ArduPilot 4.6 issues list

[rmackay9]

This is a list of issues discovered during ArduPilot 4.6 testing. Issues for Copter/Rover-4.5 issues are here

Reports requiring investigation

• Speedybee ENOMEM message (discussion)
• DroneCan M10 GPS lock issue again (raised by TimT, discussion)
• Microbrain UAV-R21 Proximity sensor (discussion, discussion2) -- @rishabsingh3003 will investigate

Confirmed issues

• Internal error involved multiheap allocation for LUA (Internal error from multi-heap allocation #29563)
• EKF3 altitude runaway EK3_SRC1_POSZ = 2 (rangefinder) and rangefinder is out of range (PR)
• AM32 passthrough (without DroneCAN) not working (issue, PR)
• AIS not working on Copter? (discussion) -- build server issue? Randy investigating
• QuadPlane EKF3 velocity divergence before arming (issue, PR) -- hoping PaulR will investigate (not a blocker)

Plane Specific Issues

• “pitch level-off at takeoff altitude” is interfered with by the feature “pitch rotate speed for ground takeoff” (discussion)
• Plane servos using IOMCU not moving (issue)
• the assisted_flight flag oscillates rapidly in fwd transition of tilt quadplanes, possibly leading to incorrect tllt slew rate Plane: quadplane tilt rate issues #28957 -- aiming to fix in 4.6.1 (or higher)
• terrain following issue: Plane: fixed terrain guided target intermediate alt handling #29685
• circle climb after terrain: Plane: fixed incorrect set of location alt fields #29765

Enhancements or Issues that may not be resolved before the stable release

• BLHeli SERVO_BLH_RVMASK mapping is achieved through trial and error (raised by xfacta, issue)
• when the user has configured ESC, FFT or RPM notch and has INS_HNTCx_REF=0 then they end up with a fixed notch. We should have an arming check to tell them they need to set the REF to 1.0 (any non-zero value will do actually) check that INS_HNTCn_REF is non-zero for ESC,RPM or FFT notches #28786
• ROS2 attitude control + thrust enhancement request (discussion)
• AutoTune should have a vehicle specific (or configurable) AngleP max (reported by Tridge)
• add a warning if the GCS system clock and UTC clock on the flight controller differs by over 20s and signing is active, as signing will fail with a 30s difference
• Custom Build Server for CubeOrangePlus with OpenDroneID shows linker error when out of flash (discussion1, discussion2, Issue)
• Add LDRobot LD19 support (discussion, issue)
• need to provide a clearer message to users when the power status flags change, indicating loss of a power source in flight. A statustext at message level CRITICAL I think is warranted, maybe "POWER STATUS CHANGE" ? (Tridge has details)

Wiki issues/enhancement requests (Wiki To-Do list)

