Commit Graph

833 Commits

Author SHA1 Message Date
Andrew Tridgell 78858bbcdd AP_GPS: switch ublox over to 230400 baud
this ensures we have sufficient bandwidth for raw data
2020-06-01 17:18:17 -07:00
Andrew Tridgell 529bfe8fcd AP_GPS: fixed issue with GPS selection for moving baseline
when you have a moving baseline pair of ublox GPS modules and the
rover GPS does not have full fixed RTK lock on the base GPS then we
should not use it as our primary GPS as it's position and velocity can
be badly affected by the attempts of the GPS to gain a fixed lock.

This was observed in a flight with two F9P GPS, where the GPS velocity
data from the rover GPS went way off when it lost full RTK lock. It's
status stayed at 4, so it was selected as the primary GPS
2020-05-27 10:59:46 +10:00
Andrew Tridgell f00392d567 AP_GPS: don't accept a zero GNSS timestamp from UAVCAN GPS
this is needed for a bug in AP_Periph 1.0 which could briefly send a
zero timestamp on first fix marked as a UTC time
2020-04-29 12:20:19 +10:00
Andrew Tridgell 4f591a338f AP_GPS: use GCS_SEND_TEXT() 2020-04-29 07:36:55 +10:00
Andrew Tridgell 6347d0adb0 AP_GPS: leave uart2 config enabled when using uart2 MB setup 2020-04-21 15:30:54 +10:00
Andrew Tridgell f3a5d915f3 AP_GPS: allow for larger average delta on MB rover than normal 2020-04-21 15:30:54 +10:00
Andrew Tridgell 36098c3221 AP_GPS: improved delay health threshold
allow for missing frames without showing unhealthy on GCS
2020-04-21 15:30:54 +10:00
Andrew Tridgell 3397ed766d AP_GPS: enable ublox moving baseline compilation option
disable for HAL_MINIMIZE_FEATURES and if max receivers 1. This fixes
the f103-GPS AP_Periph build
2020-04-21 15:30:54 +10:00
Andrew Tridgell ae2132d861 AP_GPS: added check for Z error in moving baseline ublox
this ensures we don't get bad yaw from having the right distance but
incorrect Z offset for the currently expected roll/pitch
2020-04-21 15:30:54 +10:00
Andrew Tridgell 7027eecd34 AP_GPS: added GPS_DRV_OPTIONS
this allows for configuration of moving baseline with either uart1 or
uart2 for the RTCM data. Using uart2 requires an extra cable between
the two modules, but requires less uart bandwidth which is good when
DMA channels are low. Using uart2 also avoids the rtcmv3 parser, which
saves memory
2020-04-21 15:30:54 +10:00
Andrew Tridgell 9a6de209ab AP_GPS: ensure that PVT and RELPOSNED msgs time align
this ensures we wait until we have a matching PVT and RELPOSNED
message before we report new data. If we stop receiving RELPOSNED
messages then disable yaw support
2020-04-21 15:30:54 +10:00
Andrew Tridgell 3fbeae613b AP_GPS: fixed moving baseline yaw calculation
this reverts #13955 and instead applies the correct fix, which is to
subtract the angle instead of adding.
2020-04-21 15:30:54 +10:00
Andrew Tridgell 077ab22767 AP_GPS: fixed AP_Periph build with gcc 9.x 2020-04-01 17:08:48 +11:00
Andrew Tridgell 4456732911 AP_GPS: fixed rotation of moving baseline GPS
we were off by 180 degrees. Thanks to Jaime for noticing this!
2020-04-01 17:08:19 +11:00
Peter Barker 2b08af69bc AP_GPS: fix dereference-of-nullptr warning from scan-build
This is kind-of a false positive as this in practise could never be
nullptr.
2020-03-31 10:37:16 +11:00
Andrew Tridgell f00314a5d1 AP_GPS: removed dead blending code 2020-03-17 12:13:54 +11:00
Andrew Tridgell 9fa034d1de AP_GPS: fixed antenna offset for blending
we were accessing beyond the end of the antenna_offset array
2020-03-17 12:13:54 +11:00
Randy Mackay a19634b9d2 AP_GPS: reduce POS param range to 5m and add increment 2020-02-01 08:46:03 +09:00
Andrew Tridgell 7efd5248ee AP_GPS: added get_type() API 2020-01-26 09:12:55 +11:00
Andrew Tridgell e59f32074d AP_GPS: fixed yaw error when one GPS has zero position
the M_PI correction was only valid for one zero side. Much simpler to
always calculate difference
2020-01-24 11:46:40 +11:00
Andrew Tridgell cf8becfa71 AP_GPS: make all semaphores recursive
the cost is very similar and this prevents an easy coding error which
can occur on less used code paths
2020-01-19 20:19:30 +11:00
Andrew Tridgell f0d1d3fa90 AP_GPS: added get_semaphore() 2020-01-16 20:36:54 +11:00
Andrew Tridgell db777c56b9 AP_GPS: provide yaw feedback in GPS_RAW_INT and GPS2_RAW
allow GPS to display status of GPS yaw for moving baseline and NMEA
2020-01-07 12:37:35 +11:00
Andrew Tridgell e40c570615 AP_GPS: faster config of moving baseline
thanks to MdB for suggestion
2020-01-07 12:37:35 +11:00
Andrew Tridgell e660532f67 AP_GPS: fixes from PR review 2020-01-07 12:37:35 +11:00
Andrew Tridgell 41289fe945 AP_GPS: adjust lag for MB RTK ublox
rover will lag by about 40ms
2020-01-07 12:37:35 +11:00
Andrew Tridgell b86c43aa66 AP_GPS: prefer RTK MB rover when it has yaw available
this ensures that yaw data is time aligned with position and velocity
data
2020-01-07 12:37:35 +11:00
Andrew Tridgell 261465ef96 AP_GPS: split out update_primary() from update()
this makes the selection of primary GPS a lot clearer, with less
nesting of if statements
2020-01-07 12:37:35 +11:00
Andrew Tridgell 93aaf98672 AP_GPS: added automatic u-blox moving baseline configuration
this uses GPS_TYPE and GPS_TYPE2 to setup two u-blox F9 receivers in a
RTK moving baseline configuration
2020-01-07 12:37:35 +11:00
Andrew Tridgell 71a0f3a1dc AP_GPS: added RTCMv3 parser
will be used for moving baseline support
2020-01-07 12:37:35 +11:00
Andrew Tridgell 6efbc484fd AP_GPS: cope with UAVCAN GPS that don't provide Aux message
thanks to @VadimZ for the suggestion
2019-12-05 15:48:51 -07:00
Michael du Breuil 8ca6535661 AP_GPS: Log SBF commanded disarms 2019-11-20 11:06:16 -07:00
Michael du Breuil f8808022b8 AP_GPS: Check the driver logging health 2019-11-20 11:06:16 -07:00
Andrew Tridgell 379e3d60a1 AP_GPS: fixed sizes of config items 2019-11-16 15:28:52 +11:00
Michael du Breuil 63a6a1084c AP_GPS: UBlox: Fix the assumption that all keys are 1 byte keys 2019-11-16 15:28:52 +11:00
Michael du Breuil 150a4edcf2 AP_GPS: UBlox fix not listing the time mode as a configuration reason 2019-11-16 15:28:52 +11:00
Andrew Tridgell 7bec9d87b4 AP_GPS: fixed TMODE config issue on non-F9 GPS
thanks to Michel Pastor for reporting this
2019-11-16 15:28:52 +11:00
Andrew Tridgell 8aa781206d AP_GPS: fixed signed/unsigned comparisons 2019-11-09 17:11:32 +11:00
Andrew Tridgell b7c9391619 AP_GPS: support Fix2 message for UAVCAN
this allows for RTK fix levels
2019-11-09 17:11:32 +11:00
Andrew Tridgell 632114d687 AP_GPS: support RTCMStream for RTCM injection on UAVCAN 2019-11-09 17:11:32 +11:00
Andrew Tridgell 6c31fb1bb6 AP_GPS: added automatic config of TMODE on F9
the Drotek F9 GPS ships with TMODE enabled, which means we don't get a
3D fix.

