Dev Call Feb 26, 2024

AP_RangeFinder: add support for RDS02UF radar driver on seria

ArduPilot:master ← zebulon-86:driver_rangefinder

opened 01:54AM - 28 Feb 23 UTC

zebulon-86

+360 -49

## Summary Resubmit pull request，relevant : https://github.com/ArduPilot/ardu…pilot/pull/22385 Adding new features AP_RangeFinder:add support for RDS02UF radar driver on serial。 ## Documentation Product URL：[Microbrain Intelligent](https://www.microbrain.com.cn/en/ProductCenter/list.aspx?categoryid=132) Related Manuals:[【Benewake(北醒)】毫米波雷达在无人机上避障资料汇总](https://blog.csdn.net/zoran_wu/article/details/126750128) Here are the dumped Manuals:[RDS02U_Uart_Manual](https://github.com/zebulon-86/Document/blob/master/MicrobrainRadar/RDS02U_Uart_Manual.md) For PDF:[RDS02U](https://github.com/zebulon-86/Document/tree/master/MicrobrainRadar/pdf) ## Change Some tweaks to the code related to #22385 [libraries/AP_RangeFinder/AP_RangeFinder_RDS02UF.h](https://github.com/ArduPilot/ardupilot/pull/22385/files#diff-bf4ef8c6a1897e12756a77e6efd92c249789ccd5788f54610ad9a6d1e6140622) ```c #if RDS02_USE_CRC const uint8_t crc8_table[256] = { 0x93,0x98,0xE4,0x46,0xEB,0xBA,0x04,0x4C, ``` The CRC table takes up too much flash space,so added the define of whether to enable CRC, which is not enabled by default, and only use the frame header and frame tail to judge the integrity of the frame. It is not a big problem not to do CRC when the serial communication is stable. Users can also choose to enable CRC according to the hardware situation.

AP_Scripting: add ROMFS to lua path

ArduPilot:master ← IamPete1:ROMFS_require

opened 10:42PM - 06 Feb 24 UTC

+36 -0

This allows `require` to find modules in ROMFS. Possibly helps with other file s…ystem methods too. Note that this adds the new search path even if we don't have ROMFS compiled in. I think it will be fine, but I have not tested. Adding `"./modules/?.lua;" "./modules/?/init.lua"` to get to the same place relative to a script loaded in ROMFS does not seem to work, I'm not sure why. This should allow a SD card script to require something from ROMFS, which using relative paths would not. Because of the ordering it should also mean a SD card version of the same file gets priority over a ROMFS copy. Its convenient for development but might give users a very exciting way to break stuff. We could swap the ordering so ROMFS takes priority. Again, I have not tested this.

Copter: Autotune performance and safety enhancements

ArduPilot:master ← lthall:20240217_Autotune

opened 02:21AM - 17 Feb 24 UTC

lthall

+123 -87

This PR attempts to address some of the long term concerns with AutoTune. 1. …Autotune now fails when it becomes clear the aircraft will probably fail to tune well. 2. The recovery from the rate twitch makes use of the control model to reduce the severity of the ESC command. This should dramatically reduce the chance of desync. 3. The waiting for level criteria has been adjusted to reduce the chance initial conditions cause problems with the tests. 4. The practice of sipping a side in an attempt to save time has been removed as it causes aircraft velocity to build. I believe removing this is a net gain in both time and reliability. 5. Base test angles on ratio's of ANGLE_MAX. These changes are expected to dramatically reduce crashes due to desync and successful autotune with bad/dangerous tunes. We will see an increase number of complaints and questions along the lines of "Why did autotune fail to complete" and "Autotune failed with a failed to level error". Improvements for a future PR 1) Make the test twitch more consistent on each test by continuing the twitch after the measurement is complete. This will keep twitches symmetrical and reduce the build up of velocity 2) Update the twitch rate and angle target based on the results of the last four twitches to reduce the number of aborted tests and therefore the total time taken. 3) Another protection I should add is to stop angle P term increase when negative rates are observed but angle overshoot is small. This will help prevent excessively large Ange P results.

