Quote:
Originally Posted by Turboh
I have summarized roughly a hundred WOT runs on various maps at various IATs. I have looked for the smoothest and quickest map and user settings. For each run I graphed timing, rpm, boost vs time. All on 93 octane. Same road, same elevation, different weather conditions, spring, summer, fall.
I imported the csv files with excel and ran the graphs with it. Rate changes are easily calculated using the excel trend lines. Slopes in the trend line equations are the rate of change, or equivalent to quickness. The higher the rate the quicker.
To me the rpm vs time graphs shows the effect of boost and timing best, without going to a dyno.
Looking at Maps 6, 5, 4,3,2, and 1, I think map 6 at 15.5 psi flat is the best overall. Map 5 and 3 appear a bit quicker at times but not very smooth and less adjustable. I assume that is due to internal timing adjustment spoofing by the JB4.
Comparing map results must be done at the same IAT.
Such fun....

Thank you for posting this, and for sharing your thoughts with all of us.
That's pretty cool.
Couple of questions/comments, if you don't mind.
First and foremost, can you post those summarized charts, with the datapoints?
Quote:
To me the rpm vs time graphs shows the effect of boost and timing best, without going to a dyno.

Do the slopes you produce on rpm mathematically (not just visually) correlate with boost at all? Do they correlate with mph? Have you tried to graph mph to time and see if it correlates more or less strongly to both boost or rpm? I have more on this below...
Quote:
Slopes in the trend line equations are the rate of change, or equivalent to quickness.

I agree on rate of change, assuming linear, but what you really want to do is slope of the slopes (second derivatives) because it's not exactly linear. What is your ultimate goal, e.g. how do you define "quickness?" Is it how quickly the car gets to 60mph, or maybe how quickly it covers a given distance, or some other metric? To me, boost and rpms and all that are just a means to an end, the end being how well the car accelerates, and what is its maximum speed for as long a time as possible. You could measure relative speed. If the distance is short enough, measure out the set distance, launch the car, record your times with each setting. You could measure acceleration  I think the JB4 records MPH and you could set the log rate to raw sampling or 2 seconds, then you would be able to see the second derivative (the rate of change of velocity over time a = dv/dt) approximating an acceleration curve.
I'm really curious why velocity over time is not a key metric in your measurements.
Obviously none of this substitutes for a dyno, which measures torque and power, increasing both of which would be my real endgame...
Quote:
Comparing map results must be done at the same IAT.

How do you control for IAT, since it correlates so strongly to results? Even with me injecting meth to lower IAT (which should yield better timing, etc), I can't see how you can control for a set IAT to experiment with?