Looney Tuning
03-30-2009, 08:32 PM
This Evo X belongs to Gear Grinder. It is the first Looney Tuned Evo X. The car had the following modifications before I tuned it:
Downpipe
HFC
Exhaust
AEM Intake
The car was making pretty decent power mostly because of the AEM intake. Most know by know that the AEM intake produces a lot of its power by leaning out the horribly rich stock AFR. The size of the MAF intake pipe on the AEM changes the airflow that the MAF senses. This lowers the load that the car hits. The lower load cells are also leaner than the higher load cells. That makes the car run by up to one full AFR point leaner. A leaner fuel mixture produces more power than a richer one. The same logic applies to the timing maps. The lower load cells have more timing and that makes the car make even more power. Despite the leaner AFR of the AEM intake Evo Xs still run rich and the timing is too high that the car knocks and so the car requires a tune. But the AEM intake really helps give your car semi-tune.
GG and I logged this car at 4 different sessions. We logged it with the stock tune, after the first flash, the second flash, and the third flash. There will be a fourth flash, but the last flash is as close as it gets to what we wanted out of the car. So it was time to do a write-up on it.
My biggest concern with this car was triggering a CEL. I have read many many instances of tuners triggering CEL on the Evo X. I am happy to say that the CEL was not triggered once. Tuning load based boost w/o the ability to log load error and WGDCC can be very challenging. I used to tune w/o logging these two essential parameters for a long time, but when mrfred discovered a way to log these tuning load based boost became very easy. Now, I have to go back to using an excel spread sheet to correlate the load/boost numbers to the WGDC number so as not to trigger an overboost CEL. I cannot wait for Evoscan for the Evo X to come out. Logging Load Error and WGDCC makes tuning ECU boost so much easier. W/O them tuning is more tedious, but doable.
The other issue that we had with the Evo X is logging boost. While the MAP sensor on the Evo X is a 3 BAR sensor, The ECU limits it output to 22.2 psi. I do not know why Mitsu does this crap. What is needed is for the disassembly experts to come up with a patch that will remove that cap. In the meantime, we will have to log boost through an external MAP sensor. Luckily PCMScan (the logger that opensource tuners use) allows auxiliary input for five Innovate channels. That allowed us to use a GM 3-bar MAP sensor and log it through PCMScan.
Enough of the talk and on to the charts. Here is the way the before and after boost turned out:
http://www.sr20deracing.com/EVO/Danny/boost_b4_after.gif
First, the tuned boost is higher both at peak and by redline. The average boost increase at peak is 3.08 psi and the average increase at redline is 3.45. Second, the car also hit higher load cells than before, indicating more power being made. Whereas the stock tune never allowed the car to hit the 260 load cells, the third flash hit the 260 load cell at 3000 and 3500 rpm. From there on the logs never tapered below the 200 load cell. By contrast the stock log tapers below 200 starting at 6000 rpm.
Despite the AEM intake, the AFR was on the rich side of things. It was not in the 9.xx:1 like stock, but in the 10.xx:1. I wanted it to be in the 11.xx:1. So I had the fuel map redone and this is the way it came out:
http://www.sr20deracing.com/EVO/Danny/afr_b4_after.gif
The stock AFR was nice and lean all the way to 4000 rpm. Thereafter, the AFR fell into the 10:1 area and tapered all the way to 10.35:1 by redline. While this is an improvement over a stock tune, it is still not enough. The tuned AFR stayed in the 11:1 region all the way to 6500 rpm where it dipped into the high 10:1 AFR. Note the 10.74:1 AFR dip at peak boost. I am still working on this one.
For the power charts, I did something different this time. I averaged the results from the last two runs that GG and I did and then exported them to an excel spread sheet. Then I exported the best stock run that we did into the same excel spread sheet. Averaging the results should provide for a more consistent result from a road dyno. Please note that these power runs were done in 4th gear and not in 3rd gear which tends to produce slightly higher power numbers:
Here is the power that GG’s car put down:
http://www.sr20deracing.com/EVO/Danny/dyno_before_after.gif
First, most of the gains were concentrated above 5000 rpm. In the 6000-6500 rpm range, for example, the car gained an impressive 59 hp and 49 ft-lb over stock. If you look at the entire 5000 to 7000 area under the curve you will see an average hp gain of 37.6 hp and 32 ft-lb of torque. The rest of the gains/losses are noted on the graph.
I am not done tinkering with this car. I did create a ninja mivec intake map for it, but I have not messed around with the mivec exhaust map. I am going to have a sit down with Mike Kojima and see what we can come up with for the mivec exhaust map. I have a feeling that threre is more power to be had from a properly designed mivec exhaust map.
