The Top DSM Community on the Web

For 1990-1999 Mitsubishi Eclipse, Eagle Talon, Plymouth Laser, and Galant VR-4 Owners. Log in to remove most ads.

Please Support Kiggly Racing
Please Support Morrison Fabrication

Evo 8 ECU In 2g Works!!

This site may earn a commission from merchant
affiliate links, including eBay, Amazon, and others.

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Anyone have a copy of the original pics on the first page?
 
I thought there were different ones. Mike, ever try that AC idea? Or want to tell me how to wire it so I can try it?
 
************Mods Please Update Post on First Page****************


*****Updated 3.28.2011*****
(Fixing Lost Images)

Below is a basic guide to installing an Evo 8 ECU in a 2G DSM. This is advantageous because it allows for the flashing/reprogramming of a stock Evo ECU (which can be found for around $150) via the EcuFlash program. This program has great functionality and capabilities, rivaling (or even surpassing) DSM Link in some areas.

Please note that this is just a guide, and I am not responsible for any damage caused by incorrect wiring, tuning, etc.



STEP ONE: WIRING


The first step is to go to Mach V forums and purchase an Evo 8 Harness Color Coded ECU Patch Harness - Evo 8 - ECU Tuning And Management. (Alternatively, you can cut/hack up your stock harness, but I'd recommend going with a plug-n-play harness if you can swing the $). Next, refer to the diagram attached for the wiring swaps that need to be made.





STEP TWO: ROM DEFINITIONS



Before you will be able to get a 2G to run with an EVO 8 ECU, you will need to find a ROM image with the engine immobilizer disabled. Join OpenEcu.org to obtain access to ROM downloads (Most EDM EVO 8 ROMs have the immobilizer disabled). OpenEcu.org is also where you can download (for free, legally) the EcuFlash tuning software, which allows load-based tuning and a great level of tuning flexibility. OpenEcu is a great place to learn more about tuning and the EcuFlash software.

But before you can do any flashing (including the initial immobilizer flash required to start the car for the first time), you will need to buy the reflashing cable (either the 1.3 or the 2.0 cable should work, found on OpenEcu.org).




STEP 3: INITIAL TUNING


For a 2G, the MAF size in EcuFlash will need to be changed to 260. You will also need to make the following Maf scaling changes for the EVO 8 Rom to work properly with a 2G (courtesy of Johnnyboy55):



You will also need to scale your injectors down to 400-450 (assuming you are running the stock injectors).




STEP 4: TUNING



Now the fun of tuning can begin. Log onto evolutionm.net or openecu.org to learn how to tune using EcuFlash.





Please PM me with any corrections/questions/comments.
 

Attachments

  • ecupinout1.jpg
    ecupinout1.jpg
    45.4 KB · Views: 256
Just figured I'd add this to the discussion:

I've been running the Evo 8 ECU on my daily GSX for a few weeks now. I started with the TephraMOD v7 ROM, and so far I'm loving it. The car drives better than it ever has (previously only had DSMLink V2 on this car). The huge increase in low-end torque and turbo spool after some tuning is amazing, it's night and day. It feels like a new car. i'm not having any of the aforementioned issues with just some quick adjustments. It starts up immediately at any temperature and idles perfectly. The driveability is beautiful.

I too had to adjust my low RPM knock multiplier up to get rid of part throttle knock. However, and more scary, is the fact that it seems the mid and high RPM tables are not set aggressive ENOUGH, which is very scary!

Up until last night, I didn't do a pull through the full RPM range, but doing shorter pulls, I noticed I was able to tune the car significantly more aggressively on the Evo ECU without seeing knock in boost on 91 octane.. i.e., 21 psi, 17* peak timing and 11.2 AFR which would make DSMLink show significant knock pretty much as soon as full boost was reached. So, last night I decided to have a little fun and did a 1st gear pull. About halfway through the pull, I heard very audible knock (several "tick" sounds) and the Evo ECU did pick up knock, but a only a slow flash. Unfortunately I did not log the pull. The point is, with the DSM ECU, the knock would NEVER get to the point that it was audible, even if I ran the same boost/timing etc numbers. I have a feeling the knock cal is not sensitive enough at high RPM. it's not a knock threshold vs. load thing since I have that set very conservatively. Just wanted to see if any others are seeing similar things.

I'm going to start tuning it at wastegate pressure, conservative timing, listening for knock, adjusting the knock cals until it JUST starts picking up knock when I know there isn't any present and then back it off from there. We use this same strategy setting up a knock sensor calibration on a standalone EMS system such as AEM/MoTeC etc.

