Developer Comment by R_Lefebvre on June 19, 2012 at 9:31am

Phillip, I have to Google the Whiplash as I hadn't heard of it.  Interesting machine.  Innovative design, I like a lot of things about it.  Direct servo control.  Placement of the avionics, etc.  For a while I thought that it didn't have an autorotation one-way, but I see it's in the motor pinion.  Very interesting idea and I can see the reasons.

Though, I wonder if the autorotation gear is even needed on an electric UAV?  I'm not interested in autorotating, and what are the odds that an electric motor dies in flight anyway?  I find the one-way hub to be a constant source of problems and would love to get rid of it.

And I've been thinking about converting my TRex 600 to a 550, just to get rid of the silly swash linkage.

Now, the tail servo being up front is not something I like.  Seems like it introduces a lot of slop?  I've actually moved my tail servo back right onto the tail, which obviously causes CG issues, but is not a problem for a UAV.

Anyway, any tips for our software? :D  You have an amazing performance!

Comment by Phillip Jones on June 19, 2012 at 10:20am

@R_Lefebvre I'm no expert on R/C helicopters, but I really like the Whiplash. It's a very solid and simple design and we've had no hardware issues with it.

It's really hard for a spectator to know if an electric is autorotating. With a gas or nitro, it's obvious due to the change in engine noise. But for an electric people just wonder why you're flying strangely. Still, it's a good skill to practice. The one-way gear is better (less friction) than forcing the electric motor to turn during an auto.

When people can fly like in this video, I'm not worried about tail slop ;-)

which obviously causes CG issues, but is not a problem for a UAV

If the autopilot is always holding 3/4 forward stick for hover, that doesn't leave much room for control if there's wind. It's also less efficient, so you'll get shorter flights. That might not matter to you if you drop from 10min to 8min on a flight -- but I'm aiming for several hours of flight time, so efficiency matters a lot!

Honestly I'm amazed by what you can do with ArduPilot software already. It's fantastic! I did chat with Randy and Jason after the rodeo and it was fun to compare methodologies. Clearly stuffing all the flight control algorithms into 8k ram is a major challenge! I'm lucky, I've got 32Mb to work with. 

Developer Comment by R_Lefebvre on June 19, 2012 at 12:40pm

You mean holding forward stick to prevent it tipping back due to the tail weight?

I didn't mean to say that the CG doesn't have to be spot-on.  What I meant was that with a UAV type application, we're not worried about fitting a canopy on.  So moving mass around to get the balance is not a problem!

I'm actually flying around with an extra battery pack strapped to the front as ballast until I get the camera mounted up front! 

To me it seems like, since it's almost impossible to hear the electric motor quit until it's too late.  And the autopilot as absolutely no ability to auto in the first place, that simply doing away with the auto gear might make some sense...

Developer Comment by R_Lefebvre on June 19, 2012 at 12:45pm

Oh, and as for that video, I see the comparison to 3D flying all the time, and I think it's really not applicable.  What I am trying to do it get perfect fine tail control.  Currently I get a 2° waggle that I am trying to eliminate.  Nothing you'd see in a 3D video.  It's only visible in a hover, looking at the tail closely.  But it would also really mess up a video, which is why I'm working so hard to eliminate it.  I thought the cause was because of linkage slop, which is why I moved the servo.  But now I still have it, assume the issue is "transport lag" in the PID loop.  Due to the fact that it actually takes time for the servo to move.  I'm not sure how to get rid of it.  I've been looking at a Smith Predictor, but I don't know how to actually implement it.

Comment by Samantha Oborne on June 19, 2012 at 4:29pm

Here are some photos from the day!  Sorry for the delay, I was busy getting married today... Video to follow (of the rodeo!!)

3D Robotics Comment by Alan Sanchez on June 19, 2012 at 5:31pm

Those look awesome! thanks. And congrats on getting married, I saw the name change on your account :) . Any chance you can post the group photos as downloadable?

Comment by Phillip Jones on June 19, 2012 at 6:01pm

@R_Lefebvre -- I see what you mean about CG now. I've never had troubles with the one-way gear, but I've had instances where having it saved a crash. 

I've got a bit of tail waggle, too. For me, I don't think it's mechanical slop. I'm using they yaw gyro as a simple rate dampener, so the tail does wander around a bit. I need to spend some more time adjusting the yaw gains to get better results for filming, but it's 'good enough' for flying right now. I think it will be easier to use a gyro stabilized platform for the camera and let the helicopter waggle a bit. 

I don't think a Smith Predictor will help, since the waggle (at least in my case) is not due to a time delay. It's more related to changes in main rotor torque and possibly wind effects.

@samantha -- great pictures!! thanks for posting. and congrats!

Comment by Samantha Oborne on June 19, 2012 at 6:03pm

Hi Alan thanks :) Didn't realise I hadn't set download permissions, fixed now. If you go to whichever photo you want and click "Actions" above the photo you should be able to view all sizes!

Distributor Comment by Dany Thivierge on June 19, 2012 at 6:59pm

Nice pictures! looked really a fun crew.  Happy people after crashes is always nice to see (but quite rare!) 

Congratulation Samantha and Michael!  I guess you did not take these wedding pictures yourself this time! 

Comment by Troy on June 19, 2012 at 7:56pm

I just wanted to give a quick thanks for the DIYDrones crew for arranging this event. I had a great time meeting and learning from all of the more experienced folks that were there.

Comment

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