TrackPoint's drift cancellation
Posted: Sat Aug 06, 2005 5:38 pm
Hi,
I had a small discovery concerning the TrackPoint's drift: If you press one of the mouse button the recalibration doesn't occur, and hence non of the usual mouse drifts. I say non of the usual, because I read that the mouse drift cancellation algorithm is due to that Synaptics' TouchStyk (That's how they call the TrackPoint) is made of capacitors, and is sensitive to temperature changes. And so, I set the middle button to "none" (i.e. not scrolling or magnifying) and then kept it pressed while I gave the computer something to do, and raised the temperatures from 44C to 60C. When it got to 60C the mouse started to drift, but very gently, not like we know. Since I kept the middle button pressed the mouse kept drifting, and I could negate that drift with little effort with my finger. When I released the middle button, the mouse re-calibrated, and stopped drifting. Then I did the test in the opposite direction. I pressed the middle button, and kept the CPU free of chores. Again the drift started only after a major drop in the temperature has happened, and the CPU temp was 44C. But this time the drift was upwards, while before it was in a diagonal down to the lower left corner.
I now know why Synaptics do this, but I still think that they are overdoing it. After all it took a 14C change for some temperature related drift to happen. A 14C temperature change take quite some time to occur. Yet the drift cancellation algorithm happens so often, that many times it cancels my own movement in the process.
I had a small discovery concerning the TrackPoint's drift: If you press one of the mouse button the recalibration doesn't occur, and hence non of the usual mouse drifts. I say non of the usual, because I read that the mouse drift cancellation algorithm is due to that Synaptics' TouchStyk (That's how they call the TrackPoint) is made of capacitors, and is sensitive to temperature changes. And so, I set the middle button to "none" (i.e. not scrolling or magnifying) and then kept it pressed while I gave the computer something to do, and raised the temperatures from 44C to 60C. When it got to 60C the mouse started to drift, but very gently, not like we know. Since I kept the middle button pressed the mouse kept drifting, and I could negate that drift with little effort with my finger. When I released the middle button, the mouse re-calibrated, and stopped drifting. Then I did the test in the opposite direction. I pressed the middle button, and kept the CPU free of chores. Again the drift started only after a major drop in the temperature has happened, and the CPU temp was 44C. But this time the drift was upwards, while before it was in a diagonal down to the lower left corner.
I now know why Synaptics do this, but I still think that they are overdoing it. After all it took a 14C change for some temperature related drift to happen. A 14C temperature change take quite some time to occur. Yet the drift cancellation algorithm happens so often, that many times it cancels my own movement in the process.