Mission Patch







Payload #1: the 'Margaret-Myrtle'

October 24, 1998









Our first balloon project was rather crude and simple, done more as a 'proof of concept' exercise than actual scientific mission.

In the diagram at left, you can see the basic components of the payload. The gondola chassis was made from a foam box originally used to ship grapes to a supermarket. Foam makes an excellent construction material for gondolas, since it is very lightweight, strong, cheap, easy to work, insulates well, and will cushion the components somewhat during the impact of landing. Inside the gondola at the bottom of the compartment were the two main battery packs. Each consisted of two 'D' size lithium batteries, wired in series and wrapped with black electrical tape. These batteries were taken from larger packs originally intended for use by the military and bought as surplus. Each cell provided 3 volts, and hooking the two packs in series provided 12 volts. Nestled in between the battery packs was a Samsung 35mm fixed-focus auto-wind film camera. This was aimed straight downward, and a viewing port for it was cut in the bottom of the gondola. The shutter of the camera was triggered by a Futaba radio control servo, which was in turn controlled by one of the output channels of the microcontroller. The 6 volts needed to power the servo was tapped from one half of the main battery pack. Above the camera and batteries was situated the 'brains' of the payload, a Parallax Stamp-II microcontroller. The microcontroller received power from a 5 volt fixed regulator connected to the main batteries through a single-pole switch, accessable from outside the gondola. Aside from the camera servo, the microcontroller was also connected directly to a piezo electric speaker, the audio input of the radio transmitter, the push-to-talk circuit of the transmitter, and the dust collector experiment. The speaker was used to provide audio cues, allowing us to determine if the microcontroller was 'alive' and functioning properly, as well as help trackers locate the package if it was near-by but out of sight (such as stuck in a tree). The radio inteface allowed the microcontroller to transmit a Morse code signal that the radio hams tracked, identify the package, and will in later missions be used to transmit telemetry. The radio transmitter was an Alinco model DJ-C1T FM VHF (2 meter band) transciever. This radio packs a digital tuning circuit, antenna, tranciever, and lithium-ion battery in to a package 10 by 56 by 94mm, and weighing only 75g. On the exterior of the gondola was mounted the dust collector experiment and its power supply. The battery pack for the collector consisted of four more of the same lithium cells that the primary pack was made from. The collector itself was made from four rods forming a long, skinny pyramid. At the base of the pyramid was a 12 volt cooling fan. The pyramid was covered over with sticky-tape, with folds of tape crossing the insides as well. The function of the collector was that when the fan ran, air was drawn in to the base of the pyramid and flowed along the sticky surface of the tape, hopfully trapping any particulate matter of interest on the tape. A nylon screen over the fan kept any large objects from entering the collector and jamming the fan. The microcontroller switched the power to the colletor via a 5 volt reed relay (the same model relay was also used to control the push-to-talk circuit of the transmitter). A control and sequencing program was hastily coded late the night before the launch by yours truly, and went like this:


Admittedly not my most brilliant piece of code, but it got the job done and was small (microcontrollers don't have very large memories). All events were based on elapsed time, since we didn't have an altimeter or other sensors for this mission. The mission profile was based on a computer simulation written by John Graham, and assumed a total flight time of about 1 hour and 45 minutes, to a peak altitude of 100,000 feet.

The balloon was filled from a tank of compressed Helium (which, although twice as expensive as hydrogen, is still considerably cheaper than a stay in a hospital burn ward) through a fixed regulator and an aglomoration of compressed air and plumbing parts. The neck of the balloon was tied with strong cotton string in two places, folded, then tied again. A piece of cotton string approximately 15 feet long was tied to the loop formed by the folded balloon neck. The other end of this string was attached to the apex of the parachute. The parachute shroud lines were tied to another cotton string (doubled for strength), which was attached to the gondola sling. This doubled string also had several pieces of red crepe paper streamer attached for visibility in tracking and recovery. The sling that attached the doubled line to the gondola was made from several long strips of duct tape, and also helped hold the sides of the payload together. The extra room in the payload was filled with foam packing 'peanuts', to help keep the electronics warm (it can get as cold as -90F at high altitude) and cushion the landing. An image of the complete balloon, taken just after launch, is availiable Here.

The name of the balloon, "Margaret-Myrtle", was in honor of the grandmothers of Blaise Mibeck and myself. Both these ladies passed away during the year, and we requested that the first payload be named after them. The others of the group graciously agreed.

