Plane 4.3 stable release

Do you know how long this bug has been present in Plane?

This would explain couple crashes i have seen.

the home reset bug? I think since at least 4.0.

If you have the logs I could check. It will only affect takeoff if you have a Q_OPTIONS bit set to honor reference frame in VTOL takeoff. By default it always uses the current height as the takeoff start.

@Andrew_Tridgell
i had same gps problem with SKystars h743hd boards same as @makeflyeasy but yesterday i flew Arduplane 4.3.3 without problem with Speedybee F405 V3 board. Nothing happened negative.

1 Like


I’ve just released plane 4.3.4 stable. Changes since 4.3.3 are:

  • support CubeOrangePlus BG edition
  • enable VTX power on MambaH743v4
  • probe external compasses on Foxeer Reaper F745
  • fixed home update on bad GPS quality
  • fixed GPS unconfigured error for non-M10 uBlox GPS
  • don’t allow RC protocol change on IOMCU once detected
  • fixed FBWA pitch limits when in VTOL qassist
  • fixed handling of zero compass diagonals
  • added an output buffer to MAVCAN
  • set emergency status in OpenDroneID on crash or chute deploy
  • avoid logging duplicate format messages
  • fixed bug in alt error arming check with BARO_FIELD_ELEV set
  • fixed handling of double IOMCU reset and safety disable
  • enable VTX power on MambaF405 2022
  • disable PWMSource feature in lua scripts (will be back in 4.4.x)
  • fixed throttle wait on rudder arming in quadplanes
  • fixed earth frame accel compensation for AHRS_TRIM

All users are recommended to update to this release.
Happy flying!

8 Likes

Hi Tridge, Updated to 4.3.4 yesterday - however today at the field I was unable to arm my plane…usually I do this via the switch on the M8N GPS/compass unit but that method it no longer seems to work. I also tried arming from MP actions panel but it reported plane as armed and I wasn’t able to disarm and re-arm from there either. Tried rebooting FC (Pixhawk 6C) a couple of time - no joy. Any help much appreciated - parameter file here http://richardjcox.org/temp/params040323.params

Tridge, It’s all good now…I have set up to use right rudder to arm…now the M89 unit
switch arms all but throttle when GPS, compass fixes OK and right rudder on Tx
arms motor (following a radio recalibrate). Thanks again for all your development work.

1 Like

@tridge
I ran into a problem with ArduPlane 4.3.2 today, I haven’t tested the latest 4.3.4 firmware yet.

question:
When I planned the waypoint, wrote it into the flight controller, and the ground station switched to automatic mode, the aileron and tail rudder turned to a certain angle, which was very strange.
After unlocking and taking off, the rudder returned to the center and returned to normal. I saw this output information in the log. My RCouts 2 and 4 are the tail and 1 and 5 are the ailerons.

Below is the log and video:


https://drive.google.com/file/d/13_YgHo1yAzoAjy-HGZ3n0i4Af73OcV2O/view?usp=share_link

https://drive.google.com/file/d/13VvUl6aASTYtsDHYLyWwlqa_sN6oWGsL/view?usp=share_link

On the ground this movement of the control surfaces is a normal behavior. Only in flight Ardupilot can execute the correct deflections of the control surfaces

@jreise-d
I have been using version 4.0.9 before. In AUTO mode, the state of the aileron and tail rudder remains in the neutral position before takeoff, which is normal.

@makeflyeasy this is very strange - the fixed wing desired roll and pitch rates are rising fast while waiting for arming in AUTO, but the desired roll and pitch angles are steady at the current angles:


I don’t yet understand how that happens, and I can’t yet reproduce it in SITL. Something very odd is happening. I’ll think about it and get back to you when I understand what is happening.
@iampete any ideas? Note that there are no PSC logs and no PIQ logs, which should means VTOL pos controller not running.

1 Like

@tridge
I use ArduPlane V4.4.0-dev; when the h16 remote control communication is interrupted, it will not recognize the sbus signal; only I restart the FC to recognize the SBUS; it is quite dangerous.

No, I don’t see anything. The desired angles could be lying to us, the show_vtol_view function pics which attitude frame we log, so we could logging the VTOL angles but not using.

It is a suspiciously straight line to be coming of a gain, more likely to be a constrain somewhere.

All of you guys that work on these bugs are FUCKIN AWESOME.

1 Like


I’ve just released plane 4.3.5beta1.

This release has a single bug fix for a critical bug for anyone using bi-directional DShot on their vehicle. If using bi-directional DShot and the vehicle is running its motors when 32 bit microsecond time wraps at 71 minutes (or multiples of 71 minutes) then the bug that is fixed in this release has an approximately 1 in 3 chance of causing the motor to stop.

Note that the time is the time since boot, not the time since arming, so even vehicles flying for a short time could be vulnerable if they sit for a long time on the ground before takeoff.

Testers welcome! I expect to release the stable of 4.3.5 in the next few days.

6 Likes

Thanks for the quick fix.

1 Like


I’ve just released plane 4.3.5 stable.

This release has a single bug fix for a critical bug for anyone using bi-directional DShot on their vehicle. If using bi-directional DShot and the vehicle is running its motors when 32 bit microsecond time wraps at 71 minutes (or multiples of 71 minutes) then the bug that is fixed in this release has an approximately 1 in 3 chance of causing the motor to stop.

Note that the time is the time since boot, not the time since arming, so even vehicles flying for a short time could be vulnerable if they sit for a long time on the ground before takeoff.

This bug only affected STM32H7 based flight controllers, and only if bi-directional DShot was in use.

2 Likes

Where is SCR_ENABLE ?
I also don’t see APM>SCRIPTS folder

what flight controller are you using? Scripting is not enabled on boards that are low on flash or ram

Ahhh ok, I was testing a low on ram FC

Thanks

AP 4.3.5, PH4mini. At boot, I get this message:

`ERR Airspeed 1 not initialized, cannot call`

The aircraft has no AS sensor. I looked at ARSPD_* parameters and noticed ARSPD_TYPE=1. I set ARSPD_TYPE=0 and now I’m getting a different message:

`WRN No airspeed sensor present or enabled`

For reference, here are all my ARSPD_* parameters:

ARSPD_AUTOCAL,0
ARSPD_BUS,1
ARSPD_DEVID,0
ARSPD_FBW_MAX,22
ARSPD_FBW_MIN,9
ARSPD_OFFSET,0
ARSPD_OPTIONS,11
ARSPD_PIN,15
ARSPD_PRIMARY,0
ARSPD_PSI_RANGE,1
ARSPD_RATIO,1.9936
ARSPD_SKIP_CAL,0
ARSPD_TUBE_ORDER,2
ARSPD_TYPE,0
ARSPD_USE,0
ARSPD_WIND_GATE,5
ARSPD_WIND_MAX,0
ARSPD_WIND_WARN,0