So, stay tuned, there is more to come.Â* Â*
Downpipe
HFC
Exhaust
AEM Intake
The car was making pretty decent power mostly because of the AEM intake. Most know by know that the AEM intake produces a lot of its power by leaning out the horribly rich stock AFR. The size of the MAF intake pipe on the AEM changes the airflow that the MAF senses. This lowers the load that the car hits. The lower load cells are also leaner than the higher load cells. That makes the car run by up to one full AFR point leaner. A leaner fuel mixture produces more power than a richer one. The same logic applies to the timing maps. The lower load cells have more timing and that makes the car make even more power. Despite the leaner AFR of the AEM intake Evo Xs still run rich and the timing is too high that the car knocks and so the car requires a tune. But the AEM intake really helps give your car semi-tune.
GG and I logged this car at 4 different sessions. We logged it with the stock tune, after the first flash, the second flash, and the third flash. There will be a fourth flash, but the last flash is as close as it gets to what we wanted out of the car. So it was time to do a write-up on it.
My biggest concern with this car was triggering a CEL. I have read many many instances of tuners triggering CEL on the Evo X. I am happy to say that the CEL was not triggered once. Tuning load based boost w/o the ability to log load error and WGDCC can be very challenging. I used to tune w/o logging these two essential parameters for a long time, but when mrfred discovered a way to log these tuning load based boost became very easy. Now, I have to go back to using an excel spread sheet to correlate the load/boost numbers to the WGDC number so as not to trigger an overboost CEL. I cannot wait for Evoscan for the Evo X to come out. Logging Load Error and WGDCC makes tuning ECU boost so much easier. W/O them tuning is more tedious, but doable.
The other issue that we had with the Evo X is logging boost. While the MAP sensor on the Evo X is a 3 BAR sensor, The ECU limits it output to 22.2 psi. I do not know why Mitsu does this crap. What is needed is for the disassembly experts to come up with a patch that will remove that cap. In the meantime, we will have to log boost through an external MAP sensor. Luckily PCMScan (the logger that opensource tuners use) allows auxiliary input for five Innovate channels. That allowed us to use a GM 3-bar MAP sensor and log it through PCMScan.
Enough of the talk and on to the charts. Here is the way the before and after boost turned out:
http://www.sr20deracing.com/EVO/Danny/boost_b4_after.gif
First, the tuned boost is higher both at peak and by redline. The average boost increase at peak is 3.08 psi and the average increase at redline is 3.45. Second, the car also hit higher load cells than before, indicating more power being made. Whereas the stock tune never allowed the car to hit the 260 load cells, the third flash hit the 260 load cell at 3000 and 3500 rpm. From there on the logs never tapered below the 200 load cell. By contrast the stock log tapers below 200 starting at 6000 rpm.
Despite the AEM intake, the AFR was on the rich side of things. It was not in the 9.xx:1 like stock, but in the 10.xx:1. I wanted it to be in the 11.xx:1. So I had the fuel map redone and this is the way it came out:
http://www.sr20deracing.com/EVO/Danny/afr_b4_after.gif
The stock AFR was nice and lean all the way to 4000 rpm. Thereafter, the AFR fell into the 10:1 area and tapered all the way to 10.35:1 by redline. While this is an improvement over a stock tune, it is still not enough. The tuned AFR stayed in the 11:1 region all the way to 6500 rpm where it dipped into the high 10:1 AFR. Note the 10.74:1 AFR dip at peak boost. I am still working on this one.
For the power charts, I did something different this time. I averaged the results from the last two runs that GG and I did and then exported them to an excel spread sheet. Then I exported the best stock run that we did into the same excel spread sheet. Averaging the results should provide for a more consistent result from a road dyno. Please note that these power runs were done in 4th gear and not in 3rd gear which tends to produce slightly higher power numbers:
Here is the power that GG’s car put down:
http://www.sr20deracing.com/EVO/Danny/dyno_before_after.gif
First, most of the gains were concentrated above 5000 rpm. In the 6000-6500 rpm range, for example, the car gained an impressive 59 hp and 49 ft-lb over stock. If you look at the entire 5000 to 7000 area under the curve you will see an average hp gain of 37.6 hp and 32 ft-lb of torque. The rest of the gains/losses are noted on the graph.
I am not done tinkering with this car. I did create a ninja mivec intake map for it, but I have not messed around with the mivec exhaust map. I am going to have a sit down with Mike Kojima and see what we can come up with for the mivec exhaust map. I have a feeling that threre is more power to be had from a properly designed mivec exhaust map.
So, stay tuned, there is more to come.Â* Â*