This also adds the VALGET/VALSET msgs we will need for automatic
moving baseline config
2019-11-09 17:11:32 +11:00
Michael du Breuil d623da0de6 AP_GPS: Always send GPS2_RAW after we've seen the second GPS
This significantly improves the GCS's ability to correctly display whats
happening with the second GPS.
2019-11-02 10:35:08 +11:00
Michael du Breuil 5bafd7d72c AP_GPS: Don't defeat the GCS message interval code 2019-10-28 00:26:28 -07:00
Andrew Tridgell 14cdac9202 AP_GPS: fixed GPS to UART mapping
this allows for first GPS as UAVCAN and 2nd as UART. That is not
possible currently unless you waste a uart
2019-10-24 14:37:30 +11:00
Andrew Tridgell 00260db65a AP_GPS: fixed UAVCAN GPS status when disconnected
this fixes the status reported for a UAVCAN GPS when the cable is
unplugged. It should be "NO GPS", but was instead reporting "NO
FIX". The user had no way to distinguish between not finding a CAN GPS
at all and it not getting satellites

thanks to CUAV for reporting
2019-10-14 19:08:37 -06:00
Andrew Tridgell 37fb028532 AP_GPS: improved checking for valid yaw for moving baseline
and fixed check for carrSolnFixed
2019-10-15 11:06:11 +11:00
Michael du Breuil 2587b13ae7 AP_GPS: UBLOX: Parse RELPOSNED messages
This does not yet:
  - validate the receiver configuration
  - manage timing out stale GPS heading info
  - relPosNormalized usage isn't clear, which may defeat the STRICT_LENGTH_CHECK
2019-10-15 11:06:11 +11:00
Andrew Tridgell b5217412f9 AP_GPS: fixed a bug in handling corrupt u-blox packets
when we have corrupt input due to loss of bytes on a UART we can end
up with a u-blox packet with zero payload bytes. When that happens we
need to bypass payload reading as otherwise we will end up keeping
reading bytes until the driver resets at 4 seconds, causing a GPS
outage.

This was causing GPS outages every few hours in copters running 3.6.7,
and was also reproduced in SITL using SIM_GPS_BYTELOSS.
2019-09-30 15:20:30 -06:00
Andrew Tridgell 9eb6c1be64 AP_GPS: use jitter correction on GPS_INPUT data
this allows for more accurate timing when using GPS_INPUT for indoor
positioning systems
2019-09-17 09:17:04 +10:00
Michael du Breuil f857eb4640 AP_GPS: Take responsibility for logging the GPS if the EKF wants logging 2019-09-10 09:09:17 +10:00