CE Lancer ECU tuning

OK so is it possible to flash the CM5A ROM to the 93 ECU or is there missing chips/inputs to allow that to work correctly? or is the loom just set up differently?
I am only just starting to dip my toes into the work of reflashing ancient technology (I dabble in the dark arts of PC building/coding) and am moving into the fun side of these "black boxes" (PUN INTENDED!!) haha so trying to work out what's what before I start finalising this turbo setup I have mostly assembled!

So expect a heap of 'stupid' questions (as I believe the only stupid question is the one you don't ask!

Thank you guys JimmiB

P.S. if anyone happens to be up in the Newcastle/Maitland region and able to sit down over a coffee/beer and help me get my noggin around all this I would super appreciate it!!
 
Anyone have additional information about swapping a plastic case ECU into a metal case ECU car? Pin outs the same?
 
Anyone have additional information about swapping a plastic case ECU into a metal case ECU car? Pin outs the same?
All the same. As long as it's for the same engine you be golden, transmission makes no diff.
 
Will also need the immobiliser disabled unless you swap over the key reader and immobiliser with the same plastic ecu case
 
OK so is it possible to flash the CM5A ROM to the 93 ECU or is there missing chips/inputs to allow that to work correctly? or is the loom just set up differently?
I am only just starting to dip my toes into the work of reflashing ancient technology (I dabble in the dark arts of PC building/coding) and am moving into the fun side of these "black boxes" (PUN INTENDED!!) haha so trying to work out what's what before I start finalising this turbo setup I have mostly assembled!

So expect a heap of 'stupid' questions (as I believe the only stupid question is the one you don't ask!

Thank you guys JimmiB

P.S. if anyone happens to be up in the Newcastle/Maitland region and able to sit down over a coffee/beer and help me get my noggin around all this I would super appreciate it!!

Yeap, can flash a CM5A ROM to a 4g93 without any issues. I believe that the aircon and fuel pump loom pins are swapped though, so if you have a cm5a loom, it will not run. This isnt an issue when flashing a cm5a to a 4g93 ecu, but it is an issue if you flash a 4g93 to a cm5a ecu (think along the lines of cm5a front cut swap).
 
What is the difference between the 2 ecu’s besides the maps? I am thinking maybe more memory or something to allow extending maps to allow for boost? Or are they all the same
 
Not really a lot different, you are on the right track for sure. That generation of Mitsubishi ECUs are only of a couple of types of ECU such as in the Galant, Legnum, and a couple others. The more basic cars had less development, and less information saved on to the ECU chips.

A quick and nasty explanation (without me double checking specifics) is the 4g93 had one fuel and one timing map, of 11 cells by 11 cells. The CM5A used a little more of its stored memory, and was able to do 15 by 15, but with two maps.
The Evo 5 and 6 was able to do 20 by 20 cells, in two maps (think along the lines of a map for petrol and a map for E85). In addition, it also had a low octane and a high octane map for each of the two fuels, in the event of knock.

I dont really know a lot more in depth about this, such as why the 4g93 was so basic, because to find this information out you need to power up the ECU coding and attempt to follow what hexadecimal addresses get called during certain functions (e.g, what happens in limp mode, like what timing map/cells it runs if the limp mode activates). This is a very specialised activity, and is actually really old school - nowadays modern standalong ECU's do this with IF statements, and written in C, C#/C++, or PHP or another, quicker and easy computer language.
 
Oh I miss the old days oh hex coding to “break” my games haha used to use an old hex coder to alter games god mode ammo health whatever haha some came built in but most of my games weren’t common so I had to work out what hex code did what...... now I need to go back and relearn it to do the same thing with my ecu .... sweet haha

Ok so that raises a few other questions sure maps were different between ecu’s as would pin outs but is the amounts of memory be it RAM or ROM is it all even or is more maps etc more mem better processor blah blah etc
 
Oh I miss the old days oh hex coding to “break” my games haha used to use an old hex coder to alter games god mode ammo health whatever haha some came built in but most of my games weren’t common so I had to work out what hex code did what...... now I need to go back and relearn it to do the same thing with my ecu .... sweet haha

Ok so that raises a few other questions sure maps were different between ecu’s as would pin outs but is the amounts of memory be it RAM or ROM is it all even or is more maps etc more mem better processor blah blah etc

Oh sweet so yeah thats more or less the exact same sort of thing lol.
You decipher the hex coding, change/edit the value, and reflash (rewrite the hex) back to the ecu, and the ECU will read and play the code.

