Dan's Story

gallery | dan's story |

Balloon Project 2008

We started our vacation mid-week and arrived at ‘Launch Control’ (mom and dad’s house in north Florida) on Wednesday afternoon. The rest of the week involved last minute adjustments to the experiment module and preparing the tracking module for connection to the rest of the launch vehicle. As we approached the hopeful launch date, Dad and I spent our days either working together or each on his on project fine tuning the various bits of the experiment. The 2008 project was a little more ambitious even after failing to recover the 2007 balloon. We hoped that we had learned the lessons from the previous year and we pressed on with optimism. This year we would be sending up a camera to snap pictures, one a minute, just like last year but we would additionally be sending up a simplex repeater for ham radio use in the 430Mhz band. This is normally a line of sight frequency with a normal range of around a mile or less but we would be hoping that people would be able to use it from up to 200 miles away. Additionally, we were sending up a thermochron, a temperature recording device that we would program to take temperature readings at a rate of once per minute. We would then be able to correlate the temperature readings with the altitude data from the tracking module and get some idea of the temperature at the higher altitudes.

In the evenings, we would check the projected tracks using high altitude wind data from NOAA. The Tallahassee weather center sends up their weather balloons at 8am and 8pm and we would wait for each report to see where the projected ‘track of the day’ would take our hypothetical balloon. In the days leading up to the launch, there were tracks that predicted a landing only 9 miles away from the launch point. On the day before the launch, we were beginning to notice the jet stream dipping into our area which was producing a high speed easterly wind starting around 35,000 feet. However, once the balloon popped out of the jet stream at the top, there was a moderate westerly flow that kept the balloon from flying right out to the Atlantic ocean. Unfortunately, if launched from either of our two primary launch sites or our first alternate launch site, the tracks predicted the balloon landing in some pretty nasty North Florida swamps and bogs. When calculated from our final alternate launch site, we found the projected landing to be in some relatively safe forest and/or farm land. After a couple of days of watching the winds and weather, we were pretty confident by Friday morning that this weather system was going to prevail and we started notifying the other members of the launch team of our decision to launch at my brother, Larry’s house in Thomasville, Georgia. We were finally confident of our decision after I was able to get in touch with Larry, who was returning from a business trip in Milwaukee, Wisconsin that day, and let him know that we would be descending on his home around 6am the following morning and, oh yeah, could you please have coffee ready for about 20 people.

On Friday evening, prior to launch, we were all very excited. My wife and daughter had already gone back to the hotel in anticipation of the 4:30am departure from mom and dad’s house the next morning. Dad, my son Josh, my brother Donnie and I sat up finalizing details for the following morning. The prior year’s launch included a team meeting the night before launch to let everyone know what the plan was and what everyone would be expected to do. This year, that little detail seemed to go by the wayside since most of the launch team was involved in other things and they all seemed to know it would be an early start Saturday morning. I finally excused myself to the hotel and eventually got to sleep around 2am. I double checked the alarm clock and verified the time it was set. Unfortunately, both my wife and I were so tired that we verified the alarm time to be 4:30am, instead of early enough to be on the road by 4:30.

A little explanation about the schedule: Larry’s home is in Thomasville, Georgia which is a drive of about one hour and 15 minutes from mom and dad’s house. We figured a departure time of 4:30 would get us there a little before 6am. It took us around 1 hour to prepare for launch last year so we could get there for a 7am launch. Accuweather.com had predicted 0mph winds on the ground from around 4am to around 8am so with a 7am launch, we wouldn’t have to worry about the balloon moving around on us once it was inflated and waiting for us to connect the payload.

So, we didn’t get underway until nearly 5am (a record to get my family up and on the road in less than half an hour) with profuse apologies all around. We took all the requisite ribbing for ‘being on Chicago time’ and ‘holding up progress’ but were soon on the road to my brother’s place in Georgia.