Simplify g2 object conversion

ArduPilot:master ← peterbarker:pr/parameter-conversion-static-const-array

opened 10:34AM - 21 Feb 24 UTC

+97 -229

``` Board AP_Periph blimp bootloader copter heli iofirm…ware plane rover sub CubeOrange-periph-heavy * Durandal -48 * -68 -68 -112 -96 -88 Hitec-Airspeed * KakuteH7-bdshot 48 * 20 20 -16 -8 -8 MatekF405 60 * 68 60 60 44 32 Pixhawk1-1M-bdshot 60 60 60 60 44 40 f103-QiotekPeriph * f303-Universal * iomcu * revo-mini 60 * 68 60 60 44 32 skyviper-v2450 60 ```

Plane: Manual: use `apply_throttle_limits` function in manual mode if `IDLE_GOV_MANUAL` is enabled.

ArduPilot:master ← IamPete1:manual_idle

opened 06:57PM - 21 Feb 24 UTC

+13 -16

This moves the application of throttle limits to the main `apply_throttle_limits…` function that all other modes used here: https://github.com/ArduPilot/ardupilot/blob/e01e697343cf3eb9b0991d7153f135b60b96f965/ArduPlane/servos.cpp#L507 This makes manual mode behave the same as any other mode. This does result in a couple of changes in behavior: - `THR_MAX` is also applied along with `THR_MIN`. - Battery compensation is done if enabled. - Watt limiting is done if enabled.

AP_GPS: improve reporting of backend inaccurancy

ArduPilot:master ← peterbarker:pr/gps-backend-debug-improvement

opened 03:46AM - 22 Feb 24 UTC

+2 -2

I didn't find this debug message particularly informative; add more info. Old…: ``` calculate_moving_base_yaw:372: Exceeded the permitted error margin 29.73999 > 0.052001 calculate_moving_base_yaw:372: Exceeded the permitted error margin 29.73999 > 0.052001 calculate_moving_base_yaw:372: Exceeded the permitted error margin 29.73999 > 0.052001 calculate_moving_base_yaw:372: Exceeded the permitted error margin 29.73999 > 0.052001 ``` New: ``` calculate_moving_base_yaw:372: Offset=0.300000 vs reported-distance=0.250000 (max-delta=0.050000) calculate_moving_base_yaw:372: Offset=0.300000 vs reported-distance=0.250000 (max-delta=0.050000) calculate_moving_base_yaw:372: Offset=0.300000 vs reported-distance=0.250000 (max-delta=0.050000) calculate_moving_base_yaw:372: Offset=0.300000 vs reported-distance=0.239990 (max-delta=0.047998) ``` Only affects things when you've got the compile-time debug option enabled

Correct use of various defines

ArduPilot:master ← peterbarker:pr/misc-define-fixes

opened 09:01AM - 23 Feb 24 UTC

+23 -5

Mostly excluding compass code, but a few other minor fixes ``` Board … AP_Periph blimp bootloader copter heli iofirmware plane rover sub CubeOrange-periph-heavy * Durandal * * * * * * * Hitec-Airspeed * KakuteH7-bdshot * * * * * * * MatekF405 * * * * * * * Pixhawk1-1M-bdshot * * * * * * f103-QiotekPeriph * f303-Universal * iomcu * revo-mini * * * * * * * skyviper-v2450 * ```

AP_HAL_ChibiOS: allow AP_TERRAIN_ENABLED to be turned off in hwdef

ArduPilot:master ← peterbarker:pr/terrain-fixes

opened 09:11AM - 23 Feb 24 UTC

+11 -2

adds the ifndef, and changes things to the path defines are undefined if it is f…alse ``` Board AP_Periph blimp bootloader copter heli iofirmware plane rover sub CubeOrange-periph-heavy * Durandal * * * * * * * Hitec-Airspeed * KakuteH7-bdshot * * * * * * * MatekF405 * * * * * * * Pixhawk1-1M-bdshot * * * * * * f103-QiotekPeriph * f303-Universal * iomcu * ``` I've run the fuller version of the size-check and nothing of note has changed: ``` pbarker@threads:~$ ~/rc/ardupilot/Tools/scripts/filter_size_compare_branches_csv.py /tmp/some.csv --hide-binary-identical ------------------------ ----- ------ ---- ----- ----- --- Board blimp copter heli plane rover sub SITL_arm_linux_gnueabihf 0 0 0 0 0 0 SITL_x86_64_linux_gnu 0 0 0 0 0 0 ------------------------ ----- ------ ---- ----- ----- --- pbarker@threads:~$ ```

