More advanced timing #6

The last round didn't work either, I had knock in the range of load 0.75-0.8 between 2700-3000 rpm,

So I'm going to increase timing below that range only since I've reached the limit on that.

I increase the range below by 1.0 x 2:


Then the range below by 1.0 X 1"


And the range below by 1.0 X 3:


And reduce the section below 0.2 x 2 to smoothen the map:

More advanced timing #5

My last changes din work.. I still had more knock :-(


I decided to start over and use 20080916_advance.hex and instead of increasing it by 1.0 x 1, increase it by 0.2 x 2:


This ROM is named 20080920_advance.hex

More advanced timing #4

Still no knock so I decided to increase the timing by 1.0 x 1. This time I had knock. Here's the result from mickeyd's spreadsheet:


Here's the change I made:


And here's how the 3d graph looks like:


I'm going to name this rom: "20080920_advance.hex"

More advanced timing #3

Still no knock with my last revision. I've decided to increase it further 0.2 x 2 for all cells:


Compare to Stage2 MY05 STi from OS ECU ROM, mine is still not as hard core:

Tip-in - Possible reason why running lean causes problem at low revs

Here's a possible cause to the problem I had when going lean:

Funny, my car has the same tip in stumble. I've gotten so, I blip when I let the clutch out to cover it up. What you seem to be saying is having it butt rich right off idle gets rid of the tip in sag. I haven't tried logging just that initial startup tip in. Maybe I should. I know I've seen recommendations to have the MAF set to pull 5% fuel at idle. Which is supposed to make tip in smoother.

Read the thread here.

Good getting started summary to tuning:

Disclaimer first: I'm just getting into tuning myself. Don't trust this until it's been verified by someone with more experience. I think it would be useful for the FAQ though (if/when it's correct).

First, make sure the ECU knows what it's attached to:

MAF Sensor Scaling - It's a good idea to get this dialed in before you start. And definitely revise this if you change your intake anywhere near the MAF sensor. It tells the ECU how to interpret the MAF sensor voltage to know much air is going into the engine. The learning view tool and the RomRaider logger's MAF scaling tab will help dial it in. There's also a spreadsheet in a sticky thread in the tuning forum.

Injector Latency and Injector Flow Scaling - change these if you change your fuel injectors. They tell the ECU how long it takes for an injector to fully open, and how much fuel the injector flows while it's open. Together they allow the ECU to squirt just the right amount of fuel. RomRaider's logger has an injector scaling tab which is helpful. You can also tune these by watching the fuel trims shown in the learning view tool.

Then, the "real" tuning. The main things are fueling, boost, and timing. All of these will induce knock if you ask for too much power, and they interact, so after making changes you should log to ensure that you haven't overdone it. (Knock is a leading cause of death for high-performance engines.)

Adjust the Primary Open Loop Fueling table to get the right AFRs. You'll need a wideband O2 sensor to verify your changes here, since the stock O2 sensor is not accurate under boost. You may want/need to adjust the MAF and/or injector tables as well if you car isn't hitting the desired AFRs.

Enter your boost targets in the Target Boost table. You will likely need to adjust the Initial Wastegate Duty and Max Wastegate Duty tables to get the engine to hit your targets without overboosting. Also consider the various Turbo Dynamics tables to help hit your boost targets.

Use the Base Timing and Timing Advance (Maximum) tables to adjust ignition timing for best power, without detonation.

Opinions vary about the best order in which to adjust fuel, boost, and timing. There isn't necessarily a best way to approach this, since the adjustments need to be iterative and you need to experiment with different tradeoffs to determine the which compromise across all three makes the most power without inducing detonation (more boost with less timing / leaner AFRs with less boost / more timing with richer AFRs, etc).

Miscellaneous:

Some people extend the knock detection range all the way up to redline (Feedback Correction Range, Rough Correction Range). Some say that the sensor isn't very accurate outside the stock range, I can't say who is right.

Some people zero out the Closed Loop / Open Loop transition times (CL to OL Delay and related tables).

Many people raise their redline a bit (Rev Limit) and raise the Rev Limit Fuel Resume to avoid losing power when bouncing off the redline. Some people also raise the Speed Limiting tables.

Anyone care to confim / dispute / add to the above?

Read the full thread here:

More advanced timing #2

With my last timing revisions, I didn't get any knock. I've decided to increase it a bit more:


I'm naming this rom "20080915_advance.hex"

More advanced timing #1

No knock since my last flash (20080910_advance.hex). As such, I've increased another section of my map, increasing it by 0.2 x 2:


How to know if the changes are making any differences? Well, one way is to look at Airboy's acceleration spreadsheet. I'm going to have a go at that.

Good advice on tuning timing

I found this post on romraider - I think it is a very good summary:

3D map view isnt one of romraider's strong points. ECUFlash is much easier to see the maps in 3d and smooth them out.

You need to concentrate on getting the maps smooth and logicalThere should be a progressive increase in timing with revs to allow for the extra time needed for the flame front to travel and burn the mixture. Therefore each increase in rpm should have an increase in timing. The point at 4400 on your maps actually goes down. Maybe that was to fix a det issue but the correct way to do it would to be smooth the surrounding cells to still allow this logical increase.

The timing should also logically decrease with increased load.

The post is here:

Advance Timing at Closed Loop Target 14.47

Since my map at 14.47 is very smooth with no knock (20080726_timing.hex) What I am going now is to advance the timing for this map keeping closed loop target at 14.47. This map is good just that pickup is a little slow. Let's see if I can improve it further.

After flashing it today, no knock but it's abit jerky. I'm going to drive around for a few more days and see if it improves.

Going Lean #5

Here's my final timing map for going lean:


It's everything that I mentioned in Going Lean#4 except for Step8.
:-) No knock for now. Except in the higher RPMs.

The map name is: 20080823_lean.hex

Even though this map has no knock and is quite smooth, the only problem is that sometimes when the rpm drops to idle, it drops to about 600rpm and sometimes the engine dies. :-(

I'm not sure how to fix this, though its not a major problem.

Bridgestone RE070 finally!

I went for my 5th year vehicle inspection on Saturday (had to fix my handbrake and remove stickers before I could pass). The mechanic who fixed my car saw that my front tires were gone - the steel linings were even showing.

So I went to Leong Seng Tyres and changed to RE070 (luckily they had stock!). I'm so happy coz I have been wanting to change to these tires for so long and finally I had them. They cost $290 per piece.

My first hand impression after driving about 100KM on it, they are much much better than Yokohama Advan AD07 and Falken Azenis RT215 and RT615. My ranking:

1. Bridegestone RE070
2. Yokohama Advan AD07
3. Falken Azenis RT215
4. Falken Azenis RT615

I've yet to take the RE070 to the track but I've no doubt it will outperform. :-)