When we arrived at Larry’s, things seemed to move in slow motion but always in the direction we wanted to go. I think I was still in preparation mode, where we were always going at breakneck speed to get everything tested and ready for this day. Now that this day was here, we didn’t need to go so fast to get everything going. The launch platform was unloaded from my dad’s pickup truck and trundled down to the launch area. At this point, I looked around and realized that the meeting of the launch team we missed the night before was actually unnecessary. Thinking about it now, it was probably a combination of experience from last year, coupled with the fact that this was a family operation and we all talked to each other from the start to the finish that precluded our need to organize the night before. Whatever it was, I was a little surprised to find that every time I would raise my head from my task of getting the tracking unit ready and assisting my dad with getting the experiment module ready, the entire project was a little farther along the way. As a matter of fact, the balloon was prepared and inflated and my brothers were waiting on my dad and me when everything was ready to be connected together. They knew that we needed 10 pounds of buoyancy which would give us 4 pounds of positive “thrust” with our 6 pound package and they had gone straight to work to get that ready. Even though they were ‘waiting for us’, we still wanted to be very methodical with our preparations. Everything went smoothly until we started connecting the parachute to the experiment module. This took a little longer because we were having to think upside down, considering the way we had the connection ring configured and the parachute spread out in front of the launch platform. This is yet another lesson learned for next year. But once it was connected, the launch team was able to connect the balloon to the parachute and at that point all pieces were ready to go.

I think my dad and I were both taken aback at how quickly everything was ready to go. Dad quickly checked and rechecked battery connections and then switched on the power to the camera. I checked the battery connection to the tracking module and added the ballast that was required. During our preparations we found that the entire package, including the experiment module, tracking module, parachute and all connecting lines only came up to 4 pounds 11 ounces. The parachute was rated for 7 to 14 pounds of payload but we figured we could get away with 6 pounds, so I added a plastic zippy bag of 1 pound 6 ounces of ballast (Dirt - the dirt will become yet another experiment after the fact). Once everything was connected and power applied, I insisted that I see at least 3 good tracks from the tracking module before we launched. This means that I would personally see correct time, correct position and correct altitude three times before I would give the GO signal. After the failure of last year, I wanted to know that if it screwed up again, I had at least done everything in my power to make sure it was good when it launched.

We got everything ready to launch, but wanted to make sure we had a good clean launch time. My brother Donnie held on to the entire package for about 2 and a half minutes while we waited for 7:30am to arrive. We did a 10 second countdown and he let the package fly. We have two good photos from two different cameras showing him standing with his arms outstretched and watching the entire package about 50 feet above him.

Once the balloon was up and rising into the morning sky, it turned out to be a bit of an anti-climax. While we watched the tracking coming in and saw the balloon meandering to the east, most of us stood around in small groups talking about the morning’s events. Half of the family went down to my brother’s garden to grab some fresh veggies to take home. There didn’t seem to be any urgency to “chase the rabbit” since we knew we were getting good tracking data and it was pretty much following the projected track we had calculated the night before.

Then it was time to go. Tracking showed the balloon climbing into the jet stream, heading east and picking up speed. Some of the tracks we were getting indicated the balloon was traveling at 70 to 80 mph. Our calculated track predicted the balloon would pop out of the jet stream and reversing direction, moving back to the west before the balloon burst. We decided to drive to Quitman, GA, about 25 miles east, to wait for the balloon burst, but stopped for gas and coffee prior to setting out. While we waited at the gas station, I got my second ’relief’ moment. During construction of the tracking module, I purchased a Garmin GPS engine (see my tracking module write-up for details). I was concerned about the GPS because the US government requires manufacturers to limit all GPS constructed for commercial use to stop reporting position if the altitude is above 60,000 feet AND speed is above 999 mph. Many manufacturers read the restrictions as an OR statement and their GPS units will not report anything with an altitude above 60,000 feet regardless of speed. Some of the research we had done showed that the Garmin E-Trex (an off the shelf handheld gps device) worked just fine at the high altitudes, so I hoped for the best and purchased the Garmin 15 engine. While we waited at the gas station, the balloon began reporting altitudes above 60,000 feet.