Scripting: only try loading from ROMFS if scripts exist

ArduPilot:master ← IamPete1:lua_ROMFS

opened 12:49PM - 24 Feb 24 UTC

+21 -3

https://github.com/ArduPilot/ardupilot/pull/26225 means that ROMFS will fail to …open directories that don't exist. This results in a failed to open directory warning in ROMFS: ![image](https://github.com/ArduPilot/ardupilot/assets/33176108/03755b3f-9db3-445d-abc3-2a3ef71b6d54) We can tell at compile time if there are any scripts we could try to load, so we can compile out trying to load from ROMFS.

AP_Motors: Tricopter: fix actuator indexing

ArduPilot:master ← IamPete1:tricopter_actuator_index

opened 02:34PM - 24 Feb 24 UTC

+13 -15

This corrects the tricopter actuator indexing it was off by one, confusingly `AP…_MOTORS_MOT_1` = 0. This fix means it lines up with the indexing used by quadplanes to output in forward flight in the `output_motor_mask` method: https://github.com/ArduPilot/ardupilot/blob/728d9a9f965ca9b644116887f51e964cdf0eef99/libraries/AP_Motors/AP_MotorsMulticopter.cpp#L726 This also corrects to zero the `_actuator` in the `SpoolState::SHUT_DOWN` case so the slewing picks up again in the correct spot. Nether of these bugs are a issue unless your using a slew limit on the outputs with `MOT_SLEW_UP_TIME` and `MOT_SLEW_DN_TIME`.

AP_GPS: added GPS_DRV_OPTIONS for fully parsing RTCMv3 injection

ArduPilot:master ← tridge:pr-log-gps-inject

opened 11:07PM - 24 Feb 24 UTC

tridge

+139 -29

This adds GPS_DRV_OPTIONS bit 6 (so value 64) to enable a full RTCMv3 parse of a…ll incoming MAVLink GPS_RTCM_DATA messages, with a separate parser per mavlink channel This allows for setups where you have redundency in comms links, such as is common for light show drones which may inject RTCM data via both WiFi and a SiK radio setup as receive only This also gets RTCM logging at the message ID level, allowing you to debug RTCM injection issues with a lot more detail

AP_ESC_Telem: support Extended DShot Telemetry v2

ArduPilot:master ← mbuzdalov:edt-v2-support

opened 03:16PM - 25 Feb 24 UTC

mbuzdalov

+199 -14

### Description This PR adds support for Extended DShot Telemetry v2, the spe…cification for which is available [here](https://github.com/bird-sanctuary/extended-dshot-telemetry). Currently, recent versions of Bluejay and AM32 support this specification, and I believe BlHeli32 also has some limited support. Notably, EDTv2 introduces two new types of messages, which are useful to investigate the performance of ESCs: - the "stress level" message that contains a number in the range [0..255], which is the greater the more effort an ESC had to make during commutation; - the "status" message contains three bits for different warning levels: "alert", "warning" and "error", as well as the maximum stress level encountered, which is in the range [0..15]. The existing EDT parser handles these messages, but ignores them. This PR propagates them all the way through the `AP_ESC_Telem` machinery and logs the obtained information using two new log messages: - `EDT2` logs the status messages, so it has fields `Alert`, `Warning`, `Error` and `MaxStress`; - `EDTS` logs the stress messages, with the only field `Stress`. ### Implementation details Note that two different messages needed to be introduced, because the stress messages come more frequently than the status messages, and the stress level itself is scaled differently. For instance, there is a non-obvious but linear correspondence in Bluejay code, e.g. 120 maps to 0 and 255 maps to 15, and other firmware providers can implement a different relationship. As these messages come at a relatively low rate, and my use cases in particular use logging rate limiting, we do not want the information in these messages to be