KA24DE-T Injector Resize TP Question

[iyarovoy]

Good morning,

I am finishing up my KA-T project and have a question about setting nistune.

Quick overivew of the car:
- T28 Turbo
- Stock KA24DE MAFS
- Subaru STI 550cc Injectors
- Looking to get 200HP

I have not used Nistune before the Feature Pack and now the whole K-constant and Total Injection Multiplier is screwing me up as I cannot find a definitive answer.

Because I am not modifying the MAFS, only bigger injectors, can I just use the TIM to get the car into a ball park area, and then use the Fuel Maps to tune for AFR at cruise and boost? Or do I need to modify the K-constant too and then use TIM further to fine tune?

My understanding is that since I did not change the MAF, my TP will stay the same no matter what size the injectors are, since the MAF Table is the same.

Using only TIM will keep all my tables related to TP the same.

Thank you for the help

[Matt]

Just do the injector resize (Operations > Resize injectors) which will change only TIM for your setup. You can fine tune TIM at cruise (keep close to 0% fuel trims) to reach stoich mixtures

You are correct, there is no need to rescale or change K constant (or any change to TP load)

After this just a matter of adjusting the fuel maps for your turbo setup. Just watch your MAF voltage, since standard MAFs measure about 10-20% more than standard normally before topping out (as a generalised figure)

[iyarovoy]

Finally got around to starting the car and tuning. I started with a KA24DE base tune, resized injectors + latency for Subaru yellows (535cc 1050uS latency) by selecting "TIM" + "Crank enrichment" option in the "injector resize" window. The car started right up.

Took it out for spin around the block and did sligh boost pressure (3psi) run and saw my TP load values go into the 140s (stock base s13 KA24DE max TP value is 104 in all the maps.) Called it quits right then and there, realizing I don't fully understand the situation I am in.

What I didn't think of is that the STOCK MAFS I am using is rated for around 200hp yet the stock KA maps are tuned for a maximum of 155hp aka max TP of 104. In other words, the MAFS can read more air but the NA Stock KA map is only setup for what the engine could produce without turbo.

I could not find a clear answer on how to overcome this situation and how to modify my K value + TIM to leave all the maps at a max TP Value of 104 (aka the stock KA24DE map).

Seems like my plan is:
1. Start with a base KA24DE map.
2. Resize injectors and select the TIM option + crank enrich tables (this should allow the car to start)
3. Remove 10 degrees timing from every cell beyond the close loop area in the ignition map (or the TP value at and above the fuel map close loop area)
4. Adjust K-const until Max TP reached is 104 when the car is in +psi range
5. Adjust the TIM in the opposite direction of the K-const to counter the effects of modifying K-const
6. Repeat steps 4 and 5 until the TP value does not go off the map.
7. Fine tune idle and cruise via TIM.
8. Put back the 10 degree timing.
9. Carefully find what TP the boost transition is on and adjust the timing and ignition maps to the right of boost transition accordingly.

Does that sound like a good plan or is there a better way to take an NA stock map and modify in for a turbo and bigger injectors?

How do I then adjust the crank enrich tables in the end?

[Matt]

In other words, the MAFS can read more air but the NA Stock KA map is only setup for what the engine could produce without turbo.

This is correct... so work from your current map and lower TP further (say 20%) and increase TIM by the same (20%) and see how far the max TP reaches when you reach your target range

Yes your plan sounds correct

[iyarovoy]

Hi Matt,

I would like your help again in identifying the parameter that is causing my lean partial throttle for a 0.5 - 1 second after applying throttle (ex. shifting gears)

Capture.PNG

(34.58 KiB) Downloaded 3869 times

.
I can see the AFRs go lean as I step on the throttle, feel the power go away, and then comeback when the "Enrichment" is over an the AFRs return.

You can see the injection time follows the TP slope for a certain amount of time then spikes up which is when I see my ARFs return to normal, as if there is some table adjusting I am missing.

Any input would be great. Dont mind the fact that TP goes off the scale, I set up such that I have safe fueling and reduced timing when I am troubleshooting this issue and going into a boost and slamming into the last column.

Thank you for the help

[Matt]

I've done a compare from this file to 53F1F

- Auto trans is switched off (are you running a manual with AT ECU?)
- Alpha/N table set to 0 (normally set 255) but I do notice tickbox disabled

As for the graph, the picture doesn't show anything, so I've taken my snapshot

You can see TPS (grey) going up well before the lean issue (at 0:33)
Also the MAF (blue) is well up at 2.8V before the lean issue
TP (green) similarly is also stable at this point'

At the 0:33 point you see the injection time has already increased, but then the AFRs shoot lean, until the injection time (purple) jumps from 5.86 to 8ms injection time

TPS_lean_issue.PNG

(84.19 KiB) Downloaded 3857 times

I'm not sure why thats happening. I'll check on the bench

[iyarovoy]

I'm running Subaru STI yellow injectors which are about 500cc but have a latency of 1050us (source: Subaru forums) which seems a bit high considering nissan 500cc injectors are about 700uS.