From memory, its all the same. The 4g93 maps are simple smaller, so there is more free space available. Technically it would be possible to write the Evo algorithms to the CE ECU, such as launch control, but you would need to know exactly what memory addresses are available and unused and never called upon, and that again is the tracing out of the Hex process.
 
so another question then, what sorta Power figures can you pull from a stock 4G93 Block N/A with a decent Re-map?
stock its 86KW at the crank... which if you subtract 15% the common drivetrain loss you get 73KW at the wheels (97hp) and considering ALL of these cars are far from new its probably closer to 65KW at the wheels... with a throttle body, intake, exhaust, extractors & a basic tune to richen the mixture and maybe timing slightly you probably should see near 80KW at the wheels or over 100whp... but with a DOHC head swap and some bigger cams yes you could see up to and over 100KW at the wheels, but by the time youve done that you might aswell go turbo or engine swap
 
you mean lean??? make sure you still have your MAF connected and working properly

It wouldnt be running at either end of the spectrum - the ECU for both 4g15 and 4g93 are smart enough to use the o2 feedback to bring it back in line with the OEM mapping.
 
I also forgot to ask which tactix cable is the one I need I can’t quite workout 1.3 or 2.0? And if 1.3 which one there is a 1.3D, 1.3R and 1.3U and they are all different
 
From what i know about the ecu hardware it is pretty similar between models, same CPU and board, mostly the same looking. the only known issues are the more basic cars depending on the model and market use a few less chips to handle the inputs and outputs, such as the boost control solenoid not found on n/a cars. EGR not fitted to the audm engines, etc

by running the ceddymod evo 5/6 roms i havnt noticed any funky business except the idle dipping to around 650rpm when you clutch in or rev it up a decent amount and then snap the throttle shut. it doesnt nicely return to the 800 rpm as set like the stock rom does. and also the air con not working. needs further investigation as the led in the button is blown and the last few times i had it on it made a nice grinding sound.. haha! donki does your ac work?

the tactrix openport 2.0 is required for the CE series for flashing. you can use a 1.3 cable for logging only apparently, but i havnt tested this.

JimmyB if you havnt already grab the latest ceddymod ralliart evo rom & XML i posted or from donki.
if you compare to the ce 4g93 one you will find some things are in the same locations as a base starting point.
I have managed to find a few more tables by inputting the same addresses from evo rom into my XML file for ecuflash and some do line up perfectly, others half the table lines, up and the other half is all wrong, i managed to add a few boost control tables to the xml by digging though other evo roms. if you have the skills to sus out the hex coding it would be a great way to get started.

i have also found the pajero io 4g93, cm5a 4g93t also share locations, some are defined a little differently aswell as the older non ralliart evo roms from 5-6. really intersting "boost enhancement" tables and many more.

if you need someone to do testing im very experienced in flashing ,logging, making the changes and keen to help find some more tables! :)
running on ceddymod i have everything pretty much perfect,

the way i see moving forward to correct any issues im having is to decide between these options
A: find a few more tables in the evo 5/6 ceddymod rom to ensure its as good as possible
or
B: transplant a few features from the evo roms into the stock 4g93 rom.

i think option A would be the best, but doing both and combining all the newly found info you could choose to use either rom and have the car run the exact same!
 
From what i know about the ecu hardware it is pretty similar between models, same CPU and board, mostly the same looking. the only known issues are the more basic cars depending on the model and market use a few less chips to handle the inputs and outputs, such as the boost control solenoid not found on n/a cars. EGR not fitted to the audm engines, etc

by running the ceddymod evo 5/6 roms i havnt noticed any funky business except the idle dipping to around 650rpm when you clutch in or rev it up a decent amount and then snap the throttle shut. it doesnt nicely return to the 800 rpm as set like the stock rom does. and also the air con not working. needs further investigation as the led in the button is blown and the last few times i had it on it made a nice grinding sound.. haha! donki does your ac work?

the tactrix openport 2.0 is required for the CE series for flashing. you can use a 1.3 cable for logging only apparently, but i havnt tested this.