lost, there are two important implementation details: - when the next message comes, the telemetry values are aggregated rather than replaced, that is, status bits are OR-ed, and stress values are max-ed; - if the backend refuses to log the message, we don't drop the existing values, so that we do not lose any information, and the next attempt to log them may succeed. ### Tests performed This PR was tested on two copters, one quadcopter using the FlywooF405S-AIO controller, and one coax copter using the MatekH743-WLITE controller, both running Bluejay 0.19+ on their ESCs. The logs I obtained during testing appear to be rather sane and consistent. I attached [one of them](https://github.com/ArduPilot/ardupilot/files/14396764/24-02-25_14-01-52_edt.bin.zip) as an example. ### Thoughts on firmware and log sizes Initially I thought that the amount of new code seems to be not very large and should not lead to firmware exceeding the available size. However, it apparently exceeded the size on Pixhawk1-1M, so I surrounded the code with `#ifdef HAL_WITH_BIDIR_DSHOT`, which, retrospectively, makes sense. Other than that, I think this contribution can be important to FPV flyers, so there is a merit to include it in low-flash builds that support bidirectional DShot. The rate of these messages in current ESC firmware implementations seems to be quite low, so it should also be rather safe from this perspective.

Rename AP_Logger_DataFlash to AP_Logger_JEDEC

ArduPilot:master ← peterbarker:pr/logger-rename-jedec-to-dataflash

opened 02:15AM - 26 Feb 24 UTC

+20 -20

The naming if this driver class is *really* confusing! This is just one of th…e dataflash backends we support, the other being the W25N01GV. Done as two commits, one to rename the files, the second to adjust for the renaming. ``` Board AP_Periph blimp bootloader copter heli iofirmware plane rover sub CubeOrange-periph-heavy * Durandal * * * * * * * Hitec-Airspeed * KakuteH7-bdshot * * * * * * * MatekF405 * * * * * * * Pixhawk1-1M-bdshot * * * * * * f103-QiotekPeriph * f303-Universal * iomcu * revo-mini * * * * * * * skyviper-v2450 * ```

AP_NavEKF3: move initialisation of rngBcn into BeaconFusion method

ArduPilot:master ← peterbarker:pr/beacon-move-init

opened 04:47AM - 26 Feb 24 UTC

+46 -38

This just moves the code 1:1 First step in privatising these variables. I'…ve copied the "from" text in this branch into a file, and the "to" next into a file, run perl against the former to remove rngBcn and compared the two. The difference then is just grabbing the dal object. ``` Board AP_Periph blimp bootloader copter heli iofirmware plane rover sub CubeOrange-periph-heavy * Durandal -16 * -16 -16 -16 -16 -16 Hitec-Airspeed * KakuteH7-bdshot -24 * -24 -24 -24 -24 -24 MatekF405 * * * * * * * Pixhawk1-1M-bdshot * * * * * * f103-QiotekPeriph * f303-Universal * iomcu * revo-mini * * * * * * * skyviper-v2450 * ```

Plane

Plane 4.5.x issue tracking

opened 03:03AM - 13 Feb 24 UTC

tridge

Plane DevCallTopic ReleaseAdmin

Pending: - [ ] centre throttle options not working: https://discuss.ardupilot.…org/t/plane-4-5-0-beta-release/112518/14 - fix in mission planner https://github.com/ArduPilot/MissionPlanner/pull/3291 - [ ] ramping of motors on tilt quadplanes not working on back transition - fix https://github.com/ArduPilot/ardupilot/pull/26249 - [ ] POSITION_TARGET_GLOBAL_INT alt is wrong for terrain following - [ ] ALT_ASSIST not working while completing fwd transition? - [ ] terrain following not working from long distances in QRTL Resolved - [x] https://github.com/ArduPilot/ardupilot/pull/26133 Longer term - [ ] https://github.com/ArduPilot/ardupilot/pull/25442 - [ ] rebuild newlib and remove all possible abort() calls

Copter & Rover