Replies to This Discussion

AdminPermalink Reply by John C. on December 8, 2010 at 6:23pm

Good question.

I took the XBee off, and added the satellite RX back onto the main RX.

It seems like these actions minimized it a bit more, but still did NOT eliminate the hitch.

I also read this post, where it seems like this is, or is becoming, a fairly common problem.

http://diydrones.com/forum/topics/arducopter-first-tests-and?xg_sou...

I re-inspected all of my connections, and #5 and #6 input were flipped, but that doesn't impact the motors. Everything else looked good.

I'll see if I can get it to do it while I'm holding it.

• ▶ Reply

Permalink Reply by Chris Walsh on December 13, 2010 at 5:59am

check your ESC's... 

 

in our project, the escs would often be a little temperamental and glitch up... we gave them a flick and made sure they were properly seated and it worked fine after that. But this would happen a couple of times....

 

if that doesnt work then good luck!!

• ▶ Reply

AdminPermalink Reply by John C. on December 8, 2010 at 7:02pm

Could this be a function of not enough solder on the main traces of the power distribution board?

• ▶ Reply

AdminPermalink Reply by John C. on December 8, 2010 at 9:28pm

I had two more flights tonight that were a bit more successful.

PID gains back to stock.

I turned my magnetometer so the components were down, pins forward.

I also set the declination to a reasonable estimate per this map:

http://www.sadoun.com/Sat/Installation/STAB-Installation-Angles.htm...

I smashed a couple more connectors together with pliers.

Routed the main RX antenna up through the protector, sticking up.

Satellite Rx attached.

NO XBee yet.

Still had two quicker glitches on both batteries. Couldn't get it to do it in my hands.

I'll give her another two batteries, and then try to attach the XBee again.

• ▶ Reply

Permalink Reply by Sebastian Gralla on December 9, 2010 at 8:33am

yes
check the motor connections (I can't imagine any other reason)
resolder them if necessary

• ▶ Reply

AdminPermalink Reply by John C. on December 9, 2010 at 12:02pm

I re-soldered the back of the PDB..

From this:

To This:

STILL getting random dropouts. Seem to be occurring less frequently, and less catastrophically, but it's still there. All of the motor wire connections seem nice and tight. I guess I can go through and cut them all open and resolder them all.

• ▶ Reply

Permalink Reply by Richard Gosling on December 10, 2010 at 1:36am

I have had 2 crashes now.

The first i put down to a bad choice in Rx.

I've just had a potentially catastrophic crash hovering 3 metres in front of my face which flipped the quad into the concrete. Luckily it had done a full loop before biting the dust and without much inspecting can report that the injuries appear to be superficial - that's 2 from 2 get out of jail free cards.

 

My advice is to not bother flying your Arducopter anywhere for the time being.

 

There is a real risk of terminal damage, not to mention the risk of damage to anyone within a radius of who knows what from where the Arducopter shat itself.

 

The glitches other users have reported are real.

 

Until the situation is resolved by someone who can pin the issue down to a specific source, all Arducopters in this neck of the woods are permanently grounded.

 

Regards, from a very disappointed  Rich

 

• ▶ Reply

Permalink Reply by Riccardo Kuebler on December 10, 2010 at 4:15am

Rich,

 

it seems I'm fighting the same problem.

Doug Barnett pointed me toward a good solution : change all the Deans connectors (as a starting point).

Best regards,

 

Ric

• ▶ Reply

Permalink Reply by Riccardo Kuebler on December 10, 2010 at 3:56am

John,

 

if it was a trace problem you should have voltage drop on all four motors, not glitches. If drop was important you should experience an ESC reboot (APM reboot too ?).

I don't think trace are a real problem, seing the (small) traces that are on some current sensors with 100A capabilities. Those are definitely not very big.

On Arducopter power board there is a weak point in Gnd trace, where they join in the rear side. This seems to me the weakest point of all the board.

 

Ric

• ▶ Reply

Permalink Reply by Sebastian Gralla on December 10, 2010 at 8:37am

you mean the rear side is not as good as the front side?

• ▶ Reply

Permalink Reply by Magellan on December 10, 2010 at 9:46am

I think I remember reading in the ESC Spec Sheet, that the ESC will drop to half RPM in some situations (low current, high heat, signal loss, or low voltage/battery, etc) It seems reasonable that one isn't getting enough power and is dropping to half RPM. I also remember reading that it would drop to half RPM and then after a delay, would drop to 0.

Could that be it?

I think Im going to go add a ton of solder to mine now.

• ▶ Reply

Permalink Reply by Richard Gosling on December 10, 2010 at 10:23am

The ESC's will drop power when at 110 degrees Celsius. It will also reduce output power if throttle signal lost for 1 second - further loss of 2 seconds will result in output being cut completely.

 

After moving my ESC bullet connectors during operation (with no props on) i discovered that certain motors would glitch depending on which connectors i moved.

 

I have since removed all bullet connectors and have soldered the connections directly to each other. I then did the same test and the problem was no longer there. I then ran the quad for 10 minutes (still no props) and did not see one glitch. I did see that the right ESC was noticeably hotter than the rest, but still only 45 degrees C or so.

 

Fingers crossed that the glitching is gone forever.

 

Rich

• ▶ Reply

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