Warning #1: PX4/Pixhawk users upgrading from AC3.1.5 (or earlier) may need to re-do their compass and accelerometer calibration because AC3.2 also uses the backup compass and accels.  Pre-arm checks have been added to ensure this has been done.

Warning #2: on the APM2.x the logs must be downloaded using MAVlink instead of the terminal.

AC3.2-rc14 is now available for BetaTesters through the mission planner’s Beta Firmwares link.  The full release notes can be found in ReleaseNotes.txt and changes from -rc13 can be seen below.

     Feel free to raise issues found during testing on this discussion or in the new support section in the APM Forum.

     It’s a big release with “the onion” restructure and a bunch of new features (including these 57 closed items) so we need to re-test almost everything including all flight modes, all mission commands and all the new features.  Marco and I will be maintaining (and adding to) this testing list.  Issues reported will first be checked by Jonathan, Marco and I and then confirmed bugs/issues will be put on the github issues list (and then hopefully fixed).

     Thanks especially to the beta testers who put their copters at risk testing each release.  Enjoy!

Changes from 3.2-rc13
1) Safety Features:
     a) fail to arm if second gyro calibration fails (can be disabled with ARMING_CHECK)
2) Bug fixes:
    a) DCM-check to require one continuous second of bad heading before triggering LAND
    b) I2C bug that could lead to Pixhawk freezing up if I2C bus is noisy
    c) reset DCM and EKF gyro bias estimates after gyro calibration (DCM heading could drift after takeoff due to sudden change in gyro values)
    d) use primary GPS for LED status (instead of always using first GPS)

Views: 305492

Reply to This

Replies to This Discussion

Thank you for this, Randy.

I flew this "latest" one: 22-May-2014 12:53. (didn't know the difference between Beta and Latest.)

Stabilize: ok

Alt hold: ok

Loiter: move around 70cm to 1 meter with 11 sats and hdop 1.7 in moderate wind.

Improvements in altitude control while circling slowly to do a camera pan on 360 (loiter mode)

Loiter: Perfect in no wind condition.

Hybrid: same (Just guesstimate) in no wind but fun to move around! Loved the breaking scheme!! Meaning: going in one direction fast and releasing the sticks! Great!!

RTL: no problem (with altitude set to 0 to force auto land)

Land: No problem

Simple mode: Tested with Alt hold, Loiter and Hybrid. Great! nothing to report (except it works).

Hi Randy, i want to test the "latest" AC3.2-rc1(

Date:   Thu May 22 21:13:48 2014 +0900).  Does this"latest version" already includes the fix for "unexpected roll/pitch when switch to hybrid"?.  Thanks so much.


     Thanks for the feedback.


     Yes it does include the Hybrid initialisation fix.  That bug was my mistake but Julien D (the creator of Hybrid) found the issue and fixed it.  The joy's of open-source cooperation!

     AC3.2-rc2 will include a couple of extra things beyond what is in master.  In particular we will enable the Parachute deployment and "Rally Points" from Andrew Chapman but only for the Pixhawk.  I'll see if I can get some wiki pages done for those too.

Sorry to be here again but regarding version: 22-May-2014 12:53 and failsafe

FrSky  receiver failsafe set to RTL and throttle output value to 920 (Obviously < 975)

APM set to RTL on throttle lower than 975

APM Rssi: set to ignore in APM

APM GPS failsafe: set to land

Then on switching off Tx while in loiter and banking left seems that land mode was engaged as planned. 

But as descending slowly towards the ground (as programmed in APM 30 cm/s) did continue to move horizontally until touch down on the ground. no damage but flip upside down.

Loaded the "latest" AC3.2-rc1(Date:  Thu May 22 21:13:48 +0900).  Just came back from the field and......

a) first flight:  Stabilize mode and Alt Hold -  It was quite windy but the copter was flying smoothly and nicely with predictable attitude. Switching from Stab to Alt Hold and back is seamless with no change in the behavior of the copter. 

b.  Second flight: Stabilize/Alt Hold/Hybrid - it is confirmed that the "unexpected roll/pitch when switching to Hybrid" has been corrected. When switching from Stab to Alt Hold, then Hybrid, the copter remains stable with no stick input...it has rock solid hover! Forward and backward flight were straight and the braking behavior is nice. Although, maybe because of the wind, there's a tendency to descend when flying sideways. Using hybrid, the copter did figure eight flight like an airplane at 40 meters height - great for fpv flying.

c. Third flight: plotted 4.5 km flight plan with spline waypoints - the copter did a fantastic flight. The turns were smooth with no abrupt pitch, roll and yaw behavior; maintaining its altitude ....well, it is simply amazing! 

