First, congratualulations to Brian Wolfe, the winner of the second round of the T3 Contest. Now it's time for Round 3.
This round's objective is to break the Stanford team's UAV altitude record of 7,142 feet by doing at least 24 circles with a 300ft climb and descent in each, as shown above. (This won't really beat his official record, because there won't be an official judge there. But you'll get bragging rights, at least).
The winner will have the highest cumulative altitude, but anyone who exceeds 7,142 feet will win a prize.
As usual, you must submit a KML track and video in the comments below. Evidence that fun was had is welcome (and may influence Gary's point assignment blackmagic equation), but is not required.
Deadline is Midnight PST on November 29th.
I'm impressed with the run time Mark got out of his battery.
I went back out this last weekend and made an attempt at the course but as I expected didn't get anywhere near the run time to get to 7100ft. I have made some changes to improve things plus my battery wasn't completely topped off due to a false start. After flying for about 5 min. I had to land and remove the camera. It was causing some tip stalls that for the most part the autopilot was handling, but I decided not to take a chance so I landed and removed the camera. I'll post that video when I get a chance. Unfortunately ( or fortunately depending on how you look at things) were headed to Hawaii tomorrow for vacation so I doubt I'll get a chance at another run unless things get extended again.
So here is where I'm at for an entry:
- AXN Floater (Easy clone) with Picpilot Autopilot.
- 2400mAh 3S 20C LIPO.
- 20 climbs, 1602 meters (5256 feet) cumulative climb while autonomous.
- 1500 seconds total (25min).
Stanford's record broken again - second time for me!
I refined & tuned my setup and achieved an even better result! Weather conditions were acceptable; light wind with some rain traveling along the nearby mountains (fortunately not in the flight area).
Details are:
- Easystar with Paparazzi Autopilot.
- U-blox LEA-5H GPS with ceramic patch antenna.
- 2400mAh 3S 30C LIPO.
- 33.5 climbs, 3'296 metres (10'813 feet) cumulative climb.
- 2515 seconds total (41 mins 55 sec).
GPS altitude plot (x-axis in seconds, y-axis in deci-metres):
Side view (looks like a neatly wound ball of string);
Top view. Very smooth this time - very minor course deviations:
All went very well without incident and I was very impressed with the results. Auto 2 mode (autonomous) launch stayed in Auto 2 until the battery was exhausted (2310 mAh consumedfor the whole flight). No birds chasing the UAV this time :-)
I'll post a video soon but given the flight duration, it's awfully long so will cut it short somehow.
T3-3 Pre-Run Test:
I finally found some time to start looking at the new T3 challenge but decided to take a baby step since I'm making several changes/updates for this round. Firstly, I haven't taught my autopilot how to do circles yet. I was looking at different approaches to doing that but decided in the interest of time to just approximate a circle with twelve way points. Rather than having to enter those way points multiple time though I decided to add a "repeat" function to my mission scripting language so I could tell the autopilot to simply repeat WP commands x through y z number of times then jump to command q. I also have no data on the motor/prop combo that comes stock in the AXN Floater (Easy Clone) so have no clue where the sweat spot is for efficiency. And lastly I just got a new Netbook to use in the field as a ground station, so I decided to do an abbreviated 4 loop climb/dive cycle test to check everything out. Here are the results:
The first thing I learned was that the netbook goes to sleep after 4 or 5 min and stops logging data!! Note to self: disable sleep function before test.
I decided to take a stab and run the throttle at 65% during climbs and 35% during descent. After looking at current draw and running some numbers I'm seeing that it takes 0.151Amp/hrs to do each climb at my chosen climb rate and throttle. Given the size of my battery that means about 15 cycles. I'll have to do something different to hit the needed 24 cycles. I'll look into a lower throttle and slower climb rate to see if efficiency improves enough to make it. At this point it's less an autopilot test and more a platform/battery test. It could be the motor prop combo is not up to the task.
The last thing I found was the platform had a hard time descending at the required rate. I think this is a trim issue. I noticed during the manual flyout it wanted to climb but I was too excited to start the run and didn't trim it out.
Ultimately I want to be able to tell the plane to fly in circles and simply climb at a commanded rate, taking all the time it needs until the commanded altitude is reached then change directions and start the dive.
Far from a winning entry but we got a break in the weather around here so I decided to give it a go. I used my EasyStar with a 2200mAh battery and all video equipment (pan/tilt pod, camera and transmitter), so there is no way I can compete with Krysztof's setup.
I used Paparazzi but with a few modifications, I connected an EagleTree airspeed and altimeter sensor through the I2C bus to the autopilot board. With a few software modifications I was able to hold airspeed and constant climb rate, without having to figure out the optimal throttle for the climb. The airspeed chosen was a bit above stall speed.
