| commit | 456396b25a9597ec614d60a381478b55dea68756 | |
| author | Sean Vig <sean.v.775@gmail.com> | |
| Thu, 17 Sep 2015 06:54:43 +0000 (17 01:54 -0500) | ||
| committer | Dominic Clifton <me@dominicclifton.name> | |
| Tue, 6 Oct 2015 17:33:23 +0000 (6 18:33 +0100) | ||
| tree | faf9f95ed1f983fd3030e90addbad17e4ac801c0 | treesnapshot (tar.gz zip) |
| parent | e5f3f1794fade3185bd7bba164072a3e1aed4257 | commitdiff |
Use fixed point calculation for BMP280 barometer
Substitute the floating point calculation for the comparably much faster
64-bit calculation recommended for a Cortex-M3 (floating point is
recommended only in PC applications with FPU) [1]. Even if we have an
FPU, we round the temperature to the nearest 1/100th of a degree and the
pressure to the nearest 1 Pa, both of which are within the limits of the
fixed-point algorithm's accuracy.
Conflicts:
src/main/drivers/barometer_bmp280.c
Substitute the floating point calculation for the comparably much faster
64-bit calculation recommended for a Cortex-M3 (floating point is
recommended only in PC applications with FPU) [1]. Even if we have an
FPU, we round the temperature to the nearest 1/100th of a degree and the
pressure to the nearest 1 Pa, both of which are within the limits of the
fixed-point algorithm's accuracy.
Conflicts:
src/main/drivers/barometer_bmp280.c
| src/main/drivers/barometer_bmp280.c | diffblobblamehistory | |
| src/test/unit/baro_bmp280_unittest.cc | diffblobblamehistory |