• None

Resolved issues

• RPI PiPilot / Invensense ICM42688P hang (issue, PR) -- resolved for beta6
• on ICM45686 IMUs the SPI transfer overheads can be overwhelming on some boards, especially those with 3 of that sensor. Some PRs to address: Prearm check to validate backend read rates against scheduler loop rate #28672 and Do not Read FIFO faster than requested rate for ICM45686 #27573 -- resolved for -beta6
• AHRS upside soon after startup (reported by partners, PR) -- resolved for -beta6
• IMU temp cal pages need updating -- no details on what needs updating
• Copter/TradHeli internal errors 0X4000000 and 0X4000020 (discussion, issue, PR) -- resolved for -beta5
• SET_ROI_LOCATION support (issue, PR) -- resolved ahead of -beta5
• Take Pic with no location (issue, PR) -- fixed in master (won't backport)
• Pixhawk 5x ethernet not enabled (issue, wiki PR) -- wiki PR merged ahead of -beta5
• AP_Scripting should get follow binding for Rover (discussion, PR) -- resolved for -beta5
• Plane: fixed guided VTOL command in terrain mode (PR)
• ARSPD_AUTOCAL can trigger while disarmed (PR) -- resolved for -beta5
• QuadPlane motors using BDShot making knocking noises (discussion1, discussion2, PR1, PR2) -- fixed for beta5
• Water Depth message too fast on boats (discussion, PR) -- resolved for beta5
• UBlox M10 can't get enough sats (discussion) -- testing is inconclusive
• Copter MOT_THST_HOVER param desc for lower limit should be 0.125 (discussion, PR) -- resolved for -beta5
• iBUS telemetry not working? (discussion) -- user error
• Mission option bit issue (discussion, PR) -- resolved for -beta5
• CADDX GM2 update slow (discussion, PR) -- resolved for -beta5
• Unknown message 45 (discussion, discussion2, PR) -- resolved for -beta5
• Holybro M10 waiting to configure GPS1 (discussion, PR) -- resolved for -beta5
• LD06 support not working? (discussion, PR) -- fixed ahead of -beta5
• Reboot loop after enabling Scripting (discussion, issue) -- user using incorrect compiler
• possible regression in attitude control in quadplanes? possible regression in quadplane attitude controller? #29060 -- no evidence that there's been a regression despite much effort spent searching
• Takeoff throttle behavior for planes without airspeed sensor changed (issue, PR) -- intentional change
• ESC Telemetry logging issue (issue, PR) -- resolved for beta3
• Community dev struggles with build options (discussion) -- no response from user
• Floating Point Exception in the position controller #28969
• Autotune acceleration limiting is not effective (PR) -- resolved for beta3
• Orqa 3030 autopilot config issue (discussion, PR) -- resolved for beta3
• BeagleBoneBlue RC not working with Copter (discussion) -- user error related to running as a service
• Plane takeoff not completing under TKOFF_OPTIONS[0]=0 Takeoff not completing under TKOFF_OPTIONS[0]=0 #28685 (PR) -- resolved for beta3
• plane takeoff regression since 4.5.x: regression in plane takeoff behaviour in 4.6 #28681 -- ready for beta2
• on plane when FENCE_AUTOENABLE=3 and FENCE_ENABLE=0 the fence does not enable when arming if there is a RCn_OPTION set to FENCE. This is a regression from 4.5.x. (Tridge has details) Fix: AC_Fence: fixed 2 regressions from 4.5.x #28729 -- ready for beta2
• on plane the pre-arm check for being inside polygon fence was lost since 4.5.x. Fix: AC_Fence: fixed 2 regressions from 4.5.x #28729 -- ready for beta2
• Navio2 GPS not working (discussion, conflicting report that it does work) -- fixed by refreshing/resetting firmware
• arming at mid-stick question (discuss) -- no issue
• Crossfire receiver does not reconnect after failsafe (issue, PR) -- resolved for -beta2
• Copter AutoTune can produce very low maximum acceleration values (issue, PR) -- resolved for -beta2
• memory fragmentation and multiple memory regions mean lua scripting can fail with out of memory error even if you have plenty of free memory. Addressed in AP_Scripting: allow expansion of memory for scripting at runtime #26517
• system time may be unstable when using only first GPS timestamp, consider AP_RTC: allow time to shift forward when disarmed #24022 or similar fix (Tridge has details)
• Plane fences ignored (issue) -- cannot reproduce
• Plane: auto-deploy landing gear on fixed wing (PR)
• guided error with terrain target Plane: fixed terrain alt target guided #29602
• QuadPlane fast attitude recovery issue (PR)
• disarm in fwd flight on all motor tailsitter #29555

[amilcarlucas]

Why not copy the open issues from 4.5 into this page to make sure they are not forgotten?

