Roger Meike's Blog: Sun Labs and Beyond

So we launched our SPOT controlled weather balloon, EDGAR1, yesterday. It was really a lot of fun. Here is my account of the day.

This project was started about two weeks ago by our extraordinary intern, Tennessee Carmel-Veilleux. Working in the evening and on weekends he and other Sun Labs researchers designed and executed a plan to put a Sun SPOT in near space. A weather balloon would carry aloft a payload including radios, GPSs, various sensors and of course a Sun SPOT for control. It should rise into the air some 10 - 15 miles and then gently parachute to safety where we would gather up our gadgets and analyze the data. Sounds like fun!

Well, after a couple weeks of hard work, we had a plan. We purchased a helium balloon and fashioned a payload for it. We used the MicroTrak 8000FA for tracking.

Finally, we gathered yesterday morning (April 5) at 7am for our preflight briefing. We checked the weather conditions and chose our final launch location. We picked a deserted looking cul de sac near Hwy 101 and 156 close to San Juan Bautisa, CA. We ran our simulations (using cool tools like W3BalTrak from Near Space Ventrues or the U Wyoming Balloon Trajectory Simulator) and predicted a roughly two hour flight and about a 90 mile trajectory to the south and east. The launch team headed to the launch site while a chase team headed downwind to California's central valley some 200 miles away. A third team headed to a nearby airstrip to be the eyes in the sky.

I was part of the launch team. We headed south to set up. At about 9:15 we discovered that the launch site that we were hoping for was in a gated community, so we found a nice grassy field just off the road about a quarter mile from the planned location and decided to launch there. We packed the electronics in the payload, conducted final tests and started to fill the balloon. We checked in with the chase teams and all was going well. We had to wait a bit for the clouds to burn off to the 50%-or-less that is required for a safe balloon release. Just as the skies cleared enough for the launch, the plane showed up on site and EDGAR1 was launched. It was really a great feeling seeing it head up into the sky. After a few tense moments watching it head a bit to the west toward the ocean, the winds kicked in just as predicted and it headed for the south east and out of site.

At this point all we could do we track its amateur radio transponder that was giving us continuous GPS data via a system called APRS. Some of us had ham radios, but for those who didn't, there is a nice website called http://www.findu.com where you can track these signals. We tracked the balloon as long as we could and the tracking airplane was actually able to see it go passed them which is pretty much a miracle when you see how small a spec the balloon is in the big sky.

Some of the launch crew went off to join the chase. I and my family had a social obligation that we had to go to, so we drove back up to the SFBay area. Thank goodness for iPhones that let us continually update the position from the web as we travelled. The chase crew had location information via the ham radio, but they had no course or speed information. We acted as a communication point relaying this predictive information to the chase team. We watched as the balloon climbed through 60,000ft, then 70,000ft. It was working!!!

The flight parameters had been set so that the Sun SPOT would cut the flight at 1h 55min or if it went far enough that it was going to drift into the Sierra mountains. At around 80,000ft, we lost altitude information. Apparently GPS has problems at these high altitudes, although I suspect in this case the GPS has a software imposed reality check... if the altitude says you are over 80,000ft, there must be something wrong! It was rather nerve racking not knowing what was wrong with our data feed. We did however, know that we were reaching the timed conclusion of the flight. At this point the Sun SPOT was supposed to trigger a heated wire that would melt the rope and release the balloon from the payload. The payload would them parachute down.

We waited, but there was no sign that it was descending. It was just chugging along at an unknown altitude. hmmm. Now what? It continued along for a very long 10 minutes at least. There were some pretty tense moments and then... suddenly.... we started to get altitude data agian. It was falling. And just a few miles farther than our predicted location. After a few sighs of relief we concentrated on figuring out where it was to land. It looked like it would be down in another 10+ minutes. The chase crew was able to situate themselves to the area where they thought it would land ahead of its arrival. They never did see it and neither did the chase plane although they had it in had within ten minutes of it touching down. It had a minor bump in the bottom, but was otherwise unharmed.

Post Mortem:

The flight was over 100 miles and reached over 80,000ft. I judge any flight that goes that high and ends in recovery of all the equipment is a very successful flight. However, you'll notice that I haven't posted any of the pictures from the onboard camera. That is because there weren't any. In fact the reason the flight went longer than the programmed length is because the balloon was never cut free. Instead it eventually popped because of the low air pressure. We understand that the balloon that was 6 ft in diameter when we launched it was probably about three times that, 18ft, and it must have been close to 90,000ft, when it exploded. We'll be analyzing what caused the failures tomorrow.

In the mean time, I must say this was a LOT of fun and we will do it again.

For more information about how this project went stay tuned to http://groups.google.com/group/spotballoon. We'll be posting more technical details and information there.