For the USDM roms without the MAT sensor, are we going to have to use the fuel temp sensor patch for wiring in our IAT sensor? If so, is there a possibility of wiring the IAT sensor into the MAF IAT sensor wiring? Not sure of the difficulty in code change for this.

The only reason I ask is because I have my GM IAT sensor in my UICP right next to my MAF basically, so it would be really easy to tap into the wiring harness there than run the connection back to the ECU. Since I'm an USDM 05 and still on 96940011, I'm guessing this will have to be for the 9653006 ROM.

Whatever we have to eventually do, I don't mind. I was just wondering if there were going to be any options in implementation. I'm just happy this is all happening so fast and hope to have something that I can test soon. : )

Edit: Actually, it doesn't matter...I have my SSi4 behind my glove compartment anyway, so I am already running my IAT sensor wires all the way to the ECU area. But, curious as to the answer to my question anyway, if anyone knows.

 

You could wire the MAT into the MAF IAT wiring, or into any other available input. We just need to change a byte to get this value into the MAF IAT reading if you use another input. Either method is not a problem.

 

John this is what I am talking about:


My MAT is scaled every loop of my Tephra routine (so 100hzish).

 

ok so I flashed the new scaling for GM and drove my car:

1) coldstart was good, no funny idleing issues (obviously the idle was high because it was cold)
2) cold driving was good, felt like the car wanted to get up and go - like a 'golden flash' feeling!
3) initial warm driving was good as above
4) warm/hot idling was bad, wants to idle at 1100 then SLOWLY comes down to 900 (my MAF idle is 800)
5) FWY driving was ok, HOWEVER it felt like I was driving into a strong headwind - ie more TPS needed than usual to maintain
6) after the FWY drive, during normal road drive I had the notorious rev-hang problem, I am guessing its the same as the high-idle problem..

^ all off boost driving, John I am happy to shoot a log through to you, but it still feels like something is wrong...

Thoughts?

 

so just thinking about this again.

so if I have my GMMAP logging in kpa (from my own custom scaling) should that match the load regardless of rpm (so long as the VE for RPM is 100%??

ie 50kpa = 50load, 120kpa = 120load, 300kpa = 300load?

and if my logs indicate that's NOT the case I need to adjust my SD/MAPVE table right?

 

another dumb Q.

2byte airflow, is that now being calculated from MAP or is it still being calced from the MAF if its plugged in?

 

Access Port now has the option for speed density on the Subarus.

http://forums.nasioc.com/forums/show....php?t=1695346

From what I have seen, you can pull an accessport tune with ECUFlash. Might be interesting to see how they have done it.

 

ok John - Hopefully you can get back to me on the above Q's

I have another suggestion:

What if we change the VE+cal MAP table to a 16bit + 11bit adc and just do the 0x73* in the scaling. This means we get increased resolution from using both the ADC AND 16bit map (rather than 8bit)

Also I had a look in my ROM, there is a table "Air Density Compensation" @ 0x30f3 which looks like:



Do u think this needs updating? Given the IAT's are now 10-20 higher?

 

some more notes.

I took out my 4pt MAP averaging and the car felt a lot more clunky, not sure if it was just cold or what but the car felt a lot more fragile... also had some weird occasional gremlins that Ill see if I can find in the log...

I flashed back to a non-SD map, and studied my idles etc, and they are kinda the same as the SD... maybe I just don't notice it because I have 'confidence' in the stock system vs 'worryness' about the SD solution.

 

this is from the log with SD. Just free reving, very steady TPS of 18%:



Whats up with the zigzag STFT?

edit - lol they must be considering revoking my "Guru" status with all these posts - I just hope it's giving you some insight John!

 

Re the resolution of IAT and MAP, I think they are OK for now, but they are good ideas to consider improving them.

Re hot idle, it is freezing here so I don't have chance to test high temps, but we may have to rescale.

2 byte airflow is calculated even if the MAF is plugged in.

Load should match MAP in kPa when MAP VE is 1:1, RPM VE is 100% and "IAT" is 25C.

I made my VE table by using Evoscan's maptrace and having MAP vs RPM vs Load/MAP. I think getting your MAP VE right is crucial to getting this to work.

I had a quick look at your addresses and they looked correct.

 

I struggle with the scale of your graph re WBO2 and STFT, but my STFT is sawtooth (update speeds up above idle) as well, always looks like that IIRC?

I think it is just a case of getting that MAP VE table correct, and possible temp scaling.

 

ok - Obviously I am not at 25C, more like 45C - but anyways can you tell me how to tune the MAPVE?

Is there something I can do like looking at MAF-based logs of RPM vs Load vs GMMAP and then update the SD-based ROM somehow? so that the Loads come more inline?

Obviously this will be critical to tuning the tables and thus we need a sure-way procedure so that others can tune their's!

 

WB scale is on the RIGHT, all other items are on the left.

So as the STFT reduces (down and to the right) you can see the WB02 lean out (up and to the right)

Not sure why this is happening - ie is it evoecu design or sideaffect of SD?

 

I made my VE table by using Evoscan's maptrace and having MAP vs RPM vs Load/MAP (off the engine running MAF still). Because Evoscan only wants to do load vs RPM, you have to relabel Load to Loadold and then change MAP to Load to get it to display on the x axis, and you have to calculate the Load/MAP in the csv log before loading it into Evoscan. Then you can have a display of min/avg/max % VE and can incorporate these into the VE tables.

In your graph it looks like WB is richer than stoich but you have a positive STFT. This surely means your WBO2 and NBO2 don't agree?

On mine I just tuned closed loop off the STFT and the NBO2. Then did open loop checks on the WBO2.

I drove to work this morning, cold start, traffic, MAT from -1C to 22C and it all works fine. I'm sorry it isn't for you yet, but I know this can be good!