With a limited number of data-acquisition channels, it's tough to choose what's important.
This weekend I made a small upgrade to the data-acquisition system in my race car. Briefly, I added a throttle-input sensor under the gas pedal.
Why's that? Wasn't there already a throttle input? Well, yes, sort of. I'd always had an RPM sensor that worked off the ignition, basically counting sparks, dividing by four, and plotting engine revs. The system is also smart enough to calculate what gear I'm in without actually monitoring that directly. (It looks for discontinuities in road speed versus engine speed.) But measuring engine speed isn't the same as measuring foot pressure on the throttle, I've found. Now I've got both.
The difference is a bit subtle but it helps me detect periods where I'm asking the engine to do more than it can. Normally the system can "back out" throttle data based on RPM but that data's filtered by mechanical inertia. It's also fairly low-frequency. Even with a light flywheel, the engine can change RPM only so quickly. Dancing on the throttle, in contrast, is often much quicker and more, well, fidgety.
Finally, there are times you'll be mashing the throttle and the engine just can't respond because you're climbing uphill, or in too high a gear, or at the peak of the torque curve, or starved for air, or who knows what. Those cases go unrecorded because the throttle data is interpreted from engine data. Now I've separated the two.
Sadly, once the season starts up again in April, what I think I'll find is that I'm asking for more all too often. The post-race plots will probably show nonlinearities between the throttle data and the RPM data -- places where what I'm asking and what I'm getting are two different things.
Such is the real world. But without the data-acquisition system, I won't know where those are, and won't know how to fix 'em. At least this way I'll know where to look.