Beau
 
Just figured I'd add this to the discussion:

I've been running the Evo 8 ECU on my daily GSX for a few weeks now. I started with the TephraMOD v7 ROM, and so far I'm loving it. The car drives better than it ever has (previously only had DSMLink V2 on this car). The huge increase in low-end torque and turbo spool after some tuning is amazing, it's night and day. It feels like a new car. i'm not having any of the aforementioned issues with just some quick adjustments. It starts up immediately at any temperature and idles perfectly. The driveability is beautiful.

I too had to adjust my low RPM knock multiplier up to get rid of part throttle knock. However, and more scary, is the fact that it seems the mid and high RPM tables are not set aggressive ENOUGH, which is very scary!

Up until last night, I didn't do a pull through the full RPM range, but doing shorter pulls, I noticed I was able to tune the car significantly more aggressively on the Evo ECU without seeing knock in boost on 91 octane.. i.e., 21 psi, 17* peak timing and 11.2 AFR which would make DSMLink show significant knock pretty much as soon as full boost was reached. So, last night I decided to have a little fun and did a 1st gear pull. About halfway through the pull, I heard very audible knock (several "tick" sounds) and the Evo ECU did pick up knock, but a only a slow flash. Unfortunately I did not log the pull. The point is, with the DSM ECU, the knock would NEVER get to the point that it was audible, even if I ran the same boost/timing etc numbers. I have a feeling the knock cal is not sensitive enough at high RPM. it's not a knock threshold vs. load thing since I have that set very conservatively. Just wanted to see if any others are seeing similar things.

I'm going to start tuning it at wastegate pressure, conservative timing, listening for knock, adjusting the knock cals until it JUST starts picking up knock when I know there isn't any present and then back it off from there. We use this same strategy setting up a knock sensor calibration on a standalone EMS system such as AEM/MoTeC etc.

Beau

Have you happen to have made any progress with this? I have my car back on the road again after its last 6 month sit down and with the minor changes I have made I have noticed a lot more knock in the mid throttle range.

Originally I had the car up and running with the 96 cam sensor but this last time the car was down I swapped it over to the 98-99 style sensor. Ever since then my ecu has picked up a considerable amount more knock (it went from between 3-5 counts to over 6 in the 30-50% throttle range right before the engine transitions from vacuum to boost.) AFR's are still decent they show a dip but they aren’t excessively lean or rich for part throttle driving from what I can tell, low 13's and still dropping when showing 2-3 psi positive pressure. I don’t get any audible knock that I can hear. Next payday if money is available I was considering picking up an EVO8 knock sensor and installing it just to see if there is any difference. I am going to do a fresh download of a tephra mod tonight and reinstall it over the weekend just in case I accidently changed something and didn’t realize it. Other than that the car drives awesome... Great power runs smooth no different from before swapping the CAM sensor.
 
Have you happen to have made any progress with this? I have my car back on the road again after its last 6 month sit down and with the minor changes I have made I have noticed a lot more knock in the mid throttle range.

Originally I had the car up and running with the 96 cam sensor but this last time the car was down I swapped it over to the 98-99 style sensor. Ever since then my ecu has picked up a considerable amount more knock (it went from between 3-5 counts to over 6 in the 30-50% throttle range right before the engine transitions from vacuum to boost.) AFR's are still decent they show a dip but they aren't excessively lean or rich for part throttle driving from what I can tell, low 13's and still dropping when showing 2-3 psi positive pressure. I don't get any audible knock that I can hear. Next payday if money is available I was considering picking up an EVO8 knock sensor and installing it just to see if there is any difference. I am going to do a fresh download of a tephra mod tonight and reinstall it over the weekend just in case I accidently changed something and didn't realize it. Other than that the car drives awesome... Great power runs smooth no different from before swapping the CAM sensor.

The knock low/mid/high sensitivity tables are based on RPM:

low = 0000-3800 RPM
mid = 3800-4800 RPM
high = 4800-7500 RPM

With the base settings, I was getting too much knock in the "low" RPM range, and not enough sensitivity (I was hearing detonation but the ECU wasn't detecting it as knock!) in the "mid" and "high" ranges. In these tables, increasing the value raises the knock threshold, meaning that it raises the knock voltage that the ECU considers to be knock.

