http://www.amazon.com/Modify-Management-Sy...89825264&sr=1-2
http://www.amazon.com/Engine-Management-Ad...89825289&sr=1-1
Read those two books and you'll be way ahead of the curve. The 2nd is a very well written, analytical approach to proper use of dyno and data acquisition in engine tuning. The author has experience with OEM calibrations as well as many very high horsepower racing engines. It's amazing the myths that are propagated by internet experts that get dispelled by guys that tune for a living. The first has a lot more theory and how and why engine mechanicals and sensors work the way they do.
Start with fuel. Always work fuel first. Begin with a semi-conservative timing setting. Since the ITS engine combination is stout and you pretty much have to try to burn one down you can begin with timing near what you would think is optimal. In this case we'll say begin at about 22-25* total advance at WOT. Do this by making a pull and checking timing with the light as the engine is run through it's rev range. Once you have that initial timing adjustment made you can begin fuel tuning.
If you want to seriously compare spark plugs prepare for a long day. It's entirely possible that different plugs can require very slight differences in a/f and timing for total optimization. A hotter, longer spark can light leaner mixtures. You shouldn't run lean enough to be at the point of a lean misfire on any plug but in theory it's possible. The same holds true for timing. If one plug is more shrouded than another it may require slightly more lead for proper burn. In this case I'd say any changes would come down to statistical noise and that's what I've personally found but without being prepared to optimize each plug it's not a complete comparison.
At any rate work out the fuel numbers first. If you're already seeing rail pressures down in the 30 psi range you are giving up power by running pressures that low. There's really nothing you can do about it without getting into the software/hardware and changing the injector pulse width but just to be sure there is power to be had by increasing rail pressure for better atomization while maintaining the same a/f ratio. But you have what you have and optimizing it is simply doing a series of pulls to record how the changes are affecting power.
One thing to watch for is lambda sensor placement. If you're using the dyno's sensor and it's placed in the tail pipe you may have issues. I used to say it didn't matter. And the truth is that if the sensor is good and the exhaust system is free of leaks it shouldn't. However I had a string of poor results early this year with tail pipe sensors and won't use them anymore. Even if you don't have a sensor of your own install a bung in the header and insist that the dyno shop install their sensor in the bung. Ideally you'd want the sensor as close to the exhaust port as possible but the high EGT of a rotary will both skew sensor readings and kill a sensor in short order if it's too close to the port. Somewhere between 12" from the port and the collector flange is a good rule of thumb. If you're stuck with a tailpipe sensor and the readings look strange or consistently on the lean side suspect a problem with sensor placement. Exhaust leaks before the sensor will show up as lean readings.
Tune fuel for best power. We've found best power in the 13.7 range with power falling slightly above and below that number. In my experience this is a safe number. Basically you're looking at watching what the power output is and adding or subtracting fuel for best power. With the gross adjustment of the fuel pressure regulator you are making compromises across the board and you can be faced with the question of whether a 2hp gain at peak is worth losing 8 hp at 6000. I wouldn't tune solely trying to optimize the high end. All you're looking for is the most average power. If that means everything works best at a lower RPM and your shift point falls at 7500 instead of 8500 so be it. I've never worked with an S4 intake/electronics package so I have no idea how it'll work for you. Just keep in mind that you need to give the engine what it wants and if it doesn't want to work at 8500 you can't make it.
Once fuel is done go to work on timing. Advance the timing until the power doesn't improve. An ITS engine is not knock limited so you can advance timing to MBT and be safe. Advance it in 1* increments until you arrive at the point where power does not increase and lock it there. Again with a factory ECU this may be a compromise. You should see MBT at about 25-27*. The gains can be quite dramatic between 20 and 25*. I think the reason you hear slightly different numbers being thrown around simply goes back to tolerances stacking up. Give your engine what it wants not what someone tells you it needs.
Tune with the fuel you plan to run. 87 octane is fine and you won't have to worry about knock. However, if you feel the need to run 110 leaded for racing you need to tune with that. Higher octane fuels generally have a slower flame travel speed and require more timing lead to produce optimum power. So if you tune with 87 then dump 110 in the tank you will most likely lose a small amount of power. Hint: though I've never tried them and they probably wouldn't pass an SCCA fuel tech if they ever decided to have one, the drag race community is big on the oxygenated fuels offered by the fuel manufacturers. If you don't mind paying $6-8/gallon all the time it would be worth it to at least try the fuel on the dyno. Again timing will need to be optimized for the fuel.
Something else to keep an eye on is engine coolant temp. Try to keep the coolant temps fairly stable between runs. Don't let the engine cool dramatically between runs. This may produce higher numbers but those numbers and the settings you come up with while obtaining those numbers may be irrelevant in a race situation. If the ECU allows a much leaner fuel trim at 180* than it does at 200* and you race at 200* but tune at 180* you'll be fat and down on power in a race situation.
And I'm supposed to be receiving an S4 beta unit soon. Everything is done except for testing.
