James' Civic Rebuild Thread + High Power ER Motor Theory

[James]

Made some awesome progress last night, worked out all of the analog data acquisition stuff, so with a bit more circuit reverse engineering I can work out where all the raw analog values are stored, which will help a lot with getting rid of fuel cut and things Might have fuel cut removed within the next 2 weeks actually. Shame I don't have a going car to test it on

[James]

Worked out the fuel map reading bit and some of the corrections yesterday. Also worked out today where the injector timer values get written so I now have enough information to scale the fueling for bigger injectors.

I need to get some writable memory chips so I can start testing out new code!

[James]

Just re-assembled my commented code into a form that can go back into the ECU for the first time and compared it to the file I read off the ECU. The good news is that they are exactly the same, this means I havn't made any errors anywhere while commenting and havn't changed the program at all while tidying it up Good news.

Now when my new chips get here I can get straight into testing...in my driveway

Anyone want to loan me a city.....

[Mike_NZ]

I'll be down in Palmy from the 9th - 15th of May for grad, so you can have a play with mine then.
This is some pretty exciting stuff! Good work.
If you can't wait that long... you could fix your car??

[James]

I'll be down in Palmy from the 9th - 15th of May for grad, so you can have a play with mine then.
This is some pretty exciting stuff! Good work.
If you can't wait that long... you could fix your car??

Long time no see man, hope its going well

Be mean to try out some stuff on your city then, but I think Ill leave you to play with your own tho

I heard you and Roz had a bit of a chat about my delayed repair, good things take time mate, and this is gonna be real good

[Hyperblade]

Just re-assembled my commented code into a form that can go back into the ECU for the first time and compared it to the file I read off the ECU. The good news is that they are exactly the same, this means I havn't made any errors anywhere while commenting and havn't changed the program at all while tidying it up Good news.

Now when my new chips get here I can get straight into testing...in my driveway

Anyone want to loan me a city.....

I have a couple of spare ecu's you could chip and send back down, but might be a waste if you haven't perfected it

[James]

Just re-assembled my commented code into a form that can go back into the ECU for the first time and compared it to the file I read off the ECU. The good news is that they are exactly the same, this means I havn't made any errors anywhere while commenting and havn't changed the program at all while tidying it up Good news.

Now when my new chips get here I can get straight into testing...in my driveway

Anyone want to loan me a city.....

I have a couple of spare ecu's you could chip and send back down, but might be a waste if you haven't perfected it

Yeah I have some spares too, but the changes are going to be pretty minor to start with so it would be really tedious sending ECU's around.

If you could get access to a chip burner I could just post modified Binaries up here for people to try, then youd just have to burn them and poop them in the ECU.

[James]

Found boost cut today

Its a pretty easy thing to move around too, I can move it up or down with ease, if I have the right thing that is

[James]

Double checked the value and its definately boost cut.

Map sensor B goes from 0-5 Volts, and is read by an 8 bit analog to digital converter. This means it can have 256 different values including zero. So 5/256 = 0.01953 Volts per step, and boost cut happens at like 3.22 Volts so 3.22/0.01953 = 164.864 and the value stored in the routine that I thought was boost cut is 165 so Im positive now Im right now. I should be getting the gear to burn some new chips next week so it will be just in time to try on mikes beast.

Also, now that I know how to stop injection, I can prtty easily implement some other cool features like a lower rev limit for launching, and an actual rev limiter!

[samol]

well done dude keep up the great work

[James]

Cheers man,

Crap sorry I just remembered I forgot to PM you back about stuff to get that turbo on. Did your mechanic manage it? Most of the stuff was in that list I sent you previously.

[city_cabriolet]

Rev limiter so you can't flog it when cold would be

[James]

Thats an interesting thought, I hadn't considered a different limit for cold running. I could also configure the PGMFI light as a shift lamp

Worked out quite a bit more last night, found all the stuff that works out where in the fuel maps to look, how it decides how many revs its doing and what its boosting etc. Found the code where it checks the inputs from the two map sensors and decides which one to use, and quite importantly I found where all the error codes get set, and how that is handled. It actually does a massive memory dump every time it throws an error code I guess for debugging.

Also found a routine that takes an altitude correction reading when you start the car up, uses Map A to measure atmospheric pressure. So if you are driving somewhere there is a large change in altitude turning your car off and back on part way might make it run better.

[James]

Took a lot of thinking but I finally worked out how it indexes the RPM side of the tuning tables, the RPM Bins it uses are:

500 600 700 800 999 1200 1399 1600 1999 2401 2799 3200 3599 3998 5000 5990

So there is also a column of manifold pressures for each of those RPM's

I can now change those rpms, so If anyone wants to rev their city up to like 7000rpm I can give you Tables for it

Now to work out the MAP side of things so I can get some tuning going for boost pressures higher than 15psi!

[Hyperblade]

Nice work dude, excellent detective work.