JimmyB if you havnt already grab the latest ceddymod ralliart evo rom & XML i posted or from donki.
if you compare to the ce 4g93 one you will find some things are in the same locations as a base starting point.
I have managed to find a few more tables by inputting the same addresses from evo rom into my XML file for ecuflash and some do line up perfectly, others half the table lines, up and the other half is all wrong, i managed to add a few boost control tables to the xml by digging though other evo roms. if you have the skills to sus out the hex coding it would be a great way to get started.

i have also found the pajero io 4g93, cm5a 4g93t also share locations, some are defined a little differently aswell as the older non ralliart evo roms from 5-6. really intersting "boost enhancement" tables and many more.

if you need someone to do testing im very experienced in flashing ,logging, making the changes and keen to help find some more tables! :)
running on ceddymod i have everything pretty much perfect,

the way i see moving forward to correct any issues im having is to decide between these options
A: find a few more tables in the evo 5/6 ceddymod rom to ensure its as good as possible
or
B: transplant a few features from the evo roms into the stock 4g93 rom.

i think option A would be the best, but doing both and combining all the newly found info you could choose to use either rom and have the car run the exact same!

I have almost zero issues with runnings ceddymods now. I changed my TB (and IACV) to an Evo one, as from memory ceddymods and the OEM g93 ROM use vastly different idle steps. Changing the steps to what actually suits the engine, instead of what Mitsubishi inputted for either 4g93 or Evo makes a world of difference. Its well worth the effort of doing some datalogging on that.
Air con also works flawlessly (though ive burnt out a temp motor. I also run the auto climate control setup, but ive run that since the car was on the old 4g15 without any tune).

Only real issue i have is tip-in enrichment/accel enrichment. There is only a single field for it for ceddymods, so i cant adjust differently for 98 or e85. On 98, its more or less stock, and it runs fairly well. On E85, theres a fractional longer delay before the car picks up again, with the wideband gauge verifying that the issue is fuel related. Not very often i need an instantly dumping of fuel for any reason, aside from showing off how quickly the car revs but thats stupid haha.
 
another thing to add:

Donki, you should get a GM 3 bar map sensor and wire it in to pin 75, sensor ground and ecu 5V, i took this from the tps and mounted it on the firewall. you can then log boost pressure in evoscan! i used a chinese knockoff sensor and working well. dirt cheap on ebay and makes tuning faster even if sticking with a MAF rom.

Mirage is having some paint work done and then il be back to finishing my Speed Density ROM, copy everything else over to the MAF rom, and drive the lancer and mirage and note the differences. i am running the SD in open loop so far, but will test closed loop next, i found in earlier testing when the VE table is not setup well enough it messes with the fuel trims and ends up throwing the whole tune completely off! closed loop seems to be more than just fuel trim. it is also idle control stability and learning..

SD rom is only having a couple issues.
1: As intake temp increases (sensor reading not actual temp being ingested - heaksoaked sensor) the car leans out a fair bit. usually hovering around 14-16afr normally, when sensor is hot it likes to idle and cruise at 16-18ish. cool the sensor down and it goes back to normal for abit.
I need to play with the MAF air temp/denisty/baro table i think. the Temp correction table can make it leaner, but will not make it richer from my testing..
The other option i will try is fool the temp sensor/fudge the IAT voltage table to find a happy medium. slightly rich sometimes, slightly lean others, tune for temp like a carby. lol!
this is with the GM IAT and no maf connected.
2: Due to idle dipping at throttle off normally with both MAF and SD roms, its worsened by the temp issue going very lean when dropping to around 600rpm, engine looses torque due to very lean afr in the 20s iv seen, open loop will not re-learn idle in this scenario. so you end up stuck at 700 with a very lean AFR when normally around 1000 as iv set. i think closed loop may correct it once the ve table is perfected and temp issue is rectified but may be wrong.

a few more idle tables found or better descriptions of the current ones will help these issues a huge amount!
 
Back
Top