I had to make the following changes (base --> new setting)

low: 13-->18
mid: 18-->14
high: 20-->16

After this, I rarely pick up any knock at lower RPM. Increasing the "low" to 20 gets rid of all low RPM knock, but since this car makes plenty of boost at 3800 and below, I want to make sure it's sensitive enough to pick up real knock at high boost. Also, more importantly, lowering the "mid" and "high" knock thresholds seemed to put the sensitivity right back where it should be. I'm getting knock right when I'm expecting it when I start turning the boost up on pump gas.

I adjusted these tables by logging the channels "knock base" (which is the ECU's current knock threshold based on your sensitivity settings, and the ECU's dynamic knock noise filtering), and "knock_adc_processed", which is the knock sensor's current reading. The units of both these channels are a "byte" value, converted from the analog knock sensor voltage, hence the "adc" part. Logging both of these channels along with RPM (and nothing else, to improve logging rate) should help you set your knock sensitivities correctly. You're looking for sudden spikes in "knock_adc_processed" that are not exceeding the "knockbase" value. These spikes are most likely real knock, and the "knockbase" may be too high for the ECU to consider the voltage spikes to be real knock.

With these adjustments, I feel the Evo knock sensor is unnecessary. Not to mention, I'm not even sure that it's any different.

Hope this helps!

Beau
 
Last edited by a moderator:
I had to make the following changes (base --> new setting)

low: 13-->18
mid: 18-->14
high: 20-->16

After this, I rarely pick up any knock at lower RPM. Increasing the "low" to 20 gets rid of all low RPM knock, but since this car makes plenty of boost at 3800 and below, I want to make sure it's sensitive enough to pick up real knock at high boost. Also, more importantly, lowering the "mid" and "high" knock thresholds seemed to put the sensitivity right back where it should be. I'm getting knock right when I'm expecting it when I start turning the boost up on pump gas.

I adjusted these tables by logging the channels "knock base" (which is the ECU's current knock threshold based on your sensitivity settings, and the ECU's dynamic knock noise filtering), and "knock_adc_processed", which is the knock sensor's current reading. The units of both these channels are a "byte" value, converted from the analog knock sensor voltage, hence the "adc" part. Logging both of these channels along with RPM (and nothing else, to improve logging rate) should help you set your knock sensitivities correctly. You're looking for sudden spikes in "knock_adc_processed" that are not exceeding the "knockbase" value. These spikes are most likely real knock, and the "knockbase" may be too high for the ECU to consider the voltage spikes to be real knock.

That advice should help quite a bit. I will try to see if I can make time this weekend to to do some logging and see what I can come up with. I will proceed with re-installing a fresh ROM so I can see if it helps straighten up my timing issue first. As the car sits I am showing a lot of timing advance that shouldn't be so I need to get that squared away first.

Thanks.

Mike.
 
That advice should help quite a bit. I will try to see if I can make time this weekend to to do some logging and see what I can come up with. I will proceed with re-installing a fresh ROM so I can see if it helps straighten up my timing issue first. As the car sits I am showing a lot of timing advance that shouldn't be so I need to get that squared away first.

Thanks.

Mike.

FYI, I replied to the other thread you created with a datalog of my car and some info.

Beau
 
FYI, I replied to the other thread you created with a datalog of my car and some info.

Beau

I got that, I just haven't had a chance to say thanks yet so.... Thank you very much. Your log is what is starting to make me think there was something screwy with my timing. I was seeing 43 degrees of positive timing in areas where the timing map was aiming for -4. But after taking the day off and doing some other things then looking at it again I think there may be something wrong with my EVOscan install. I was logging all the way up to 280 load even though I was only seeing 3 - 5 PSI of boost. normally I am only in the 100 - 120 load range at that boost.

Edit... Guess I should have posted this in the other thread...:coy:... Oops.
 
Just as a general FYI...I'm now running Speed Density (thanks to Beau!), and I love it! Still in the tuning phase, but it appears to be working quite well so far. So, if any of you were considering going SD (not sure if anyone on here besides Beau and I have done this yet on a 2G with the Evo ECU), it can be done and works well.
 
I have been running SD on my evo ecu for the past year and a half or so. I'm actually running SD 2.0 with evo live map, i like it BUT there is a consistent missfire/stutter at a certain rpm which i can't get rid off and it's really annoying.
 
I have been running SD for a year now, and ive put thousands of miles on teh car (its made 3 or 4 trips to and from arkansas to ny) and ive never had an issue with a missfire/studder at ANY rpm's..if your curious i can try and send one of my tunes and logs so you can see and try and compare.