I'll try lowering the latency down to 800us and lean it out Via TIM to see if that has any effect on this lean issue.

EDIT: The Latency + TIM changes did not improve the issue. Same dead spot for about a second at throttle tip in. :/
EDIT2: Yes I am running a Auto ECU on a manual transmission (Did the 5 speed swap many years ago)

Thank you for looking at this

[Matt]

Must be something to do with the auto code in this ECU (maybe?)

I've been pulling your tune file back to the original 53F1F (bar the fuel map)

This looks similar to a problem we saw on some S15 SR20DET ECUs, when you change the fuel map substantially, and then hold at a certain load/RPM it would hold lean and suddently jump (about 4 seconds)

Other parameters don't affect this (just the fuel map)

I'm seeing this here on the bench. We added an extra table 'EGR injection offset' to S15 and looks like we have that table also for the earlier models. If you replace your current address file with the one (attached) you will get this table. We normally set to 0 to remove the problem

See screen shot:

S13_KA24DE_53F1F.png

(177.08 KiB) Downloaded 3831 times

So fuel has been increased in the maps because you need more fuel on boost (assuming turbo here?)

Try adjusting this table (perhaps set to 0 firstly) and also increase the map more gradually (currently it is very flat for half the map) and this might help remove the jump you are getting. Let me know how that goes

[iyarovoy]

Very exciting! Thank you for looking into this. Ill get to it tonight

I too think it's related to the Auto ECU code or some throttle enrich table we are not seeing. If this is true then the 20 year myth that you can use an Auto ECU with a manual transmission will go out the window as there were some minor side effects in doing this swap (fuel economy, jerky fast shifts).

I looked at the base address file (S13_KA24DE_256_E.adr) and a couple of maps I found commented out that seemed of interest to me. Anything you can tell me about these:
#Unused
THROTTLE_ENRICH_FAST_MOVE_TEMP,&H0000,0,0,0,0
THROTTLE_ENRICH_FAST_MOVE_RPM,&H0000,0,0,0,0

The car has a turbo and a TP ~75 is 0 vacuum. I realize I did not compress the NA map very well yet, but it will try to reduce the rate of change for the fuel.

Do you think this will go away if I use the same Nistune board on a S13 manual ECU and load in a manual KA24DE Feature Pack bin? The base image is a auto one for my board so would it need changing? Just thinking of a backup plan.

[Matt]

If this is true then the 20 year myth that you can use an Auto ECU with a manual transmission will go out the window

This all depends on ECU... most cases I would use a manual transmission base image in the auto ECU with no problems. This includes the later NEO ECUs which work fine

However several models (S13 KA24DE, B13 SR20DE and possibly P10 SR20DE/S13 SR20DE) do not like manual transmission BIN files loaded on their auto transmission hardware. What I have found is that there is a prematurely early RPM limiter which occurs when putting MT code in an AT ECU.

My assumption is that there is a logic chip in the ECU which is programmed specifically for AT/MT part numbers to match the ECU code. I will need to check if the actual ECU code between AT/MT is the same (and only the maps, and flags are specific to trasnsmission type) since as you know, the tickbox which is the AT/MT specifier which tells the ECU to use AT or MT processing (particularly interfacing with the AT controller, but also other functions but I've never dug into that part too deeply)

Given that we have most of the maps covered, you should in theory be able to upload MT maps into the AT base image and it should run appropriately. However not every AT/MT parameter is defined in our address files either so no guarantees

Anyway see how you go with that EGR map adjusted. Those two tables are removed since they are in the 'base' SR20DET address files but not for this SR20DE ECU (that I have found)

[iyarovoy]

No success on the new "EGR" Map, still lean for 1 sec then injector time spikes to where it is supposed to be.

I used the MT Image and copied the Fuel Map, Timing Map, and a few other TP related parameters keeping everything the same, and the issue still persists.

I attached a new log and bin file for review.

I will try disabling closed loop, maybe there is a "Close Loop Off Delay" parameter somewhere that keeps the car in closed loop for X amount before letting it go.

Thank you again for your time

[iyarovoy]

Adjusted latency to 850uS and offset fuel via TIM, no luck, still lean for 1 second and stalling at decel

Turned off O2 Sensing, still lean for 1 second.

Not sure what else to try. Any Ideas?

[Matt]

I'll have to look at it again once I get back from vacation. Will check it out on Tuesday

[iyarovoy]

Tonight I went through most of the Expert Parameters and set them to 0 or 255 to see if I can find one that changes the behavior with out any luck.

I also corrected the "Neutral Position" sensor via hardware such that the flag is never on thinking this might fix the lean issue and that had 0 effect.

Im going to try to remove the AT Control Unit from the car to see if it has any effect on the issue.

Besides this Im not sure what else to try.

[Matt]

Sorry, I've had some bad family news after returning this week, so just trying to keep on top of things.

The AT control unit will not make any difference. I'm needing to try out a few things including the MT verison of the ROM, and also the AT ROM without FP using your fuel maps to see if it does it from factory. Once I round down the exact cause, will let you know and then work from there