HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Dec 29, 2007, 09:45 PM
13brv3's Avatar
Navarre, FL
Joined Mar 2002
3,839 Posts
Discussion
Uavp Fpv

Greetings,

I've been looking forward to trying FPV flight with a UAVP (quadrocopter), and finally got the change to try it. For both flights today, I did the takeoff and landing by looking below the goggles (head tilted way back) at the UAVP directly. Once about 5 ft up, I looked back at the video and flew that way until I was ready to land, or until I was hopelessly lost.

The FPV2 video is the first flight today. As you can see, the effect of the fisheye lens is pretty significant, but still the best for what I'm trying to do. For this flight, I had the compass hold turned off, by accident really. Everything was going pretty well, pointing into about a 8 mph wind, until I got the brilliant idea to try to yaw around a bit.

It didn't take long before I couldn't tell which direction I was pointing, and started to look for the UAVP under the goggles. MAN, was it a distant speck!!! The next moments were spent looking at video, then directly, all while turning to try to figure out which way it was pointing. During this time, I was just drifting downwind farther, and farther, with no hope of recovery.

At the very end of the video you see what looks like it's diving to a fatal crash, but at the last minute, I just let go of everything and goosed the throttle. Much to it's credit, the UAVP used the extra power to level and pull out just a couple feet from the ground. You can just see that it's still alive when the video ends. Best I can figure, the DVR automatically shuts off if the video drops out too much.

http://www.radrotary.com/FPV2_raw.wmv

After the video ends, things didn't get much better. The UAVP was level, and flying fine, but it was so far away that I just couldn't determine the orientation by looking at it. Looking at the video didn't help either, because it was low, and over scrub weeds which all looks the same. Finally, I just lowered the throttle and let it come down before it got too far out of range. It looked like it hit pretty hard, and not level.

Amazingly, when I got to it, I couldn't immediately see any damage. It was sitting upright, on it's narrow camera base, almost as if it had made a perfect landing. If it weren't for the battery laying on the ground (still plugged in), I could have throttled it up and flown it away. Of course two of the arms were bent, one pretty badly. I brought props with me, but no arms, so I didn't have any choice but to fly it with the bent arms, and it didn't make any difference.

Flight number two (FPV3 video) was made with the compass hold turned on. This worked out MUCH better, and the entire flight you see is made without looking at the UAVP, except for takeoff and landing. In part of the video, I let it drift behind me, and you see me and the van in the picture. I'm facing North, the same direction the UAVP is. It's very odd hearing it get closer to you from behind, and not being able to trust your ability to judge distance. The video is actually pretty boring, but in this case, boring is good.

http://www.radrotary.com/FPV3_raw.wmv

More testing to come, but I do think this is doable.

Rusty
UAVP quadrocopter
HeadPlay goggles
KX171 and 900MHZ Tx/Rx
13brv3 is offline Find More Posts by 13brv3
Site Sponsor
Reply With Quote
Sign up now
to remove ads between posts
Old Dec 30, 2007, 12:23 AM
Registered User
PeteSchug's Avatar
Elmhurst, NY (Queens in NYC)
Joined Apr 2004
7,051 Posts
Interesting, I liked the second flight better.

Could you post some pictures of your quadrocopter?

Pete
PeteSchug is offline Find More Posts by PeteSchug
Reply With Quote
Old Dec 30, 2007, 04:12 AM
Happy FPV flyer
Kilrah's Avatar
United Arab Emirates, Dubai
Joined Sep 2006
4,872 Posts
Hmm, interesting.
I think a bit of the problem comes from that way too wide angle lens. The thing is that it can deform things and bend distances beyond recognition. At the start of that first video, I thought you were on a beach and those water surfaces were 100m wide, and looked like pretty good but maybe too similar looking references for positioning. However once you started flying what ended maybe only being 100m away, you simply couldn't even see what was your main reference anymore (those spots that seemed big now being too small to recognise from a distance with the wide angle ).
It's not too much of a problem if you do it correctly, we're pretty good at "calibrating" our vision to what we get, but pretty much like a touchscreen needs to be calibrated to know what is where we need to give our vision known references first to adapt to it, then we can keep that in mind and figure out where we are.
That view you had after takeoff didn't have enough visual references to get a precise idea, everything looked the same. What I did on my first rotor FPV flights was to take off that way in front of me, then turn 180 in order to look at myself, and hover around myself, going a bit further, coming back, but always keeping what was both a known size reference and the target I'd need to come back to in the end in the view. No risk of getting lost that way. Then, once you get used to that (quickly) you know you can look somewhere else, and do the same with any other reference like for usual FPV navigation.
I guess I still wouldn't feel that comfortable with that wide lens though.

