AP_Baro: read data from BMP581 twice, reject if not identical#32171
AP_Baro: read data from BMP581 twice, reject if not identical#32171peterbarker merged 2 commits intoArduPilot:masterfrom
Conversation
|
Closes #31935 |
peterbarker
force-pushed
the
pr/bmp581-read-twice
branch
from
5436ac5 to
2a54a8b
Compare
|
Hi @peterbarker, Tested on hardware. |
|
Hi @peterbarker , can we make the change above and merge this? |
peterbarker
force-pushed
the
pr/bmp581-read-twice
branch
from
2a54a8b to
a80843c
Compare
I have adjusted the hwdef - please verify it's correct. |
@zebulon-86 Please double check. |
|
|
||
| # barometers (Holybro 6X more recent versions) | ||
| BARO BMP581 I2C:2:0x46 | ||
| BARO BMP581 I2C:0:0x47 |
There was a problem hiding this comment.
For newer Holybro 6X hardware (RC02), the BMP581 on I2C:0 uses address 0x46 rather than 0x47. This should be updated to:
# barometers (Holybro 6X more recent versions)
BARO BMP581 I2C:2:0x46
BARO BMP581 I2C:0:0x46
|
I also validated this on three RC02 Holybro 6X boards with long-duration tests of 2.2h, 8.9h, and 2.3h. No abnormal barometer data jumps were observed during any of these runs. Based on these results, the BMP581 driver changes in this PR appear to be working correctly on RC02 hardware as well. Representative log screenshots are attached below. 
|
vincentpoont2
left a comment
vincentpoont2
left a comment
There was a problem hiding this comment.
For newer Holybro 6X hardware (RC02), the BMP581 on I2C:0 uses address 0x46 rather than 0x47.
reports of corruption on this device
peterbarker
force-pushed
the
pr/bmp581-read-twice
branch
from
f967e72 to
201807d
Compare
|
Merging as discussed on call |
3 participants
reports of corruption on this device
needs testing on real hardware; hardware may not permit double-reading of data registers.