Found a issue. I took a 1hr trip today and right a 3k rpm I hit a fuel cut type deal for a split second then it was fine. This happened one other time but I just ignored it. Both times it happened right at 3k and part throttle not i boost. I wasnt logging tho either time. An I can go try to make it happen when i logging but I cant get it to do it.

 

Ya. I was calling that "jitter". See my previous posts in this thread. I'm traveling in Asia right now so no access to my car. If you could capture it in a log for Roadspike that would be awesome. It's really the only thing that was bugging me on this ROM but I coudn't rule out bad tuning on my part.

 

sometimes I feel it too !

 

No, my map afr 17.2 and sometimes bouncing between 2 numbers 13.3 and 17.2 ! It is the value what the ecu want to follow in open loop ! But in the fuel map it is 14.7 !
With the bouncing between 13.3 and 17.2 doing a loopy idle.
If i rev it a little or cruesing always following the fuel map ! In closed loop no problem afr map is 14.7 and idle is right.

Here is the picture with my setting : link

I know I can make my ve maps smoother, but any help on base settings for kpa to load map how shoud look like on stock or small hks cam like 264 with omni4bar and evo 8 would be nice..
Anyway How to find out that values ? What i Did is cruesing in different gears, speed, to hit as many kpa levels and rpm as I can and compare the afr value with the 14.7 what is in the fuel map until load 140.

KPA to load table : Between the kpa values the ecu is interpolate ? What KPA values shoud be on the left side of the table ? How I think we need better resolution under 0 psi or 100kpa..

 

Tested the evo 9 or 8859.. variant the past couple of days. Works fine as far as I can tell, I've boosted up to 18 psi with no problems and i'm using fix #3. Only one issue may be that the MAT scaling needs to be updated in case someone forgets to copy-paste the correct values.

Made a few minor changes to VE tables to get it running good, but the initial values work well for something to start with. Thanks roadspike and everyone else for making this possible!

 

Looking at your VE map its probably bouncing between the 60's VE point and the 70+ ve point when idling. You realize that the VE map is in pressure (kpa) vs RPM not load so when your adjusting it make sure you log the pressure in kpa so you have something to compare with.

A second look and I see your kpa to load table isn't set up quite in the way I'm used to seeing the evo's. Most of the time the kpa is higher than the load under 100 kpa and past that its usually like 10-15% higher than kpa when over 100kpa.

Personally I would reset the VE table to stock even though its probably not perfect for you and adjust the kpa to load table to get it running closer to where you want it then adjust the VE table to fine tune it later.

Its not really a surprise that closed loop fixes it as the ecu is adjusting it back to 14.7 afr on its own and it can correct up to 25% of error. You should be able to easily get the car to run a consistent 14.7 afr without the ecu's help just needs the right settings.

The ecu interpolates between any two values but basically the left hand side equals the right hand side for kpa to load. You shouldn't need better resolution for the kpa to load table it should be a nearly straight line going up until boost then a slightly higher incline then straight from boost to the end of boost. The whole point of the 3d VE table was to give you the resolution to fix all the oddities for SD.

 

RoadSpike, have I understood this correctly, load used in fuel and ign map lookup is calculated as follows:

Load = 3dVE x (kPa2Load) x (MapVolts2kPa)?

If this is correct, could I set also kPa2Load to ratio 1:1 and adjust unlinearity with 3d VE map (ie setting VE higher than 100% if needed)?

 

If you go back to page 1 i have a big picture of how the maps trace out.

basically its just load = load from kpa to load X VE

The map volts to kpa is just a calibration table for the map sensor.

 

Ok, then it is possible use only VE table to compensate on-boost unlinearity to get correct load for correct fueling? In my opinion this might be easier way, especially if 3d map is moved to RAM and tuned with livemapping. By the way, I have 88580714 + Live if you are willing to patch 3d on that

 

That should work however i don't feel comfortable playing with the DAM livemapping bit. I would like someone who knows that patch to implement it rather than me just guessing at what it should be doing.

 

Hi Roadspike !

Thank you the answer ! Yes I'm logging kpa ( x6.895 my omni 4 bar psig readings ) ! My car idleing at 36kpa maybe because the hks 264 cams ?!

So you said to reset my 3D ve tables to stock and first time only playing with Kpa to load table until my windeband afr is close to my high octane fuel table afr numbers 14.7 under 100kpa ?

Under 100kpa and 3000rpm the numbers in the 3D ve table is not extrem high to play with the kpa to load table ? OR these rom is working complety different as the 2D speed density rom ? I ask that because I saw 3D Ve table made from logs what is looks like completly different exspecially under 100kpa and 3000rpm.

 

Neat chart but the 3D portion of it is not correct. You chart is multiplying the kpa difference from load to the rpm but thats not entirely right.

The correction in the chart which would make sense to change would be the load corrected values.

You could use the RPM VE and see how it suggesting a 70% vs 65% VE @ 500rpm that its suggesting a 7% increase in VE at 500 rpm across all kpa ranges.

Don't use the 3d portion though there just isn't a good way of figuring out the chart from the 2d values. Each cell needs to be individually tracked and thats a better job for evoscan and a custom tracked value for determining the difference between afrmap and wideband vs kpa.

 

That is the chart I created way back when to create a 3D VE table by multiplying out the 2D tables (jcsbanks 2D SD tables). It multiplies map VE% by RPM VE percentage, which (when I created it) is how John had SD running.

 

Nice maybe another chart can be made for the 3d version. Going to be tough though since the kpa to load table would effect the 3d VE table values as well as percentages of change from desired to actual afr.

 

I just recently downloaded the 94172715 ROM provided here by roadspike. I was using 96533706 with no starting issues (immobilizer disabled). Then switched to 94171715_v7 to see if I could get a better running car. Still started and ran with immobilizer disabled. I then copied everything to the 94172715 for us of the 3e VE table and the car just cranks all day. Everything copied to match the previous rom with immobilizer disabled.

Has anyone had this issue? Should I post up each rom to see if someone can give me an idea of whats going on?