I don't think the heading hold of the second flight changed much. If you watch the first video, the thing isn't at all moving around the yaw axis anyway... that's of course until you got lost and started moving it yourself. But that was a bit late. The advantage of those rotorcraft is that you can do a 180 turn immediately whenever you want. So when you get lost, instead of dropping the googles, try to find the thing visually, only to see it's too far and you can't tell its orientation anyway, just stay in FPV, use that yaw stick to do a 360, find a known reference (you'd have seen your car, or at least the runways or Xs printed on them), and there you are, you can calmly fly back.

So, most of it is not enough "planning", that is thinking about what situation you'll be in, and what to do in those cases in advance. People who fly full scale aircraft have procedures they are taught and repeated for all the current stuff, and written ones in case of problems, then personal initiative is a last resort. I.e people have been planning as many possible cases they could, and thought about what to do in all of those cases. We have nothing of that kind, so it's our job to do it, and figure out how to react to "problematic" situations, and it's always easier to do it before it happens like they do instead of improvising on the moment

And I definitely need to edit that video together. I wanted to wait for more flying to have something better, but it's been 1 month I say that, and it still doesn't look like the weather will be kind with me soon enough...
Kilrah is online now Find More Posts by Kilrah
Last edited by Kilrah; Dec 30, 2007 at 04:20 AM.
Reply With Quote
Old Dec 30, 2007, 10:47 AM
13brv3's Avatar
Navarre, FL
Joined Mar 2002
3,839 Posts
Quote:
Originally Posted by PeteSchug
Interesting, I liked the second flight better.

Could you post some pictures of your quadrocopter?

Pete
Thanks Pete. I'll attach a couple pics, but there's more of my UAVP info in this thread- http://www.rcuniverse.com/forum/m_6405898/tm.htm

Rusty
13brv3 is offline Find More Posts by 13brv3
Site Sponsor
Reply With Quote
Old Dec 30, 2007, 10:52 AM
Happy FPV flyer
Kilrah's Avatar
United Arab Emirates, Dubai
Joined Sep 2006
4,872 Posts
Quote:
Originally Posted by Kilrah
And I definitely need to edit that video together.
There

And vimeo:

Mikrokopter FPV tests (9 min 52 sec)
Kilrah is online now Find More Posts by Kilrah
Reply With Quote
Old Dec 30, 2007, 10:57 AM
13brv3's Avatar
Navarre, FL
Joined Mar 2002
3,839 Posts
Hi Kilrah,

Thanks for the excellent comments. Absolutely spot on. I had realized a bit of this already, mostly about the lens, and needing better references, but your post really sums it all up.

Cheers,
Rusty

PS- I just saw your video. Wow!!! Absolutely awesome I'm not worthy
13brv3 is offline Find More Posts by 13brv3
Site Sponsor
Last edited by 13brv3; Dec 30, 2007 at 11:17 AM.
Reply With Quote
Old Dec 30, 2007, 09:39 PM
Tri-Quad-Hexa-Octo-copters!!
United States, TX, San Antonio
Joined Feb 2007
14,624 Posts
Most excellent quality! Did you post process? What did you edit with? What kiind of camera and tx/rx set up did you use?
Thanks,
Jim
Quote:
Originally Posted by Kilrah
jesolins is offline Find More Posts by jesolins
Reply With Quote
Old Dec 31, 2007, 12:01 AM
Registered User
Old Man Mike's Avatar
Joined Sep 2007
1,803 Posts
Quote:
Originally Posted by 13brv3
Hi Kilrah,

Thanks for the excellent comments. Absolutely spot on. I had realized a bit of this already, mostly about the lens, and needing better references, but your post really sums it all up.

Cheers,
Rusty

PS- I just saw your video. Wow!!! Absolutely awesome I'm not worthy
Let me second Rusty on that - very impressive, Kilrah!

I think the Quads are going to set the new standard in FVP flying and I will certainly be trying it with both Kquad UAVP and Kquad DF.

I also would like to hear more details about the goggles, video camera, lens and transmitter used.

