KA24DE-T Injector Resize TP Question

[Matt]

You can see here the rise in injection time with TPS increasing (MAF of course too) goes from 8.4ms to 7.2ms over a 0.4 second period

TPS_Enrichment_inj_time.PNG

(12.56 KiB) Downloaded 1181 times

[Matt]

The throttle enrichment itself is not causing the lean spot. The 'hump' or lean spot we are seeing appears not to come from injection time (since that is steady (as we can see from the GREEN line in your log, but the AFR purple shows the hump)

The short term trim (light purple) in your log happens before the AFR lean spot. So try adjusting the maps in this area first

[jongerlings]

Hi Matt and OP, sorry to bring up this old thread, but I am currently dealing with a very similar issue where I have a lean spike/delay when shifting up between gears at moderate cruising speeds/loads. Issue sounds extremely similar to what OP describes. My problem seems to be timer related as well, as the problem is far less noticeable when driving the car harder, under higher loads/rpms, and quicker shifting. Problem occurs the worst when taking more time between gears. It seems like maybe this issue is from a fuel cut/recovery timer or setting? It seems to me that if the throttle is released, there is a period of time where I can "catch" the next gear in time to avoid this lean spike. If I shift too slowly, there is about a .5-1 second delay before my throttle input makes the car accelerate.
Wondering if anything ever came of this thread? Searching similar terms here also brings up some even older threads with the same issue, also KA24DE ECUs...


'96 KA24DE ECU
Turbo

[Matt]

We would really need a log file to see if it is the same, so we can see injection time other parameters, if you record it

[jongerlings]

Hi Matt, sorry it took me a minute, but here is a log. Log starts from a standstill, and is just an easy drive up to 40ish mph easy through the gears. Lean spike/hesitation happens between 1-2 and 2-3 pretty consistently.

[jongerlings]

And here is the bin file, if you need it.

[Matt]

Attached is the image from your log file traced over the tune file

We can see when the RPM (red) drops and we have injection time about 0.93ms min, which is about the same time as your latency (0.94ms = 940us)

So this is happening on deceleration when there is a very low TP (calculated load)

The TP index is about 8 (so minimum of the table).

The problem with these later ECUs (ones with the VE map) is they have no minimum TP table, which lets us increase the lower limits of when the calculated load drops.

Since the injection time is calculated directly from TP, this may be why you are getting those lean points. With R34/WC34 I ended up adding a minimum TP table to it

What you can try here is either
1. Increase the min TP load value in the fuel map, and see if this affects the logged minimum TP value (currently 8, so maybe change to 16). See if this affects idle fueling however
2. Increase fuel values in column 8 around 1600-4400 rpm to larger values (start around 150) and see if that helps
3. Increase the MAF VQ map values around where the low load is reported, This seems to always be about 1200-1360 in the VQ map (3 cells). Lifting these values will increase the calculated TP, so increased perceived load