For the rest of the story, I'll cut and paste an email I sent to my friends explaining the events of the launch, chase, and attempted recovery of the payload.
Email from launch day (edited) - all photos courtesy of Blaise Mibeck

Since about two years ago, John Graham and I have been kicking around the idea of building and launching science payloads on small weather balloons. Other groups of ham radio operators around the country have done this for years, and it sounded like a fun thing to do. John finally got fed up with all the "we should do this" talk and decided to 'just do it'. He bought several 1200 gram latex weather balloons, as well as a bunch of military surplus lithium battery packs. This was based on the info we gathered from the web pages of other hams who have already done this. Lithium cells have the highest energy/weight ratio of affordable batteries, and are not as bothered by cold temps as other types of batteries.
Deciding we needed to add more people (and thus, expertise) to the group, we brought in Chris 'Milly' Milford to handle the coordination with other radio hams, as well as Blaise Mibeck to design a physics experiment for the payload. Milly brought in some more hams to form the core of the design and chase team. In about three meetings we had the design hammered out: the first balloon payload would consist of a 2 Meter radio beacon for tracking, an automatic 35 mm camera to snap pictures, and a dust collector to sample high altitude particulate matter. Our FAA liason, George Kelley, provided a copy of the FAR 101 rules regarding balloons and rockets, and told us that if the payload was less than 4 pounds in mass, we didn't need to file NOTAMs or get clearance to launch, provided we launched at least 5 miles away from any airport and the first 1000' of ascent wasn't over a populated area.
The weather didn't cooperate for the first few launch windows, which was fine since we needed more time to build the payload. I, of course, procrastinated due to the delay and ended up starting on my part of the project 3 days before the launch date. My part consisted of dissassembling the battery packs and building the packs we would use for the flight, constructing the camera mount and servo to activate it, building the radio-to-computer interface to allow the onboard computer to transmit the tracking signal, designing and building the interface that would control the dust collection experiment, and programming the onboard computer itself. I was also responsible for buying and assembling the parts of the balloon inflation assembly. The other hams found and tested parachutes for payload recovery, located an ideal styrofoam box to house the payload in, and planned their chase strategy. Blaise built the dust collector, Milly designed the artwork and coordinated the chase communication frequencies, and John coordinated everyone (and paid for just about everthing. Thanks, John!).
By thursday night I had the radio interface done, and that was about it. I used a piece of corregated cardboard for a base, and secured the componenets to it via hotglue. Wiring was done point-to- point, dead-bug fashion (componenets on their backs with their leads sticking up). Probably one of the ugliest pieces of electronic work I've done, but it was functional. By friday I was rather concerned. The weather forecast was beautiful (ideal, really), but I hadn't gotten all the hardware together, nor had I written one byte of code for the controller. John and I went to a welding shop to buy the helium and a regulator. We decided on a cheap regulator designed for filling toy balloons (it was $40, versus $100 for a nice regulator), and wrestled a 100 lb tank of helium into the back of his Taurus (not an easy task, let me tell you!). We all agreed to meet that night to do the final assembly, and launch the next morning.
After work that day, I headed out to Menards to buy stuff to build the filling adapter. The regulator had a tiny nozzle for filling toy balloons, but that unscrewed from the regulator to reveal a 1/4" standard pipe socket. I got a 25' long air hose with 1/4" fittings on each end, a 1/4" to 1/2" adapter, a 1/2" gate valve (water valve), a 3" x 1/2" pipe nipple, and a 1/2" thread to 1" glue PVC adapter. These were all screwed together (with a goodly amount of teflon tape added) to make the filling adapter.
The electronics section was not as easy to assemble. One relay controlled the radio transmit function. I added a second one to control the dust collector. I also added a piezo speaker to the parallax basic stamp II microcontroller to allow it to make beeping noises. This would allow me to know the payload was working when I switched it on, as well as providing a way to find it on the ground if it was hidden in brush, etc. The servo for the camera mount was also added (the stamp can run all these things directly, with no interface circuitry). I added a voltage regulator to step down the battery power, and was ready to test. Of course, it didn't work properly. The servo wouldn't turn the right direction, the radio wouldn't send the Morse code id, and the program I hacked together wouldn't loop. After about an hour of frustration, one of the other hams (Charlie Hofferber), looked things over and told me the regulator I was using needed at least 3 volts over the voltage it was supposed to produce to work correctly. Since I was running 6 volts into it and wanted to get a steady 5 out, this wasn't the case. I added another battery pack, and viola! It worked like a charm. Next time I won't buy cheap regulators! The camera servo was tested, being closly monitored by Blaise and Milly. Both were watching it intently, about two feet from the front of the camera. When the servo worked, the camera fired its flash right into their faces, causing Milly to spill his tea all over the place (but luckily not on the payload). This provided some much-needed amusement. The Morse code on the radio signal was very weak (I had made an error building the audio interface), but it was too late to correct it (it was 11:30 pm by then), so we decided to go with what we had. I returned home, built the battery pack for the dust collector, and got to bed by 12:30.