Mike
Old Man Mike is offline Find More Posts by Old Man Mike
Last edited by Old Man Mike; Dec 31, 2007 at 12:42 AM.
Reply With Quote
Old Dec 31, 2007, 03:02 AM
Happy FPV flyer
Kilrah's Avatar
United Arab Emirates, Dubai
Joined Sep 2006
4,872 Posts
My usual TX40 camera 2.96mm lens (90 diagonal), and it was the first time I tried the Oracle with it too.
No post-processing, edited with Premiere and rendered at 1Mbps WMV.
If you watch the original file you'll see that in the first part the image is pretty good except for being after sunset, cam a bit out of focus and a bit where it was "faded" (RX battery empty, Oracle had already been beeping for 10 minutes, the last landing I had lines on all the bright zones ). For the day part the cam went a bit loose and gave some vibration, but the 3rd part is pretty correct. It actually seemed better to me when editing the video than when I was flying, I guess that's me being a bit too picky as usual
Kilrah is online now Find More Posts by Kilrah
Reply With Quote
Old Dec 31, 2007, 03:26 AM
Free Flying
anci3nt's Avatar
Joined Nov 2006
525 Posts
Kilrah, I noticed your altitude display went a little screwy at the end. I experience the same thing. Any ideas?
anci3nt is offline Find More Posts by anci3nt
Reply With Quote
Old Dec 31, 2007, 03:36 AM
Happy FPV flyer
Kilrah's Avatar
United Arab Emirates, Dubai
Joined Sep 2006
4,872 Posts
Yeah indeed, just checked the original recording and the GPS module did go wild about altitude once or twice, the rest being OK. Hadn't even noticed before
That's something we have to live with, altitude is far less precise than position, and depends on the position of the satellites. It's good when you have a few overhead, but when most of them have low elevation altitude precision decreases (while it logically increases the horizontal position's one).
Kilrah is online now Find More Posts by Kilrah
Reply With Quote
Old Dec 31, 2007, 07:34 AM
FPV Desert Beta Test Center
Mesa, Arizona
Joined Nov 2006
2,433 Posts
GPS vertical accuracy has been found, at least in one test, to be better or equal to aircraft baro systems.
http://tinyurl.com/ypg3sk
My own experience with sailplane recorder logs show a +4% variation between the two methods. The top trace is GPS.

http://tinyurl.com/2h9umd

This doesn't address the bigger question of GPS accuracy and my equipment may be the cause differance.
Gary Evans is offline Find More Posts by Gary Evans
Reply With Quote
Old Dec 31, 2007, 08:53 PM
13brv3's Avatar
Navarre, FL
Joined Mar 2002
3,839 Posts
Here's a much better attempt from today. I used the "wide" lens, rather than the "super wide", and used the van as a reference. This was all goggles from takeoff to landing, no peeking at all

http://www.radrotary.com/FPV_123107_raw.wmv

Rusty (long way to go to catch Kilrah)
13brv3 is offline Find More Posts by 13brv3
Site Sponsor
Reply With Quote
Old Jan 01, 2008, 09:10 AM
Tri-Quad-Hexa-Octo-copters!!
United States, TX, San Antonio
Joined Feb 2007
14,624 Posts
Rusty,
Nice job! Have you tried using your FX07 with video out? That gives superb 16X9 format results and you can put a small servo on the zoom too. Is that a special antenna set up in front of you to your left? What kind of video glasses were you using? You also have a Eagletree Seagull data logger too right? Is there a capability with that to put compass and altitude (for a start) info in the video?
Happy New Year!
Cheers,
Jim
Quote:
Originally Posted by 13brv3
Here's a much better attempt from today. I used the "wide" lens, rather than the "super wide", and used the van as a reference. This was all goggles from takeoff to landing, no peeking at all

http://www.radrotary.com/FPV_123107_raw.wmv

Rusty (long way to go to catch Kilrah)
jesolins is offline Find More Posts by jesolins
Reply With Quote
Old Jan 01, 2008, 10:34 AM
13brv3's Avatar
Navarre, FL
Joined Mar 2002
3,839 Posts
Quote:
Originally Posted by jesolins
Nice job! Have you tried using your FX07 with video out? That gives superb 16X9 format results and you can put a small servo on the zoom too.
Hi Jim,

The FX07 is my ultimate goal, because it will record the flight on SD card, so I won't have any of the wireless video problems on my recorded video. Unfortunately, there's a bit of a lag on the video out of the FX07, which will make it harder to fly. My hope is to give it a try when/if I ever get comfortable with the current setup.

Quote:
Originally Posted by jesolins

Is that a special antenna set up in front of you to your left?
Yes, it's a 900MHz 8dbi "patch" antenna. I'm using this with a 500mw 900MHz TX, and dual output RX. Camera is currently a KX171, and the recorder is a Coby DVR. All but the Coby came from Range Video.


Quote:
Originally Posted by jesolins

What kind of video glasses were you using?
I wanted the best, and from everything I read, that seems to be HeadPlay. I don't have much to compare them to, but they do seem to work well. I'm still in fit and fiddle mode to try to get them to stay exactly where I need them for the whole flight.

Quote:
Originally Posted by jesolins

You also have a Eagletree Seagull data logger too right? Is there a capability with that to put compass and altitude (for a start) info in the video?
Eagletree is working on an OSD module that's expected to be ready to sell in mid to late January. I'll be surprised if it's as well tailored for FPV as some of the other OSD models are, but I guess I'll have to wait and see.

Cheers,
Rusty
13brv3 is offline Find More Posts by 13brv3
Site Sponsor
Reply With Quote
Reply


Thread Tools