Cumulative climb was 6152 ft (1875 m). This will likely be my last entry until the weather improves, the hardware in my EasyStar will go to my winter project (quadcopter).
Comments
I went back out this last weekend and made an attempt at the course but as I expected didn't get anywhere near the run time to get to 7100ft. I have made some changes to improve things plus my battery wasn't completely topped off due to a false start. After flying for about 5 min. I had to land and remove the camera. It was causing some tip stalls that for the most part the autopilot was handling, but I decided not to take a chance so I landed and removed the camera. I'll post that video when I get a chance. Unfortunately ( or fortunately depending on how you look at things) were headed to Hawaii tomorrow for vacation so I doubt I'll get a chance at another run unless things get extended again.
So here is where I'm at for an entry:
- AXN Floater (Easy clone) with Picpilot Autopilot.
- 2400mAh 3S 20C LIPO.
- 20 climbs, 1602 meters (5256 feet) cumulative climb while autonomous.
- 1500 seconds total (25min).
KML File:Templet.kml
GPS Plot:
Alt Plot:
Voltage and Current (I Told the plane to come home when the voltage got to 9V):
Brian
I refined & tuned my setup and achieved an even better result! Weather conditions were acceptable; light wind with some rain traveling along the nearby mountains (fortunately not in the flight area).
Details are:
- Easystar with Paparazzi Autopilot.
- U-blox LEA-5H GPS with ceramic patch antenna.
- 2400mAh 3S 30C LIPO.
- 33.5 climbs, 3'296 metres (10'813 feet) cumulative climb.
- 2515 seconds total (41 mins 55 sec).
Here is the KML file: T3-R3_EasyStar_Givrins_2400mah.kml
GPS altitude plot (x-axis in seconds, y-axis in deci-metres):
Side view (looks like a neatly wound ball of string);
Top view. Very smooth this time - very minor course deviations:
All went very well without incident and I was very impressed with the results. Auto 2 mode (autonomous) launch stayed in Auto 2 until the battery was exhausted (2310 mAh consumedfor the whole flight). No birds chasing the UAV this time :-)
I'll post a video soon but given the flight duration, it's awfully long so will cut it short somehow.
Cheers, Mark
I finally found some time to start looking at the new T3 challenge but decided to take a baby step since I'm making several changes/updates for this round. Firstly, I haven't taught my autopilot how to do circles yet. I was looking at different approaches to doing that but decided in the interest of time to just approximate a circle with twelve way points. Rather than having to enter those way points multiple time though I decided to add a "repeat" function to my mission scripting language so I could tell the autopilot to simply repeat WP commands x through y z number of times then jump to command q. I also have no data on the motor/prop combo that comes stock in the AXN Floater (Easy Clone) so have no clue where the sweat spot is for efficiency. And lastly I just got a new Netbook to use in the field as a ground station, so I decided to do an abbreviated 4 loop climb/dive cycle test to check everything out. Here are the results:
The first thing I learned was that the netbook goes to sleep after 4 or 5 min and stops logging data!! Note to self: disable sleep function before test.
I decided to take a stab and run the throttle at 65% during climbs and 35% during descent. After looking at current draw and running some numbers I'm seeing that it takes 0.151Amp/hrs to do each climb at my chosen climb rate and throttle. Given the size of my battery that means about 15 cycles. I'll have to do something different to hit the needed 24 cycles. I'll look into a lower throttle and slower climb rate to see if efficiency improves enough to make it. At this point it's less an autopilot test and more a platform/battery test. It could be the motor prop combo is not up to the task.
The last thing I found was the platform had a hard time descending at the required rate. I think this is a trim issue. I noticed during the manual flyout it wanted to climb but I was too excited to start the run and didn't trim it out.
Ultimately I want to be able to tell the plane to fly in circles and simply climb at a commanded rate, taking all the time it needs until the commanded altitude is reached then change directions and start the dive.
Comments welcome,
Brian
Cheers, Mark
YouTube Video
I used Paparazzi but with a few modifications, I connected an EagleTree airspeed and altimeter sensor through the I2C bus to the autopilot board. With a few software modifications I was able to hold airspeed and constant climb rate, without having to figure out the optimal throttle for the climb. The airspeed chosen was a bit above stall speed.
Cumulative climb was 6152 ft (1875 m). This will likely be my last entry until the weather improves, the hardware in my EasyStar will go to my winter project (quadcopter).
KML file
seen that Chris does not have a simple up/down pattern but a "birthday cake" one (;P), would it be allowed a pattern with climbs in any direction?
Ric
-
2
-
3
-
4
-
5
-
6
of 6 Next