[rmackay9]

@amilcarlucas,

We could but I worry that the result will be that we have two really long lists of issues that we don't actually clear up. If we can get focus from the dev team on investigating the issues and resolving them then that could work. We should discuss at the next dev call I think

[tridge]

@rmackay9 @CraigElder I think we need to restructure the dev calls. By the time we get to the copter/plane issues section we are all exhausted (I know I am). I think we should start with the release issues, and look at PRs that address those issues. PRs for new features or cleanups should come after we've looked at critical issues

[Ryanf55]

Here's a new one. I'm not sure whether you want me to add it to the list:
#28666

[rmackay9]

@Ryanf55,
Thanks for that. Tridge has put an item on the list that involves the auto-enable parameter. Maybe it's the same? In any case, I've added this to the list as well under confirmed issues.

[menschel]

@rmackay9
In case you want to list it separately for plane.
Issue Takeoff not completing under TKOFF_OPTIONS[0]=0 #28685
PR attempting to fix it Plane: Fix inability to climb the last few meters of takeoff

[rmackay9]

@menschel thanks for that, added!

[menschel]

I have created another issue.

Plane 4.6 beta - Takeoff throttle behavior for planes without airspeed sensor changed from 4.5.7

and the corresponding PR
Plane: Fix level-off and throttle without airspeed sensor

[rmackay9]

Hi @menschel, thanks, added to the list!

[MattKear]

There is a missing battery monitor option in the param meta data that will need to be brought across. PR here:
#29017

[timtuxworth]

This: Holybro M10 waiting to configure GPS1 (discussion, #29320) -- resolved for -beta5

This fixed a problem I was seeing with the QioTek M10S GPS on master.

[mikedkelly]

When I arm my plane the status changes to DCM Active and it throws continual Bad AHRS errors. As soon as I disarm it changes back to EKF3 Active.

[Ryanf55]

When I arm my plane the status changes to DCM Active and it throws continual Bad AHRS errors. As soon as I disarm it changes back to EKF3 Active.

Can you set LOG_DISARMED to 1 and get logs of it? Have you tried to reproduce on 4.5?

[mikedkelly]

Here is the log and my params. I confirmed this happens on 4.5 and 4.6 beta 4. I guess I didn't notice it before.
36 3-10-2025 10-31-44 AM.zip

[rmackay9]

Hi @mikedkelly, @Ryanf55,

It's tough to do support on this issue so I've created a new discussion here https://discuss.ardupilot.org/t/mikedkellys-report-on-dcm-active-and-bad-ahrs/131592

[timtuxworth]

There is definitely a problem with the M10S @tridge . I had a plane that was struggling to get satellites on it's M10S, so I replaced it with a brand new one out of the box. After 2 hours after booting with master no (zero satellites), I flashed 4.5.7 and it immediately came up with 11 satellites.

[andyp1per]

There is definitely a problem with the M10S @tridge . I had a plane that was struggling to get satellites on it's M10S, so I replaced it with a brand new one out of the box. After 2 hours after booting with master no (zero satellites), I flashed 4.5.7 and it immediately came up with 11 satellites.

I tend to agree - seeing issues on an M10 on my Pavo20.

[timtuxworth]

It seems that a recent change (or changes) might have resolved this issue. If I build AP from the latest master an M10S GPS that comes up with zero satellites on 4.6 will now come up with satellites. This seems to involve two commits (or at least commits somewhere near these):

This commit gets me from zero satellites to about 12
22a70a9 GCS_MAVLink: avoid use of OwnPtry during DeviceOp read/write

This commit (or something just before it) gets me to about 25 satellites
e431b1e Plane: ATT and RATE logging in sysid log at full rate

I found this by a laborious process of building 18 versions of AP for the QioTekZealot H743 and trying them one by one.

[timtuxworth]

I tend to agree - seeing issues on an M10 on my Pavo20.

Try building the latest master Andy (more recent than 23rd March) - do you still see the problem?

[rmackay9]

Hi @andyp1per @timtuxworth,

This issue probably isn't the best place to discuss this. Can we move the discussion to this new topic?