Comment by Ellison Chan on May 18, 2012 at 9:35am

Jason, yeah that sound about right.  The problem with that behaviour is that you've just lost the bottom half of throttle control.  For high powered drones, like hexas and octos, that doesn't leave much control of the throttle left.  With AC, I have enough control to hover it inches above the ground.  This will never be possible with the Naza's behaviour.

Comment by Tim - Arduino for Visual Studio on May 18, 2012 at 9:37am

John. I don't think it should get wrapped up in the global world of diydrones and 3dr because that will not provide clarity and confuse things further. Maybe on rc groups which is a little less specific

Developer Comment by John Arne Birkeland on May 18, 2012 at 9:46am

@Ellison: On the WKM half throttle in Att. mode means "hold this height", not 50% throttle. So if you raise the stick to 60% the motors will slowly increase throttle until you have an altitude lift speed that is 20% of 6ms/sec total I think it was. And then you just center the stick again when you want the copter to stay put at the current altitude (inches of the ground if you want). For security, you usually punch the stick to quickly get some ground clearance.

Comment by Ellison Chan on May 18, 2012 at 9:51am

John, I don't think sensor latency is the issue with AC stability.  If you saw my blog about the AeroQuad that's Patrice was flying, you can see how well a 2560 with basically the same set of sensors can control a quad. 

Comment by Ellison Chan on May 18, 2012 at 10:00am

I think I need to fly it some more before I comment further on the Naza behaviour, lest I start introducing false impressions.

Developer Comment by R_Lefebvre on May 18, 2012 at 10:07am

@Ellison: are you guys saying that the Naza if like a Fly By Wire type setup, where stick inputs result in a rate of movement, rather than an angle, or...

Comment by Ellison Chan on May 18, 2012 at 10:11am

Robert, from my short current experience with Naza, and what I'm hearing, it's not quite fly by wire.  My only complaint so far is that throttle position is not a one-to-one map to motor speed.  

Comment by John Wiseman on May 18, 2012 at 11:04am

"The code is quite large, and can be hard to figure out.  And when a new dev jumps on board, it can take a lot of time to get up to speed."

I wish the development mailing list was publicly readable so that I and other interested people could browse and learn from the wisdom of experienced developers, and read the historical context around design decisions.

Comment by Tim - Arduino for Visual Studio on May 18, 2012 at 11:43am

@JohnW I think the problem with that would be that discussions and work on the new arm boards might halt sales of apm2. it has been said that most of the existing code will be re-written :)

Comment by Ellison Chan on May 18, 2012 at 11:56am

The AeroQuad Baloo (STM32 F4) board is being tested right now.  I suspect it could be in wide circulation by the late summer.  Check out this thread, if you want to follow the progress.

Comment

• ‹ Previous
• 1
• 2
• 3
• 4
• 5
• 6
• 7
• Next ›
• Page