Working on it. We need a code disassembly guru on board though. We need help tracking down a few things.

-Jamie

 

That is things like this that makes me wants to continue study the SH and learn dissasembly but my god we will understand ISCV in 2017 if that relies on me.

Jamie what I meant about wrong defs was:



Obviously not even close to point the right adress. These should be anything from ISCV to absolutely something else.

 

Thanks guys.

Without the right definitions for all the demand% tables, some encountered issues could be unrepairable because like Jamie points, the right way to start investigate demand% tables is to zero all the demand% and test them one by one. The problem is that there is still some demand% after zeroing the current tables. We'll be stuck on this until we have someone to find them.

For the ISCV guide, I'll finish it with Jamie and post it ASAP so we can update it with our results. This will make things cleaner.

But for tonight I still plan to annoy the neighboorhood by cranking/idling/logging my car .

 

The IATS table doesn't even have the correct address for IAT scaling. Clearly it uses the same scaling as other IAT related tables we already have, not a coolant temp address.

I will tell you this, you will only find the startup table to show a loggable difference with all corrections disabled. Other tables...not so much.

On an Evo 8 rom which has the idle stepper table raise incrementally to 160 steps as coolant temps rise, you may actually see the cold start idle INCREASE as coolant temps increase because of this.

I do plan to build a tutorial on what everything does since I have been working on these settings for a long while now over the past year or two when I had time. I have a lot of notes to go back over still.

-Jamie

 

Probably wont happen now, because development on the Tephra ROM for Evos 7-9 and such has kinda died, but it would have been great if they just simplified the idle setting to a low number of tables, without so much code and flags and such.

 

I agree that most dissasembly efforts are going to the X, and this is okay.

 

So I logged the car at idle with the engine ice cold and coolant temps at 66*F until it exceeded 180*F with every single table zero'd out, even the initial ISC steps-AC OFF table. Only the Idle stepper table was left untouched.

I noticed that the STARTUP table is the primary ISCV Demand% table at startup. I set the startup table to all zero's for testing. Once the RPM exceeds about 500rpm, the ECU starts interpolating towards whatever 'target' ISC step value it has. In this situation, it started at 0 ISC steps and interploated UP. This interpolation seems to last about 2.5-3.0 seconds and then the startup table no longer has any effect on the ISCV. Idle RPM settled at 700rpm and slowly rose to 1000rpm by the end of the logging session. Target idle was 1398rpm, tapering to 850rpm.

The vehicle instantly settled at 18.43% Demand and stayed locked there the entire time until I ended the log after coolant temps exceeded 180*F.

I also noticed that the Demand TRIM did not move at all until coolant temp exceeded 180*. Even though the Demand TRIM was showing movement, it had no effect on the ISC position.

I then reset the demand trims and restarted the car and noticed everything basically acted the exact same except after the startup 3 second timer was done, the ISC settled at 6.27% demand. Idle settled at around 800rpm.

I then reloaded the Intial ISC Step table settings. At 180*F the Demand% is 2.75 in this table (most see this table as ISC steps as a value of 7). Now when the car settled in after the startup timer, demand% was 9.02% (6.27%+2.75%).

So it seems as though some tables are primary tables that switch under certian conditions and others are additive to whatever the BASE value is, which I dont think we have figured out yet.

-Jamie

 

Nice work. So mrfred had the VIII initial steps named and scaled correctly "ISCV demand CTS adder" for the IX.
To reach zero trims with all these demand tables left at zero, one could calculate:
18.43% (target idle demand) - 6.27% = 12,16%
This car need 12,16% Demand in the "initial steps table" at 180F. Roughly 31 steps instead of 7. Obviously this is because other tables are zeroed out.
Did you continue putting back values in the other demand% tables? I bet that if some have effect on demand% at idle, they would be absorbed by the trims readjusting to reach 18.43% for his target idle.

 

I'm totally lost by what your math was referring to from my post. Why were you subtracting the full temp demand% at idle from the cold start demand%?

 

I just realized you seemed to appear to be subtracting what you think to be a demand trim of sorts. There was zero demand trim in my entire test session. The values are direct values, most likely from an unknown map that may be the true 'base' idle map.

 

Nevermind, I tought the car at operating temp and at target idle was at 18,43% demand. That was the part I wasn't sure.

I meant if a car need 18% demand to idle at target, to have zero trims you need the adjust the initial (your 2.75%) so that the total of the:

initial + unknown base demand = 18%

I need a

 

Tomorrow I might try shutting off and restarting the car at each coolant temp value in most of the idle tables. Maybe I can capture a curve of data and give us some hex code to search for in the rom coding.

 

Good idea, tell me if you need some of the colder ones.
It could also help to know where to search in the ROM, maybe you could get the average routine location from mrfred.

 

I can go through the code again if you tell me what you are trying to figure out. Prioritize what you want.

 

Here is what I have done so far to try and make an educated guess on what parameters may be in a demand% table that we haven't found. Whatever this table is, it is only a starting point for the demand% and it appears that it expects the short trims to take over to make the rest of the adjustments since once the engine starts (i.e; 9417 "Idle Error", 8859 "ISCV Demand Target Idle Trim"), the demand% doesn't move until restart at a different coolant temp. For this reason, I shut off the car and restarted it and logged the initial demand% starting point at each of the coolant temps listed below. This was done with EVERY known table zero'd out except the IDLE STEPPER TABLE.

I used the coolant temp tables used in all of the other idle tables as my target temps to try and build this curve. I started at 68*f as this was the lowest temp I could start logging at.

coolant temp scaling:
-26
18
45
68
93
122
171
180

Demand%
-26F
18F
45F
68F 18.0392%
93F 14.5098%
122F 8.62745%
171F 6.27451%
180F 6.27451%


Hex8
-26F
18F
45F
68F 2E
93F 25
122F 16
171F 10
180F 10



I don't know if that combination of code can be found easier in the 9417 rom or if this data is a combination of more than one table still. The coding is still something I have yet to have time to dive into heavily.

Also, I'm not sure if the address' are correct for the following tables either. The IATS table uses a coolant temp scaling instead of IAT temp and the right column is definitely not the correct address.

The Baro table is all 100% demand so I expect that to be incorrect as well.

-ISCV Demand IATS Adder**** (FFFF7C0E) (sub_1F6E6) [sub_1EFC0]

-ISCV Demand Baro Adder**** (FFFF7C0E) (sub_1F6E6) [sub_1EFC0]

The IATS temp scaling seems like it would use an address like 689e

-Jamie