Ive set up the roms and configured them myself so i could have done something that most others havent. Im running Tephrav7 SD non live rom with a GM IAT in the throttle body elbow and an Omni 4 bar map. I also have ALL emission stuff deleted and am running a wideband for the front o2 and no rear o2/cat/egr.

Thinking about it quickly, the only thing ive changed that i couldnt find to many others who have done it too is the ignition coil charge times or something like that.

and using the same tune/base rom ive had no studder issues in any of my setups. (Stock 7 bolt with HX35, evoIII in/cams, 1g green top CAS, pte 880cc inj; and Stock 6 bolt with 14b, s16g, and b16g with 880cc inj) only one time did i have a studder issue and it was because my plugs were fouled from tuning and changing them took it away immediately
 
I had no idea people were running SD on these...nice!

I really haven't driven it too much yet...i do notice some weird behavior around 2k or so when im just cruising at a constant speed in closed loop, but I'm hoping that will go away once the tune is dialed in. During partial acceleration or WOT i haven't noticed any misfires around that RPM range though.

What ROM were you guys running?
 
The "phenem" and "l2r99gst" SD ROMs use the (in my opinion) annoying separate 2D VE tables for RPM and load, but include several "jitter" fixes for the known light-load bug that occurrs around 2-3000 RPM. The newer "SD 2.0" ROM has a more intuitive 3D VE map, but from what I've seen so far there are no successful jitter fixes for that ROM. Unfortunate, since I think it's a better implementation.

Even more unfortunate, there aren't any very well implemented ECUFlash SD ROMs at all, even the SD 2.0. For instance, engine displacement is a crucial part of the SD airflow calculation, and it's just assumed to be 2.0L. If you have a larger displacement engine, the VE values will have to be incorrectly skewed to get fueling where it needs to be, which in my opinion is silly. I only tune SD systems by adjusting the VE values to match the engine characteristics, so that the target AFR table matches actual measured AFR. This makes it much easier to make changes to AFR later down the road, or to track changes in engine efficiency as components on the engine are changed.

It seems that a lot of the ROM development for the Evo 8/9 and DSM ECUFlash ROMs has been dropped and more of the development has been focused on the Evo X recently.. Oh well.

Beau
 
I have been running the evo dwell times and havent had any misfire issues yet, did you notice a difference switching over? I have a map setup with the dwell setting xhackerx posted up for the coil on plug setup that im going to try later today since thats whats currently on the car.
 
Anyone running 96531706 and using the AEM UEGO and have the rom setup with the Wideband patch. I have tried to impliment this a few times with zero luck. I always end up with an 02 sensor code and alt code. The patch im referring to is here:
how-to: control closed loop AFR with your wideband O2 system - Page 16 - evolutionm.net

Im pretty sure im just missing something somewhere and would like to do a compare with some one that has this setup and working and see WTH im looking past.

UEGO is installed and I successfully log it in Evoscan with it wired to the rear O2 input.

I thought initially the MUTT might be the issue but I had the info there correct. Its more of an annoyance i cant figure out whats wrong than anything. Thanks

You can email rom to g23laserrs @t yahoo dot com.
 
Anyone running 96531706 and using the AEM UEGO and have the rom setup with the Wideband patch. I have tried to impliment this a few times with zero luck. I always end up with an 02 sensor code and alt code. The patch im referring to is here:
how-to: control closed loop AFR with your wideband O2 system - Page 16 - evolutionm.net

Im pretty sure im just missing something somewhere and would like to do a compare with some one that has this setup and working and see WTH im looking past.

UEGO is installed and I successfully log it in Evoscan with it wired to the rear O2 input.

I thought initially the MUTT might be the issue but I had the info there correct. Its more of an annoyance i cant figure out whats wrong than anything. Thanks

You can email rom to g23laserrs @t yahoo dot com.

Thats the patch im running with the uego and it runs perfect for me.....ill email you my rom but is speed density so just use it to copy the related data
 
Support Vendors who Support the DSM Community
Boosted Fabrication ECM Tuning ExtremePSI Fuel Injector Clinic Innovation Products Jacks Transmissions JNZ Tuning Kiggly Racing Morrison Fabrications MyMitsubishiStore.com RixRacing RockAuto RTM Racing STM Tuned

Latest posts

Build Thread Updates

Vendor Updates

Latest Classifieds

Back
Top