Fuel and caffeine onboard, we headed east to Quitman. I was sitting in the front of the van, my wife Mary was driving and my son Josh was sitting in the back with the laptop keeping an eye on the balloon tracking. About half way there, I looked back and saw a strange look on Josh’s face so I asked if the tracking was still looking good. He shook his head and said that the last track he had was about 4 minutes old, not good since the balloon should have been reporting its position every 30 seconds. I, of course, immediately went into panic mode. The last track he’d seen put the balloon at 75,000 feet, and I couldn’t believe that it could let me down in the crunch. Finally my brain started working again and I asked him if anybody else’s tracking was updating. Since we were transmitting our tracking on the common APRS frequency, we were also receiving tracking from all the cars and trucks around us that were taking part in the ‘chase’. Josh checked and sure enough, no tracking from them either. I jumped into the back of the van, checking radios, modems wiring and everything I could think of. Eventually, after deciding that everything was as it should be, it was apparent that I would need to reboot the laptop. We rebooted on the fly and by the time we arrived at the meeting point, a large parking lot on the main road into Quitman, we were once again receiving tracking data from the balloon. Unfortunately, in the rush and chaos of getting everything back up and working, I neglected to turn the recording portion of the program back on. So while we were getting good immediate tracking data showing up on the laptop, none of it was being recorded for later analysis (aaarrgghh!).

Now that we were in the parking lot, we watched the balloon track back to the west. It slowly tracked back and almost went all the way back to Thomasville, but we knew that the balloon would have to pass back through the jet stream on the way down and that would once again push it back toward us. Several people used this time as a bathroom break and a couple went across the street to get a sandwich at the burger place. Suddenly, the tracking became erratic again. We started noticing tracks 2-3 minutes apart and showing only time and position. We weren’t receiving direction, speed or altitude. After a couple of minutes we received one good track, showing an altitude of 69,000 feet! It was on its way down, and not behaving very well as it descended. We hurriedly gathered everyone up and started heading in the direction that the tracking was taking. Since we were traveling in my wife’s minivan, I tried to keep us on paved roads as much as possible. But eventually that became too much of a challenge. Fortunately the tracking software that I was using showed even the smallest of roads in the area we were driving. We took a few dirt road short cuts and eventually started seeing other people of the recovery team, who had been following the balloon as well. Small cars, Big antennas, they were pretty easy to spot. Our tracking eventually took us onto a private farm about 7 miles northeast of Pavo, GA called Okapilko Plantation, managed by Mr. Jim Davis. We followed the tracking via dirt road to a point that we thought was closest to the package. We knew that we were close because we began getting live tracking data once again. The tracking module’s transmitter was transmitting only 300 milliwatts and once below the tree level the transmit range dropped to between a quarter and a half mile.

Once we stopped, Donnie was extremely eager to get going. With E-Trex (as mentioned earlier) in hand, he was all over my laptop trying to get the current position of the package. I read out the coordinates to him while he stood over my shoulder and punched them in. When I finished with the last of the coordinates, I looked over my shoulder to see if he was able to get them into his E-Trex correctly. He wasn’t even there. I looked up and saw him 100 yards into the field following his GPS. My dad and I started following him and it seemed that everyone knew how to do one of these searches. Everyone spread out, walking in the same direction about 50 feet apart, looking on the ground and in the trees (just in case).

Then the shouts came. Donnie had walked right to the package. The relief I felt was indescribable. Not only had my tracking module worked all the way to the top and back but, as it lay on the ground, it had provided a good enough track so that Donnie could walk straight to it with his handheld GPS unit. Dad and I whacked each other on the back and congratulated each other. Even though I knew it was too late, I shouted to leave the package where it lay. I wanted pictures of the package as it landed because you never know what little tidbit of information you can pick up from anywhere. But, alas, in the excitement of the moment, everything was picked up and moved about 50 yards before he even heard my shouting. He put the package back down on the ground and spread it all out, which was great. We were able to get a picture of the entire package in the state it was in when it landed, even if it wasn’t in the ‘place’ where it landed. This gave us valuable data about the state of the package after the balloon burst on it’s way to the ground.

Once my dad and I arrived at the package, dad went straight to the experiment module and pulled the smart card from the camera. His first observation was that it was wet. (We later had an explanation for that. Extremely cold temperatures at high altitude and a very fast rate of descent kept the camera from ‘thawing out’. The high humidity at ground level produced condensate over the whole package. We think that this condensate actually froze for a short time when it came in contact with the cold equipment. The pictures taken by the package after it was on the ground show the camera lens thawing out).

