Building a Tricopter

While Joshua is doing good work with his rockets and high altitude balloons, I started to work on the autonomous control of a flying vehicle and decided to build a tricopter to gain experience. The design of a tricopter is pretty easy, three motors, 120° offset in between, CoG in the middle, one rotor (in my case the tail) needs to be move able around the x axis to counteract the centrifugal forces of the other rotors and to allow yaw turns. For the electronic (IMU, GPS, Controller) I decided to go for something pre build and open source. I found the ATMEGA approach of the DIY-Drones very interesting and simple and I use it as basis to get into the matter. Last weekend I had finally my first flight in the wild, unfortunately I lost one propeller during flight and did a double three sixty, but since I landed in snow, no damage occurred.   Have a look on the video, sorry for the poor video quality.

Flip the tri

Flip, the tricopter

The Moon seen from Near Space

Moon seen from WikiSat camera

Moon seen from WikiSat camera

Few minutes before the burst, a wanderfull view of the Moon was recorded from our wikisat camera in the Launch 08 past Saturday July 9th, 2011. This camera is also used in the PicoRover for the GLXP.

Simulation of the Moon seen from WikiSat camera

Simulation of the Moon seen from WikiSat camera

Sunrise was scheduled at 06:34 am and sunset was at 21:39 Spanish local time. Lucky, the Moonrise was 14:46 pm thus Moonset was 01:46 am at the following day. Burst time was 16:46 and flight time at this moment was about 3 hours.

Previous pictures show this moment from the Moon2.0 simulator and the WikiSat onboard camera. Also the real frame at the same time, the same coordinates and same attitude are showed.

The femto-satellite WikiSat V4

The femto-satellite WikiSat V4

In one hand, this fact validates our simulator as a proper tool for camera view prediction. In the other hand, the onboard GPS reported a maximum altitude of 50 km which can not be. Our theory about this fact is that the GPS reading algorithm has some bugs and sometimes GPS messages are wrong. See this report of how we recovered the Arduino onboard computer:

WikiSat Launch08 – Arduino Altitude Recovery

WikiSat Launch08

Last Saturday July 9th, 2011 we have got another high altitude balloon launch (Number 08).

Near Space view 8th launch

Near Space view 8th launch

The intention was to test the recovery method and to launch the second stage of the WikiLauncher (See a ZERO coke at the end of this post). The launching point was Zaragoza (Spain) as usual and the landing point was Huesca (Spain). The trip was more than 3 hours and the distance was 83 km (44 nm).

Launch 08 TV report

Launch 08 TV report

We have got an APRS radio link and an IRIDIUM satellite based communications attached to the balloon ramp. APRS only failed in the landing but IRIDIUM worked well.

APRS and IRIDIUM recovery devices

APRS and IRIDIUM recovery devices

The onboard computer failed to activate the second stage ignition so we recovered the rocket together with the launch ramp and the rest of the balloon. The burst was at 33 km of altitude (108,000 ft) thus launching point should be at 32 km (105,000 ft) so we have validated the balloon as well. This is a video of the final N-Prize mission.

Thanks to Pere Renom & Kari (TV3), Javier Perez (IDeTIC) and Juan Martinez (UPC) for they help.

These are some launches we made:

Launch 08 IRIDIUM validated – 2011 Recovered

Launch 07 WikiSat communications validated – 2011 Not recovered

Launch 05 Valve and launch ramp validation – 2011 Recovered

Launch 01 Balloon and HTC magic validation – 2009 Not recovered

WikiLauncher version 1

WikiLauncher version 1

WikiLauncher parts

WikiLauncher parts

Why not going reusable!?

I hate the thought that 80% of any rocket is not reusable. Hence, I decided to start to design a 2-stage -winged launcher. On the two pictures you can see the lower left half from two different perspectives. I try to lower the Cw as much as possible and to include the fuselage as aerodynamical element .

wikisat_wingedtwostagelauncher

wikisat_wingedtwostagelauncher

lowerfuselage01_2

Laser I

I recently started to do some experiments with laser – nothing big, but my final goal is it, to develop some tool to measure distance by help of a laser beam. Currently I see there are four different ways to do so:

  • Measure the time of flight – pulses
  • Measure the phaseshift – amplitude modulated
  • Measure the superposition – amplitude modulated
  • Triangulation – measure the angle between laser point and a optical sensor (CCD)

Method #1 is most probably the easierst one – if either the distance is far enough or fast equipment is used – to measure the delay between a outgoing and reflected incoming pulse requires something that can measure at least nanoseconds if I want to measure distance below 1m. Whereas #2 and #3 are pretty similar and much better for shorter distances, since the phase shift or the superposition is quiet easy (….relative…) to measure. It can start to be tricky after a certain distance, when the phaseshift is again (X byPI with X is 1,3,5…), since I have the same result for different distances. Determining the distance with method#4 is simple too, as long a program is availabe to calculate the the fixed angle between CCD sensor (picture)  , laser point and  laser. There are some out in the I-net, using regluar webcams for doing so succesful.

I have started now with an arduino board, an easy circuit (transistor…) and a green laser pointer to collect some experiences by myself. 

Yesterday I did some pulse width modulations (very slow about 20ms) , the transitor I use (NTE46)  is way not fast enough to switch the laser pointer in a useable speed. Today I’ll try the AM modulation. Let’s see what my results are.

Silence

by Carlos Sieiro del Ni… © All Rights Reserved

Sun setting over Teruel, Spain
by Carlos Sieiro del Ni…
© All Rights Reserved


It’s now T plus 100 hours, and we still have not heard a blip from our test balloon since launch on Friday. Joshua estimates that the battery of the recovery system can last about 5 days (120 hours), making it pretty unlikely that we are going to hear anything at all from it at this point.

Joshua inscribed his phone number, along with the address of this website, on the side of the payload before launch, so our last faint hope is that someone will find it, and give Joshua a call and/or lookup this website for more information. (Should that someone happen to be you, we would of course be very grateful to hear from you – our e-mail address is listed under Contact.)

The balloon was launched on Friday at 4:08 AM in Zaragoza, just off the E-90 highway. Launch of the balloon went well, and wind was consistent with our trajectory forecast.
Continue reading