Well, today I tested my theory about the OBD EGR function test. It was wrong. I drove the car for 30-40 minutes on the freeway and then went and did two logs. The SAME exact thing happened as before. The AFR runs rich in the 3000-6000 rpm and then goes back to where it is supposed to be.

I did two logs then pulled over and flashed the ECU with the SAME flash. Then did two more logs. The AFR was perfect. Right where it should be.

So here is the pattern. It seems that after a period of driving the car on a new flash, the ECU notices something and makes the AFR rich in the 3000-6000 rpm region. When I flash the ECU anew, the old "memory" is erased and I get the exact same AFR that I wanted.

I also logged the fuel trims. The fuel trim low and high stayed @ ZERO, but the fuel trim medium gradually went up to 7 and stayed there. Is that ok?

 

Hi

I know its been asked before but i don't remember you saying what the differences between morning and afternoon ambient temps are? The temp and baro corrections are quite aggressive.

MB

 

I noticed that the funky AFR curve that I am getting looks very familiar. Then I looked @ old logs that I have done before I started messing with the boost.

If you recall, I posted about how a test pipe made my car run rich. Looking @ the log it is the same exact behavior as the current AFR. It is rich in the 3000-6000 rpm and then goes lean @ 6000+ rpm. I installed the test pipe on a Sunday, drove to work on Monday and then logged on Tuesday. This is the same pattern that is happening now.

When I posted that a test pipe made my car go rich everyone said that this is not possible. I thought I had an exhaust leak, but looking under the car those black marks are tar. I tried to wipe them out and they were sticky like tar is supposed to be.

I think this has been happening even before I messed with the boost. I am going to remove the JDMMAP and reflash to stock MAP tables. I want to see if this solves the problem. Worth a try.

 

This last test that I did was done @ the same temp/baro/humidity. I flashed the car and drove it on the freeway for ~ 35-40 minutes. Then did two logs and got the rich afr in 3000-6000 and then lean afr @ 6000+ rpm.

I pulled over and flashed the ECU with the SAME map. Did two new logs and the AFR in the 3000-6000 rpm leaned out and matched the 6000+ rpm.

All weather conditions remained the same in this latest test.

 

Is there anyone else with a JDM MAP sensor on an Evo 8 that is reporting this problem?

 

I removed the JDM MAP sensor and re-flashed the stock values into the Rom maps. I also installed my trusty GM 3 Bar Map sensor so I can read and display boost while I test and see if the JDM MAP sensor is the cause of this rich section in the AFR.

 

I know Bryan has the sensor installed. I wonder if he has witnessed this issue?

 

Today, I logged the car with the USDM MAP sensor and stock Rom maps. I drove the car around town yesterday to make sure that I followed a similar pattern to previous testing.

Well, it was NOT the JDM MAP sensor. The car did the SAME thing as before.

Next, I am going to go back to stock boost and stock boost maps. I am also going to use a clean xml file for base9 and my rom number. Basically the car goes back to the way it was stock with only the TBE on the car.

 

Don't want to sound like a broken recored, but have you tried disabling lean spool? I know its going to mess up your tuned value, but disabling lean spool would let you see if it is responsible for the funky behavior.

 

I already did that and mentioned it in this thread. I had a flash with lean spool disabled. The car exhibited the same behavior as with the lean spool enabled. It is NOT lean spool.

Yesterday, I sawpped back to a stock boost pill. I reverted to stock xml files and flashed the car with my stock flash. So there are no lean spool tables, 2 byte load, and JDM MAP patches in the flash.

I will log the car on my way back from work and see what happens.

 

Today I logged the car with the stock flash and no patches in the ECU. The car exhibited the SAME behavior as before. I have no idea what is going on here. I am at a loss.

Just to re-cap the problem: I flash the car and log immediately after and the AFR is where it should be. Then I drive around for a while (approximately 30-50 miles) and log again and the AFR goes rich by about 0.5-0.7 points in the 3000-6000 rpm and then jumps back where it should be from 6000 to cutoff. After I log, I re-flash the ECU on the spot with the SAME flash and log again and the rich spot is gone. So I know that the weather is NOT a factor

Now I am thinking that I have a defective ECU. So I am going to pick up an ECU from the junkyard and flash that with my rom and use it and see if it cures the problem. If it does not, then I know it is NOT the ECU.

In the meantime, I will work around this. I will flash the car with an overlean section where it becomes rich, then drive it around. Then I will do some log, given the previous pattern, that overlean section will become a bit richer and hit the target spot. That will take care of this problem for now.

 

Something does not like heat, the MAF maybe?

 

How did you disable lean spool?

 

Maybe you could post up your logs and a copy of your ROM. Might make it easier to find the problem.

MB

 

Actually, I was having this issue while tuning. I had it nailed down at 11.5 till around 5500. I made a couple of changes and logged in the exact same location not 20 min later and the AFR read 10.8. Confused, I went back around and logged again ... 11.5. the car was at normal temps the entire time, boost an load were nearly identical as well.

I have not read all the way back ... are you using Innovate for WBO2?