I'lll try to asnwer a few questions, before I set up a proper page with all the results....
How did you track it?
The usual electronics setup for these kind of missions is to have a microcontroller connected to a GPS module and to a radio system. The system sends co-ordinates back to earth. This can be done in several different ways, generally the NMEA data is directly modulated via radio. Donny, for our launch made in december from Florida, use voice synthesis to read the co-ordinates over a radio.
My capsule was much, much simpler. First of all, for cost reasons (I funded the project all alone and I didn't want the cost to rise too much) Second, I wanted to only use COTS components (i.e. no soldering required)
The heart of the tracking system was a GPS-via-SMS tracker.
It's a tiny little box containing the "brains" of a GPS and of a GMS mobile phone. The tracker sends you SMS messages with the co-ordinates, in several different ways (for example, "you call, and the tracker sends an SMS"; or "the tracker sends an SMS every X seconds")
You can see the tracker here:
http://www.mr-lee-catcam.de/pe_cc_o2_en.htm
PDF manual here:
http://www.mr-lee-catcam.de/BINARY/Manual_CatTrack_Live2.pdf
The first problem was obviously that the GSM signal isn't available so high, so the project was to send up the balloon, get SMSes up to a certan altitude, then no more until the capsule was returning down.
To have a clearer ideas of the range of GSM signals I placed the tracker in the pocket of a skydiver. Going up, I was able to track him up to about 2.000 meters, but then surprisingly as soon as he jumped (at 4.400 meters) as he came out of the plane's fuselage the tracker was able to get signal again.
So I was optimistic about this. Other possible problems could be from poor GSM coverage.
If the capsule landed in an area without GSM coverage the payload would be lost. I got a SIM card from the italian carrier that officially has the best coverage.... anyway, I placed a small label on it, saying "please call XXX" but that would be a "last hope" method...
Also, in case the capsule ended up in an area with mediocre GSM coverage, it might end up for example is a small valley that has not GSM coverage because of the surrounding hills. I hoped that even if at the landing spot there was no signal, a few hundreds meters the capsule would be "outside of the valley" and that I could anywat get SMSes sent during the descent phase, a bit before touchdown. This proved to be false (i.e. in reality, I only got signals when the capsule touched down. I still have to understand why)
In reality, the capsule recovery was completely different from what I expected..... Will post more details later 😀
In addition, and not for tracking purposes but just for logging purposes (i.e. to get more info if I recovered the capsule) I had a GPS logger like
this
What was the payload weight?
Balloon 1.200 g. (KCI 1200 listed here:
http://www.kaymont.com/pages/sounding-balloons.cfm )
Total payload (camera, GPSes, enclosure and parachute) 460 g.
How far did it land from the launch point?
About 80 kms total
This distance varies a HUGE lot, and is usually much larger than this.
You can simulate launches in real-wind conditions using this tool:
http://www.srcf.ucam.org/~cuspaceflight/predict/index.php
My distance was rather short because high-altitude winds were in opposite direction of low-altitude winds, so the capsule actually was something like 90 kms to SE, 40 kms to NW, 30 kms to SE. so the landing was only 80 kms from my place, but actual trajectory was about 160 kms....
What kinda script did you use?
apart from times, the script was not changed from the one of the launch in Florida:
https://chdk.setepontos.com/d/2877
did you log temperature as well?
Yes. As I suspected that in the Florida launch the batteries died during re-entry because of very cold conditions (-10° inside the camera) I paid a LOT of attention to insulating and padding the camera as much as possible. This time the temperature never went below 20°
and how did you make those nice panoramas? i mean, html, flash, and stuff?
- Extract frames from a section of the video where the camera does a 360° turn
- Select frames
- Use autopano.net to automatically stitch the frames
- Use krpano.com to display the resulting image. At the end, I selected krpano because it's in flash and requires no additional plug-ins
