Pixhawk total power loss during autotune

Ok so after flying my Tarot 680 Pro hex for a while on a kk2 board while I was learning the Pixhawk on my quad. I decided I was ready to switch the Pixhawk over to the Hex. After everything was setup I went and flew it off the stock PIDS. It flew and behaved as it should have. Now after a recharge, I set up auto tune. The auto tune was running fine for just over 4 minutes when all motors just quit and she fell straight down from about 15 feet. Not too much damage occured. Bent motor mount, broken battery tray mount. The battery took the tray with it right off the frame (16,000mAh 4S) and unplugged it’self, I think… Upon checking all solder joints, power connections, I came up empty handed. Now I really can’t see the EC5 battery connector coming loose. It’s a snug fit. But I’m out of ideas. I’m using the 3DR power modual that comes with the Pixhawk. I also only powered the receiver off the output side of the PX4. All ESC’s are OPTO’s. Anywho, if any other info is needed ask and I’ll post it. I came up with this from the datalog file. Looks like a total loss of power.

Tarot 680 Pro extended to 810mm
Pixhawk, powered by 3DR Modual
Futaba 14SG, R7008SB Rx
Sunnysky v3508 580KV
AFRO 30A OPTO ESC’s flashed with Blheli
Multistar 16,000 4S

Could you please post the logfile?

Some people have seen this already as a motor sync issue. It happens on large copters with pancake style low rpm motors. What happens is that the auto tune changes the motor speeds so fast that the motors can’t keep up and just frees.

Mike

Right! Here is the log. Upon getting the auto analysis to work, I see that there is a brownout fail at roughly 21 feet which is about the height the hex dropped dead from.

dropbox.com/s/fxwbonxja9ej8 … 2.bin?dl=0

[quote=“iseries”]Some people have seen this already as a motor sync issue. It happens on large copters with pancake style low rpm motors. What happens is that the auto tune changes the motor speeds so fast that the motors can’t keep up and just frees.

Mike[/quote]

I certainly don’t have any sync issues. The AFRO ESC’s with SimonK had them but after flashing to Blheli, all issues vanished. I have battle tested the Tarot 680 with a KK2 board for over a week. And man, with a 4S 5200mAh pack with no gimbal gear, that thing flys like a bat out of hell! Besides, all six motors lossing sync at the same time is probably worse then jackpotting the PowerBall in terms of odds.

Tkopper,

You’ve described what I think that happend: total power loss. How did you power px4? only through 3dr power module?

I have a very similar 680 pro. The difference is that I am running quanum v3508-700 and 2 packs of mulitstar 10ah.

I had a crash a few monts ago with one of my quads due to failure in one multistar 5.2ah-4s. I was flying and then, suddenly , without any warning, total power loss. My quad felt from 27m to the ground, destroying my gimbal.

Further analysis revealed that the culprit was a multistar battery. One cell, besides apparently normal, under medium amperage ( around 25A ) suddenly drops the voltage to 0.5-1v. Without draw it remains “normal”. Not easy to detect until it happens. That’s why now I fly with two packs in parallel…

Try to put your pack under stress… maybe could be the very same problem… maybe not.

PS. your z vibration level is a little bit high. See if there is some play in your motor bell.

OK I never thought about cell failure. I will definitely monitor each cells voltage under heavy load. Yes I was only powering the Pixhawk through the power module. Which I now know only supports 60AMP current sensing with the Pixhawk and 90AMPs max. Attopilot 180 is on the way! Now I believe my hexacopter is capable of pulling close to 120AMPS at full power. Do you think that could have causd the module to fail temporarly? Also the dual pack setup is genius and will be applied before next takeoff. Voltage redundancy!

Now the only question is two 8,000 packs or two 10,000? Thats a lot of battery lol. My goal is longevity while lifting a GoPro and gimbal. Thanks for your help tabascoz.

Well the battery seems to be fine. Cell voltages all remained relatively the same under medium and heavy loads. My only thoughts are with the power module now. Which seems to run fine on my little quad.

Weird… is px4 the very same device on the quad? Could it be something loose?

You could power the px4 from a bec through the output line in order to have power redundancy. This way you could track down any voltage spike or brownout by analysing both voltages log graph.

Yea its the same unit. I just did a 22 minute flight with it on my qaud. No problems what so ever. Is it a must that I use a zener diode when having a secondary power source at the servo rails? I use the Multistar SBEC 5 and 12v outputs.

I had the same trouble, see posthttp://ardupilot.com/forum/viewtopic.php?f=111&t=14433

Solved by setting up triple power to the Pixhawk.

http://copter.ardupilot.com/wiki/common-powering-the-pixhawk/

Yea I’ve been running the Pixhawk on my qaud with a backup power supply running to the servo rail via a 5v SBEC. Don’t have a zener diode on there yet but the HK SBEC seems to regulate very well though. Today I’m going to take the PDB plate off and flex it around a bit while powdered up to see if maybe there are bad trace’s.