Comment by Emile L. on January 18, 2013 at 6:23am

Great gimbal control!

I have had similar spontaneous response when in loiter with APM 2.0. Log review showed a GPS change to the mid Atlantic from western Canada. It returned within a couple of seconds but took the quad for a little ride.

Emile L.

Developer Comment by Randy on January 18, 2013 at 8:13am

I'll have a look at your logs if you post them here.

As Emile says, your gimbal was working well at least!

Perhaps we need some more checks around ridiculous positions from the gps...

Distributor Comment by Dany Thivierge on January 18, 2013 at 9:10am

Hi Randy, I've seen this as well, exactly as Emile documented... but I was not flying, just connected to Mission planner testing the products before shipping to customer.   I found out the GPS (Ublox) was not "clean" so I had to reload 3DR settings on that GPS and it was fine.  

Discarting GPS extreme changes like that would be nice... maybe you can trigger a mode where GPS is not needed it would help... but I can see some issues as well as there is no feedback sent back to the Pilot.... unless connected to MP on a ground station... Anyhow, avoiding a crash on bad GPS info would be nice, up to the devs guys to figure out the best actions to put once the GPS is "disabled until data returns more normal".... 

Dany

Comment by Alejandro Martin Pirola on January 18, 2013 at 9:16am

Glad you didn't lost your work in the crash!
By the way, nice spot, beautiful view.

Comment by Quadguy on January 18, 2013 at 10:31am

I crashed mine because of loose battery. It got slipped out of its hold quad was performing Auto profile. It got dis-balance and crashed, nose dive negative pitch 30 degree

Comment by Crashpilot1000 on January 18, 2013 at 10:46am

Beautiful flight (except for the end)

Did you try to switch back to stab. mode during the bank right? What Hardware/FW were you using?

Was the gimbal controlled by apm?

Cheers

Kraut Rob

3D Robotics Comment by Joshua Ott on January 18, 2013 at 2:24pm

Thanks for the comments and willingness to help me diagnose this : )

Sorry it took so long to get the log uploaded, I had to re-post the blog to get access to the "upload" function. 

APM 2.0

AC 2.8.1

Gimbal controlled by APM, 20A v-reg on a separate battery powers input and output with no jp-1, this battery had plenty of juice left in it.

1208 MHz video Tx on it's own battery, with low-pass filter. not the first flight with this powered up, but the farthest from home position with it powered.

Once I noticed the misbehavior I centered the sticks and then switched in to stabilize but it was on the ground before the mode could change. The log shows no change in ch5 at the end.

The noise of the still spinning motors helped me find it quickly. There was no response to transmitter(Futaba T7C) inputs after the crash, I had to pull the motor power to stop them.

The rattling sound you hear during the death plunge was from some slop in the tilt pivot transmitting to the gp housing, it was not new to this flight.

I had 11 sats and 3dfix, My cell phone was in the car on airplane mode, the spot I was flying has no RF noise sources that I'm aware of.

Thanks again for any input!

I remain a loyal purple flight control fan boy 'till the end, can't wait to get 2.9 up in the air.

3D Robotics Comment by Joshua Ott on January 18, 2013 at 2:30pm

Forgot to add-

I'm using 3DR uBlox LEA-6 with a toroid on the cable.

3D Robotics Comment by Joshua Ott on January 18, 2013 at 2:32pm

Crash happens at 96% point on the log.

Comment by Greg Fletcher on January 18, 2013 at 4:40pm

You had your gimbal gains dialed in nicely. Don't know about the crash. I just worked on the gimbal/mount code.

I do think it may be gps related. Switching sats causing an instant change in velocity can mess with the dcm.

Comment

• ‹ Previous
• 1
• 2
• 3
• 4
• Next ›
• Page