Though we never said so formally, there seemed to be an implicit hierarchy of recovery. First and foremost was picture recovery. If we recovered nothing else, we were in it for the pictures this time. Next was the tracking data. We wanted to know where it flew and how high. This was my responsibility and we had gotten about two-thirds of what we wanted (remember the laptop reboot, grrrr!). Next was the temperature data. We had installed a Thermochron on the unit, a device that included a thermometer, a real-time clock, and a data logger to record the time and temperature all in a package the size of about 3 watch batteries stacked one on top of the other. The Thermochron was missing. A closer look showed that the boom that the repeater antenna was mounted on was missing. The antenna was still attached through the module wall to the radio but the boom was missing. And the Thermochron was mounted on the boom. Things obviously got a little rough up there. As we were getting everything sorted out and put away, someone asked why the audible alarm that I had put on the unit wasn’t beeping. I was busy doing something else and sort of explained it off to the extreme cold causing the alkaline battery to fail. Only later did I actually go back and check to find that the audible alarm along with the wiring and battery were missing as well. Since the battery was tucked inside a pocket on the tracking module and the piezo alarm was connected with a cable tie, it must have gotten REALLY rough up there. We recovered the rest of the experiment module including the simplex repeater, which we found out later worked brilliantly. Parachute and packaging were the last in the recovery list and these were mostly intact. We did notice that the fabric on both sides of the tracking module where everything was connected was torn right around the connection points.

That was our cursory inspection. We collected all the hardware and headed back to Perry, FL, mom and dad’s home town and Launch Control Central to gather at the local Barbeque restaurant for lunch. Mission Accomplished.

What we learned

One of the first things we tried to ascertain was what happened to the antenna boom, thermochron and audible alarm. Once we took a look at the pictures from the mission and realized that the balloon, once it burst, did NOT passively fall to the side of everything and float down on the parachute. We had used about 30 feet of line to connect the top of the parachute to the balloon. The thinking was that when the balloon burst, the parachute would eventually begin slowing the entire package and the 30 feet of line would keep the remnants of the balloon well below the parachute and payload. This was obviously not the case.

According to the photographs and final inspection of the payload we think that the balloon connecting line became entangled in the parachute shrouds. This would have caused the entire package to spin uncontrollably. Obviously, the parachute worked somewhat, and thankfully, it worked best toward the end of the fall to earth. We think this is why we could only get time and position data while the package was coming down, it takes a solid lock on 4 satellites to get altitude data and if the package was spinning, and especially spinning at a steep angle the GPS antenna would be all over the place. At the same time, we think that the shredded balloon was flailing all around the experiment and tracking modules, ripping off the antenna boom and audible alarm.

I noted above, with great consternation, that I had failed to restart the recording function when we did the laptop reboot. Here’s one for the power of the internet. A couple of weeks later I was cruising the internet and looking for my dad’s web page. I used google to find my dad’s callsign (KJ4ZI) and was surprised by the large number of hits I got. Out of curiosity, I plugged my own callsign (N4OSB) in just to see what it would return. To my surprise, there were many references to N4OSB-11, the callsign I had assigned to my tracking module. After a few minutes of searching, I found a German website, hosted by a German amateur radio operator, DB0ANF, that had every byte of data that had been received by any of the APRS receivers within range. As a result, I was able to retrieve the data that I had missed and we were now able to apply that data to the rest of the mission. We were able to correlate the altitudes with the photographs and we are now able to tell how high most of the photographs were taken from, the rest can be intelligently interpolated from the rest of the data.

And our ballast? The dirt that I dug up to provide our 1 pound 6 ounces of ballast had a few ants that went along for the ride. The results: Ants survive very well in the rarified air and extreme cold of near space.

Plans for Next Year

Next mission, we would like to have 2-way communications with the balloon, send commands in real time as well as receiving GPS data. This is done routinely in other balloon projects so it’s just a matter of deciding the best configuration for us. Additionally, I would like to set up a camera with a compass interface to show the direction the camera is pointing when the picture is taken. A declinometer would be nice so we know the angle the picture was taken at. There were many all white, and all black pictures in this set and it would be interesting to know what angle the camera was shooting from. Perhaps we could include that in the data that we could transmit in real time. We’ve already had requests to include a video camera of some sort as well.

The mission this year was our second launch but our first successful recovery. We still understand that there’s more to learn and iron out. Hopefully, in future launches we will continue to increase our data collection and learning opportunities.

Comments or Questions? Email webmaster@barntek.net