Let me start off by saying that the dyno day results were somewhat dissapointing, yet enligtening at the same time. The weather was nice, not too hot or humid like the last time. Regardless, intake temps measured at the air filter (pre-turbo) were still in the 96 degree range due to lack of airflow when the car is stationary on the dyno.

Run 1:

They strapped the car in and we got a baseline run. I had previously set up the car for about 21 PSI of boost but that was during 30 degree weather. On this pull, with the temperatures being what they were, I peaked at about 19 PSI. During the run the wideband O2 sensor blew out of my exhaust so I was unable to tune the car after the first run. As shown above, I logged about 10 knock at ~4200 RPM, which coincides with the peak torque production. I attributed this to running too lean, so I added a few percent of fuel to the 4500 and 5000 RPM ranges to see what that would do. Injectors are not yet maxed out and timing advance looks good for pump gas (93 octane Shell) peaking to over 20 during the run. The knock falls off at the higher RPM ranges as expected. A/F ratio (O2) looks good by me through the high RPM ranges.

Run 2:

Ran the car again without changing the setup beyond adding fuel to eliminate the mid-range knock. After Run 1, the FMIC was very warm to the touch on the inlet side and still cool on the outlet side. Insufficient airflow to really cool it between or during runs. This time we got a clean wideband O2 reading. The result of which was astonishing. From 5000 rpm to redline, my car was running so pig rich (that explains the black smoke you see) that the wideband O2 couldn't even monitor it. It was off the chart! Funny, my stock O2 sensor reads about ~.74v during this period, which indicates to me that everything is okay. Not good. I took out 5% fuel across the entire 5000-redline RPM range and we dynoed again.

Run 3:

By this time, the FMIC is very hot, scalding to touch. Insufficient airflow indoors to cool it. This will adversely affect my power production and create more heat/more knock for each succive run. Despite this, the subtraction of fuel from the high RPM ranges yielded an increase of 10 horses and ft lbs at the wheels! Still, during this run, the car ran so rich over this range the wideband O2 was unable to log it! As you can see above, injector duty cycle (in black) has fallen off, so I'm usling a lot less of my injector now, so I figure I can safely increase the boost again. Up to 22 PSI it goes. I lean it out an additional 5% across the board from 5000 to redline and go for another pull.

Run 4:

By this time, my FMIC is heatsoaked to death and the little airflow provided by the fan in front can't save it. As you can see, knock spiked to 22 at the peak torque production. This is sufficient knock to pull back timing advance which steals your power. The combination of heatsoak and too much boost caused this. In effect, my 16g cannot efficiently produce 22 PSI of boost under these conditions with pump gas. You can see the injector duty cycle is even lower now. I have plenty of fuel left which would tend to indicate you could run more boost, but because of the heat issues, it's just impossible today. Regardless, the subtraction of 10% fuel across the high RPM ranges finally yielded some noticable results and the wideband O2 graph finally peaked into the visible range for just a second. Another 10 horses and ft lbs were gained at the wheels. Even so, the car is still PIG rich and needs further tuning at the high end.

The results:

As you can see, the car made a paltry 231 all-wheel horsepower and 251 all-wheel torque, running pig rich at the high end. Due to the heat issues and subsequent knock, the timing retard caused the severe drop off of power over 4000 rpm which you can see above. When driven on the road under non heatsoaked conditions, this power curve should peak and smooth out without dropping off. With additional leaning out, more power is to be gained at the top end as well. I leaned it out another 5% after I was done on the dyno. If I had more pulls (and time to cool the FMIC down) I could have tuned it a lot better. But hey, four pulls for $43 is a pretty durn good place to start. At least now I know I can't tune by the stock narrowband O2s anymore. They are misleading and caused my tuning to be crappy going into the event. You can see how at 5500 rpm the wideband O2s started showing themselves at long last. Still too rich though.


My car makes significantly more torque than horsepower. This seems to be the case for other DSMs that dynoed that day. The EVOs I watched were opposite, they made more horsepower than torque. In fact, my car made as much torque as one modified EVO that day.

On compression ratio, my 16g is about maxxed out with my current set up even though I have extra fuel to offer, it is running out of breath. A lot of people are making well over 300AWHP with their 16gs, but my motor is too old and beat up and my 7.8:1 compression pistons are holding me back. The car would benefit significantly from a rebuild and to use some 2g pistons (8.5:1) or even higher ratio aftermarket. That would really unleash the power of the 16g. For now, though, the car makes more than enough power for what I use it for (namely autox and rallyx) so I'm perfectly content to run it like this for now. I am pleased with the amount of torque it makes (over 330 ft lbs at the crank) which will be great for pulling out of turns in rallyx this year. I had been hoping for higher numbers but this is to be expected on a motor with 170k miles on it. I shouldn't be too upset I guess.