This is a discussion re the ArduCopter-3.1 AutoTune feature.  A place for people to report successes and failures, ask questions and get answers.

The AutoTune wiki page is here.


Views: 74302

Reply to This

Replies to This Discussion

Ouch!  that is tough.  I don't think I've ever seen such a clear video of a propeller stopping in mid-air like that.

This whole discussion and the fairly large number of ESC sync errors it's uncovered has certainly put smart ESCs sooner on the roadmap....and maybe we need to add a sync test so people know that the ESCs are not going to be able to keep up.

@Cody O

Sorry for that. Which motor / esc combination did You use?


Couldn't we collect a data base of all combinations (motor/esc) that are known to NOT work properly?




Oh, that sucks!

It looks like you lost the first motor from the autotune stress have been talking about. You then stopped autotune, or at very least you were never horizontal long enough for autotune to initiate another test. And it looked like you were going to be able to save it but the stress of running with 1 motor down caused a second motor to go.

I am not sure how serious you were with your fix request but even if we could sense a motor stall, aborting autotune would not have changed anything since it was flying in stabilise from the moment it lost the motor. Autotune would leave you in stabilise provided you are not within 2 degrees (might be 3) of horizontal without any stick input on roll, pitch, or yaw, for a quarter of a second. So as you can see, autotune had nothing to do with the second motor outage.

I like Randy's suggestion to create some sort of bench test, sort of like compasmott, that is a bit of a stress test for motors and esc's. If we could work out the best way to recreate loss of sink this could be very effective, a bit dangerous maybe....

I have to say I really admire how clearly you were thinking at the time of the crash. Many people wouldn't have been able to take in as much during such a stressful event.

Could that also have happened in the log I posted? The suddenly see a large throttle and yaw output.

Hi BasTijs,

Sorry it took a while to get to you. Going over these logs isn't a quick thing.

This looks like your friend lost a motor but it doesn't look like he is flying a quad. Is it a hex or octo? What ESC's and motors is he using?

We are having a pretty common problem with Simon K ESC's loosing sync during autotune. It looks like the repeated roll/pitch can get too much for the ESC/motor combination and sync is lost.

Thanks for the repy.

He has a quad with Turnigy Sentry 25A and HexTronic DT750 motors (11x4.7 props).

He said that after the crash the APM did not respond and the motors were turning idle. After un/replugging the battery, everything seemed to work again.

That sounds like a great idea.  I've been battling sync/stall problems on this aircraft on all motors from the beginning.

Thanks; I did have a sinking feeling, but this is also the excuse I needed to rebuild from the ground up :-)  I appreciate the compliment; those few seconds felt like minutes when it was happening!

Normally I don't have a problem landing with 5 motors (yes, this has happened plenty of times).  I only suspect that autotune flipped the aircraft when I took my hand off the cyclic to disengage, but I can't prove it.  I didn't see a second motor stall, but the motor in question did hesitate a few seconds before stalling.  From my testing, it seems there is a particular RPM range that this motor/esc combo doesn't like.  It would be nice to know how to test and/or fix a mismatch such as this.  I'll also mention that autotune works beautifully on my quad, which has different speed controllers.

I was serious about the request.  It seems like a simple, common sense fix that shouldn't have any drawbacks.

I really like the idea of having a bench test to check the motor sync, as long as it's reasonably safe.

Motors: Turnigy Multistar 4225-390Kv 9392000007
ESC: TURNIGY Plush 40amp TR_P40A

Cody O.

Since Marco had the same problem with the smartdrone SD-M550G motors one could think that only the pancake motors (resp. motors with many poles) have this problem. Hextronic DT 750 had this problem too, but it has the standard number of poles (12?).

I am running Standard Sunny Sky Motors (2212) with 12 poles and HK F30A ESC or DYS 30A ESC and had no problems with autotune on 3S and 4S. 

I am not a specialist on electronics. Could anyone comment on the idea that the sync issues are due to rpm or number of poles?

I'm with you 100% in wanting to get to the bottom of this.

I don't think there is anything inherently wrong with the pancake high-pole design.  The 610kv version (9392000008) of the motor I listed has performed flawlessly on my 4s quad with various speed controllers tested; I am very pleased with it.  I kid you not... the 390kv version has had a 90% failure rate (!) on my 6s hex, with failures including simple stalls to flames bursting from the coils.  The 390kv motors would also experience the stall problem with the Turnigy Multistar 45 Amp Multi-rotor Brushless 2-6S (OPTO) (9351000009) speed controllers.

So with all the concern about ESC sync issues I’ve added a motorsync test to the cli.  I’ve also made a short video which says how to use it.

The code is only in master but if you’re feeling brave (i.e. use at your own risk) you can pick it up from
             Apm2 quad:
             Pixhawk quad:
To run the test:
1.       Flip your propellers over and rotate one position (similar to compassmot)
2.       Turn on your transmitter with throttle at minimum
3.       Plug in the USB cable and battery
4.       Connect with the mission planner’s terminal screen
5.       Type “test”, “motorsync”
6.       Raise the throttle to the level you want to test at and press “a” + <enter>.  The motor should spin up to the level of your throttle for 2 seconds, then stop
7.       Repeat #6 at higher and higher throttles to see if any sync issues appear
8.       Press any other key to exit the test

Marco and some others have made suggestions that we should allow testing of a single motor and also that we should allow testing the ESC/motors while they're already spinning.  I haven't incorporated those suggestions in yet though.

Reply to Discussion


© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service