2009 Aussie RA - Tweaking and Tuning
Someone suggested I start a separate thread for ideas and general approach relating to the ECU tuning of my '09 RA. Seems like a fair idea...
It has the added advantage that this one thread will become my main stream-of-consciousness dumping ground for piddling questions and pointless tweaks... rather than the entire EvoM Ralliart forum section! :) At this point, I've got an Innovate LC-1 wideband O2 sensor on its way from the US. Also on its way is a boost pill (GST map at the ready!). However, before I get to flashing, logging and checking the higher boost GST map+pill, I want to learn a little more about EcuFlash/EvoScan tuning. I'm running a "lightly tweaked" version of the GST 93 oct stock pill V2 AUDM base map right now, and I'm performing some systematic on-road tests of different load/rpm points, logging in EvoScan. Currently, I'm checking low-rpm, high-load, high-gear behaviour. What I would like is a safe tune that handles stuff like drivers who, how shall I put it, "lack mechanical sympathy". ;) So I've got EvoScan set up with an alarm that sounds a tone if it sees knocksum of 2 or more. I'm getting out on the road at 60km/h in 5th gear (manual gear select) and going WOT. Disclaimer: I don't drive like this. Seriously, does anyone? We've all heard from granddad that you can't expect a car to climb a hill or tow a trailer in high gear / low rpm without knocking. And the RA 4B11 is indeed suffering from a bit of detonation. Good old granddad - right again. Once the rpm gets to 2300/2400 (boost rising to 0.9 bar), load hits around the 140 mark, and ECU serves up knocksum values to trip the alarm. Foot comes off the throttle asap. I've spent a bit of effort retarding timing in this area, but it feels like it needs at least 2 degrees taken out, which will in turn require smoothing around it , more testing, etc. My question is this: Do I spend time on this, the "idiot driver" section of the map when tuning? My obsessive/compulsive side says I must ELIMINATE KNOCK wherever I find it. This is repeatable, non-trivial knocksums in a specific area of the map. And this is an engine management system with precise control over pretty much everything. My realist side says the ECU doesn't have control over morons who lock the transmission into 5th at 2000rpm and proceed to put their right foot to the floor. So don't waste time trying to tune out driver idiocy. To those experienced in the art/science of map tuning, which way should I go on this? Skip the low-rpm, high-load stuff, or use it as a valuable learning experience? Oh, and I'll also accept suggestions like "stop dicking around until you can check your AFR, dummy". I'm telling myself that too. Happy to learn from people who have been here before! Rich |
For anyone (eg. Aussies) using ROMID 53610010 and logging, try the following for "Intake Air Temp" and see if it looks right to you...
Address: 80865b (EvoScan "2380865b", OP2.0 "0x80865b") Scaling: x-40 Units: Dec C Let me know what you think. And yes, this came from random byte logging and eyeballing the results. ;) Rich |
I personally don't see much point in spending much time tuning the extremely rarely hit load cells of your various maps. In contrast to your current test conditions, would you plan to rev your engine out to 7k at 5% throttle also!?! You can hit high load at low rpm without being so hard on your clutch to floor it in 5th. Find a hill and use 3rd. Tune the bulk of your map, then continue the trend to the extremes of the map. No need to try to simulate harsh conditions on the engine over and over again. During the tuning of these areas, you probably generate those conditions more than you ever would with your typical driving.
I can identify with the obsessive/compulsive thing though. Almost all summer I was driving myself crazy tuning in circles till I finally upgraded from my apparent leaky BPV that was causing me all kinds of grief. I'm looking forward to seeing this thread progress. {thumbup} |
1 Attachment(s)
Here's a file that could be useful to other 2009 AUS Ralliart owners.
The attached zip contains the EcuFlash definition for the 53610010 ROM image. Extract and place in the EcuFlash "rommetadata\mitsubishi\lancer" directory. Note that this definition file has the three "extra" fields needed if you wish to enable Mode23 logging. Beware of later versions of EcuFlash that may come with this... Rich |
That's great Rich, I really have to find my Openport 2 cable now. Keen to log a few runs with this.
|
^richard, win7 reports the zip file corrupted and wont open it.
some help please... |
Originally Posted by merlin.oz
(Post 8797130)
^richard, win7 reports the zip file corrupted and wont open it.
some help please... I'm not sure what happened. When I uploaded it, the file was a simple .zip. Now, when I download it again on my home XP machine, it's a zipped file within the .zip, called "53610010" - that needs unzipping twice. When I d/l it on another XP machine, it complains the format is junk. I reckon my own zipping tool is outdated - and maybe EvoM is doing something funny with zip uploads to boot! Seeing as the file text is pretty small, here it is in good ol' raw text... until I can sort out the awesome complexity of uploading a zipfile. 53610010.xml : Code:
<rom> |
Many thanks Richard, I will make a start on the Oz Ralliart.
And, glad you like the Evo Tuning Guide, because I have now started one just for the RALLIART. |
Originally Posted by merlin.oz
(Post 8797236)
Many thanks Richard, I will make a start on the Oz Ralliart.
And, glad you like the Evo Tuning Guide, because I have now started one just for the RALLIART. I'm dead keen to read an RA version of that guide... Rich |
A tune guide for the RA?
Subscribed!!! :) |
For anyone trying to tune the ignition timing map on the AUDM 2009 Ralliart, here's something to try:
The spark map seems to be referenced by the 2-byte Load value found at mode23 address 808776. Until prodded to look in a different area by Merlin, I was using 80876e when looking up load/rpm cells. This often resulted in less than accurate cross-referencing - when the car was nice and hot, the actual TimingAdv value being used by the ECU was often at a higher load point than indicated. I've been watching 808776 for a couple of days now, and it's bang-on accurate for the ignition timing map (so far). I'll continue to keep an eye on it, but this is now my "Load" variable of choice. Here's my current OP2.0 "logcfg.txt" file: Code:
; sample logging configuration file for openport 2.0 Rich |
So what would make the ECU's ignition timing suddenly deviate from the rpm/load table lookup... when knocksum is zero?
Or, alternatively, is it possible for the ECU to "hear" detonation (and react to it) without the logged knocksum moving off zero? I'm not talking on a single line here, where there can be a time delay between two "adjacent" values. I saw the following very strange deviation from the spark map... Code:
RPM Load Timing Knocksum So either the ECU sees the future (it's a miracle!), or the logged knocksum isn't the whole story. Is the ECU dialled in to the knock sensor in a far more subtle manner than just the general "rule of thumb" knocksum stuff? I find it unlikely that this behaviour is unrelated to knock sensing... seeing as the odd "red" area immediately precedes the only non-zero knocksum in over ten minutes of logs. Very interesting. Lots to learn... :) Rich |
Richard, it is almost certainly air temp modifying the advance.
I see we dont have the relevant table defined, so I shall try and find it - tephra probably has it already though for the EvoX. |
Originally Posted by merlin.oz
(Post 8820316)
Richard, it is almost certainly air temp modifying the advance.
I see we dont have the relevant table defined, so I shall try and find it - tephra probably has it already though for the EvoX. That's kind of what got me stumped - it seemed to be a sudden, measured but strictly momentary retardation. Good luck with the detective work! :) Rich |
Have you adjusted your max ignition timing map? Your log values follow the max timing table on my stock rom. I don't have the address for your rom id, but I bet thats you problem. You can't ask more of the car in the high octane map in that area without changing the limit.
|
All times are GMT -7. The time now is 10:02 PM. |
© 2024 MH Sub I, LLC dba Internet Brands