Thanks to James for doing the first flight.

To Randy, Julien, Dev Team - You guys are the BEST!


Thanks Randy,

So what your saying is that for the APM the optical flow is as good as its going to get and all future development for the OF is going on the Pixhawk? (but not using the mouse sensor)

If you want to refer to 3.2-rc1, the link I provided is the right one. The other more recent build weren't released as such and are "work in progress", not guaranted to work

Hello Chris,

I decided to participate in this thread because I too have serious concerns about 3DR's approach to managing the intersection between an open source code base, and your own commercial product distribution and documentation.

Upfront, I absolutely say bravo to you for your initiative and success in building this brand. I'm 100% behind your efforts! But to be blunt, I fear you may lack the software engineering background you need to fully appreciate these issues.

While I do think Oliver's frustration is coming through in his posts, and perhaps is a bit out of place in a beta thread, he's quite right that 3DR's approach to commercializing an open source code base is entirely, technically speaking, half-ass.

Please go see, for example, how Barracuda manages its use of the Freeswitch telephony project in its Cudatel commercial offering. You won't find them sending people off to the Freeswitch IRC channel, or suggesting that, actually, it's the community's job to provide in depth documentation of things like you describe in the disclaimer (Important Note) at end of the quick start guide.

So, as much of an improvement as the current "quick start" guide is.... A "quick start" guide is not remotely sufficient in the documentation department, pretty much by definition, and no, I don't agree, at all, that your documentation is 'quite good.'

But beyond documentation, You must create a degree of separation between your published and approved versions of the software, and the underlying trunk. You must have competent developers to make decisions about when to advance features from the open source trunk into the released products, and you must have competent technical writers to maintain the related documentation. I gather you don't, or that you (the manager) fail to assign enough resources to the task.

That's not to say folks shouldn't be able to use other branches at their own risk, but that you need to have a sanctioned, conservative, release that you maintain, thoroughly document (read, accuracy and continuity), and have qualified confidence in.

Yes that means added liability. And yes, that's a necessary cost of your business because not doing so creates its own liability in a critical human-factors product. In short, if you're going to make things that weigh several pounds and fly, and sell them to consumers... You need to up your game in this area.

Finally, and this isn't meant to sound too rhetorical, but didn't you just raise $13M? My advice... Don't get out in front spending that on marketing, because you can't handle a flood of new consumers in this situation. People are going to get hurt, and fair or not, you're going to get blamed. Perhaps not as reckless as flying drones into trees at nuclear facilities, but you get my point.

I hope I don't get overly flamed for this post, and it is just one guy's opinion. The real point is, this an advanced hobby suited for those of us with the engineering perspective (and patience) to engage in and respect the risks involved. If you want to sell to consumers, you need to take more responsibility for educating and protecting them (and bystanders)from themselves and your products. That means documentation and code control. Perhaps your counsel disagrees with me on this.

Yes, it's a big job, and as a commercial venture promoting it for profit, you need to own it and not pass the buck to the community of uncompensated developers making 3DR's business possible. I read what you said about donating to the community, and that's laudable. But it doesn't cut it.

Thanks for listening, and I hope you receive this as the positive criticism it's meant to be!

Best Regards,

Erich Greenebaum

Small issue with AC3.2-rc1: I have some pitch twitching whenever I go to RTL. Not much, just short, seemingly periodical twitches. Only pitch, not roll.

In the attached log you can see that both pitch and desired pitch twitched in RTL. GPS HDOP was fine, no fluctuation there. 

Also, the copter doesn't seem quite as stable as with 3.1.x. On straight descents, it's really noticeable. And even if there's just a slight wind, the copter is not perfectly horizontal in stabilize mode.

I had to re-autotune and finetune with 3.2 however, and now the pitch PIDs are quite a bit higher than before, and now it's almost a 3:1 ratio between pitch and roll whereas previously it was 2:1. So maybe the problem isn't 3.2-rc1 after all - but I have no clue.

3.2-rc2 is now strait from the beta download. Though don't know were "motor sync test" disappear?

Will test the rc2 shortly.


    that's actually AC 3.1.5-rc2 in the beta firmwares area so it's quite a step back.  testing is appreciated but none of the AC3.2 features will be there.


     Unfortunately I'm getting a 'permission denied' error when trying to download your logs..

Reply to Discussion


© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service