Copter-4.3.5 released!

I think you mean 6.19 - but should be fine.

All the same settings from 4.2.3 should work on 4.3.5 correct. Same wiring ? I am using the nano diversity. With tbs unify evo and crossfire micro tx. Is there any specific options I need to change in order for crsf to work with

Should be fine. Make sure you turn the TX on first. Maybe try attaching the TX to agentX and check the CRSF outputs are correct

I have a couple craft with similar hardware and it’s working fine.

  1. Micro Tx, Nano diversity Rx, no Vtx
  2. Micro Tx, Nano Rx Pro, Unify Pro32
1 Like

Would you mind showing how your tx is configured in MP?

Thanks Yuri, I’ve updated the post with new logs (also 4.4 dev where OA worked).

Maybe @rmackay9 and @rishabsingh3003 can help us fix this?

Is it possible to build 4.3.5 for Orange Cube+ w/bdshot?

It needs a hwdef, but it’s possible

Ok thanks! I will start learning the build process.

Is there any more documentation/posts on this bug? I had a very mysterious crash on a large quad that had the FFT on for dynamic notch. This could possibly explain that incident. The logs of the crash didn’t seem to provide any explanation.

The bug is theoretical - I have only ever seen it in SITL. If it occurred in the wild then it would manifest as a watchdog - i.e. an abrupt stop to logging + crash

1 Like

Yesterday’s testing: 250 g. I was in Loitor with a nice hover and 3D lock, then I induced a RTH. Quad will climb as normal, dropped the stick to zero and switched back to loiter… Quad will drop out of the sky and crash. Not stating this is a bug but I was wondering if we could look into this action. Not 100 sure what to do to deal with this user error if anything.

1 Like

Hi @Quadzilla,

Thanks for the report. This sounds very much like the auto-armed flag issues we use to have many years ago where the vehicle state machine would think that the pilot had never raised the throttle and would then not try to stabilize the vehicle. I thought we had resolved all these issues though.

I tried to reproduce this issue in SITL with both master and 4.3.5 but I wasn’t able to. I did this:

  1. took off in Loiter and flew to 100m
  2. moved sticks to neutral positions
  3. RTL
  4. moved throttle stick to zero (e.g. rc 3 1000)
  5. Loiter

At this point the vehicle maintained attitude control and descended at PILOT_SPEED_DN (0 by default which means same as PILOT_SPEED_UP = 2.5m/s). By raising the throttle I was able to bring the vehicle to a stop.

If you have an onboard log I’m happy to look into it further. My guess is that the vehicle momentarily went into stabilize mode with zero throttle or just that it descended at 2.5m/s which is quite fast.

1 Like

No worries, I am installing the old FC and parts in a new printed frame and will pull the log as soon as i can.

  1. Took off in Stabilize.
    2 . moved to Loiter with good lock.
  2. Hit the RTH switch by mistake. copter started to raise.
  3. moved stick to zero to stop ascent.
  4. switched back to Loitor.
  5. craft fell like a rock into a pile of parts.

I am ok with user error reducing the throttle to zero was dumb…

I will get the log for you, it is possible I did switch to Stabilize it is inline and it happened fast. :slight_smile:

1 Like

@Quadzilla,

OK. We could potentially add a 1 second delay to the auto-armed state machine so that it doesn’t immediately go false in stabilize mode with zero throttle. We actually already have a delay on the “throttle zero” flag… but in this case, because the throttle was at zero for a long time it would have become “true”.

BTW, I hardly use stabilize anymore. It use to be important because of the potential danger of “climb aways” due to heavy vibration but this probablem has been mostly solved by the vibration failsafe.

1 Like

i have very low vibrations well under the chart consistently with a large variety of frames I never trigger that.

For me this has always happened in GPS mode without a Sat lock and changing out of GPS would recover now that error is gone and defaults to stab. Perfect.

The Pixhawk pro is fantastic I was a beta tester. I only mount with screws or tape I think why tuning is not needed a lot with this FC and other mRo products. Stabilize is always used as my main flight mode for years now. I find Loiter mode a little dull. :slight_smile: The key for a good frame is to have Stabilize hang in the air with little input from the sticks. I get a lot of good feedback this way in flight. Not unlike the old betaflight days when we had no GPS.
Once stabilize is working well and 90% of the time it is without the need to tune. I move to Loitor and test.

1 Like

Dropbox - 2023-03-20 15-42-40.bin - Simplify your life Here you go I not looked at it yet…

Edit: just looked at the bin at it i see i engaged Stabilize. user error.

Of interest when I first start up I do get a odd vibration that goes away. I seen this before on a few crafts. I believe it’s a prop issue or battery power. turning down the PIDs more should remove it.

New body today old parts. We getting a ton of rain so i removed the holes

At low throttle that’s called RRTE mode.
Rapid Return To Earth.

2 Likes