at 125. And my MAP lag factor is set to 60.

Now under realtime display when this happens my Acc value only peeks at 11ms. Shouldn't it be 20 ms at the least? How come I can't get this to dump in enough fuel?

MAP based accel seems to be the only value that "sees" this gulp of air from high vacuum from transitioning from engine breaking to applying power. Got any ideas how to get this to dump more fuel?

Thanks,
A.J.

Bernard Fife · Post by **Bernard Fife** » Sat Jun 17, 2006 4:45 pm

Please note in your post if you are using MS-II/MS-I/MSnS-E, the code version, MegaTune version, as well as V1.01, V2.2 or V3 main board, etc.

bluetrepidation · Post by **bluetrepidation** » Sat Jun 17, 2006 5:13 pm

MSII

2.36 code

V3 board.

http://70.226.38.10/~admin/MS/datalog20 ... 6accel.xls

A.J.

bluetrepidation · Post by **bluetrepidation** » Sat Jun 17, 2006 5:30 pm

OK I set things to 100 % MAP based just to see what happens and I found that it now dumps the 20 ms shot of fuel. Something is wrong. Even if I set my MAP lag factor to 100 the fuel seems to make it to late. I get a lean blip followed by extremely rich AFR. I posted a datalog of what is going on. My setup has always done this and I'm kinda lost at this point. Stinks considering this is a stick shift car and when I shift gears sometimes the car bucks very hard. Other times it shifts smooth as it was stock. Two things seem to have a factor. Low RPM and transitioning from engine breaking to accelerating. If you just let the engine sit at idle and blip the throttle it is much much less noticeable. Under load things are a bit on the rich side when acc kicks in but it is very acceptable and drivable.

A.J.

bluetrepidation · Post by **bluetrepidation** » Sat Jun 17, 2006 5:35 pm

If I get time tomorrow I'm going to move where my MAP sensor is connected to. Currently it is hooked to the vacuum line that supplied vacuum to a emissions valve that doesn't exist anymore. This line comes directly off the throttle body. The stock MAP sensor is located more down the intake. Maybe this is my problem. I have a old MAP sensor I'll drill a hole in and mount a line up. This way I'll be getting a reading off the stock sensor location. This is the last think I can think of to do.....

A.J.

Jedrik · Post by **Jedrik** » Sun Jun 18, 2006 10:25 am

I wonder if you are actualy getting too much AE fuel and causing a missfire that is readding as lean but actually rich?

You are getting prety huge fuel pulse width. (20ms) I have found about 50-75% req_fuel is good AE for me.

MAPdot can be troublesome because it tends to give negative values as the engine accelerates even though the TPS stays constant.
For this reason decel can not be used off of the MAPdot.
Personaly I do TPS only accell.

A trick I like for testing this area is to set the fuel cut revlimiter with a very low recover RPM then you can just open the trottle and leave it open letting the rev limiter halt the acceleration. It makes it easyer to see what the EGO is doing in the datalog.

mops · Post by **mops** » Sun Jun 18, 2006 12:45 pm

jerdik - i think it dependent onthe car.

they only way i could make it work on my engine is by mapdot.

actually i developed this strategy for tuning AE (applies only to MS2): i set my mapdot PW's high and my TPdot to all 0's. then i just move the slider to give the right amount.

bluetrepidation · Post by **bluetrepidation** » Sun Jun 18, 2006 2:24 pm

Do you mind posting your settings?

mops · Post by **mops** » Sun Jun 18, 2006 2:47 pm

setting will difffer from engine to engine.
my guess is 20ms is flooding engine...
If i was you i wouldnt bother with AE at all, just get 2.672 code and tune with x-tau.

put mapdot values to simulate logaritmic curve. try to overshoot (i think 20ms is massive overshoot). ... you might wanna even start with sometinhg random like 2, 5, 7, 9ms....set all tpsdot values to 0.

set the slider to 100% tps (i.e. accelenrich = 0)

rev the engine. notice behaviour. try revving engine from like 2k rpm onwards. idle point is not a good place to start AE tuning. aswell, when you rev, try to be consistent with throttle... including consistent speed.

now move the slider to, say 30% map, 70% tps (effectively 30% mapdot values). repeat testing. notice difference ? yes ? no ? better, but still not there ?

move the slider some more depending on the behaviour. repeat...

krisr · Post by **krisr** » Sun Jun 18, 2006 2:59 pm

Interesting approach to AE tuning, I quite like it!...

FYI, X-Tau = Thumbs up!