Saturday dawned bright and nearly clear. Winds were negligable, and the temp about 50F. The launch site was a park south of Walmart, on Columbia Road. I arrived about 8:15 am to find others there already milling around. We spread out a tarp over the grass to protect the balloon, spread out the balloon, and managed to carry the gas bottle through the muddy ditch out to the launch site without any injuries or cursing. The regulator was attached to the tank, and the balloon was secured to the filling adapter via an automotive hose clamp. John donned a pair of cotton gloves to keep from damaging the balloon, and started filling it. [Photo] The rest of us attached the dust collector to the outside of the payload [Photo] , made a sling out of duct tape to hold the payload, and attached the various strings and parachute to it. When the lift of the ballon was judged to be 'adequate' [Photo] , John turned off the fill valve and waited for us to finish preping the payload. The job of holding the balloon was turned over to Tom Mote's son, who is about 3 years old. Seeing a little kids reaction to getting to hold on to a balloon about three times his size was cute (kind of like the apes in '2001' staring in awe at the monolith). [Photo]
Finally, at around 9:15 am, everything was ready. We tied the balloon shut, tied the payload to it, decoupled the filling assembly from it, eased the balloon up to take up the slack in the line, and released the payload! [Photo] It promptly came back down: there was insufficient lift. Ugh! We hurridly hauled the balloon back down, cut the strings free, clamped it back to the filler, and resumed pumping gas into it. We improvised a test weight out of the gas cylinder cap and some tools, [Photo], and when the balloon would lift that, we tried again. After double checking all the switches and camera lens, at 9:27 am CDT, we launched the balloon. [Photo] [Photo] [Photo] We all watched the balloon take off with an ascent rate of about 1000' per minute. After about 5 minutes, the hams on the chase teams took off, while the rest of us cleaned up the area and got the gas tank back in Johns car. The launch team then did what we thought was sensable, which was go and eat breakfast. After that, we took the helium tank back to the welding shop, secured maps and radio, and went out to find the chase teams.
John and I drove about three hours around northern Minnesota, but were never able to contact either the chase teams or balloon. While we headed back to Grand Forks, I finally contacted Milly on the repeater. When asked if they had recovered the payload, he replied that they were having a meeting about what to do about that. After a few more evasive responses, we found out that they had in fact not recovered the payload. It turned out that rather than shooting up to 100,000' and bursting like it was supposed to, the balloon had stabilized in altitude and cruised merrily along. The chase teams had it in visual range until about 12:05 pm, when it went behind a cloud. The beacon signal then stopped transmitting (either dead batteries or the balloon burst), and when the cloud cleared, the balloon was nowhere to be seen. [Signal reception reports]
Hams and the sheriff in the local area were alerted, so that if anyone came across the payload they would know who to contact (we also put a phone number and notice of a $50 reward on the payload itself), so there is still a chance we will get it back if some farmer or hunter finds it. I don't know if the dust collector will still be useful after a long stay out in the rain, but I'm anxious to see the pictures from the camera if it survived.
All in all, it was agreed that the project was a success. Our goals of seeing if we could build, launch, and track a balloon and payload were met, and everyone had a great time building and chasing the payload. We documented what bugs and flaws in hardware and proceedures we'll need to fix next time, and we got signal reports from hams as far away as South Dakota, who picked up the balloon beacon with no difficulty (not bad for a little 300mW transmitter). Plans are currently to build and launch the next payload in about 1 to 2 months. We'll see if we can get the kinks worked out.
At the date of this writing (November 13, 1998), the balloon has still not been recovered and is considered lost. While this is dissapointing, the main missions of this first attempt: the design, construction, launch, and tracking of a scientific ballon, were all successful. Even the pros don't get all their payloads back, and we would have if the balloon had gone to its proper altitude and burst. This malfunction was our fault (insufficient gas), and will not be repeated next time. A cut-down mechanism has also been built for the next payload, which will allow us to seperate the balloon and gondola at our command (and/or after a pre-set time or event by the microcontroller), which should also increase our chances of successful recovery next time. Even with the loss of the payload from the first mission, everyone involved in the project enjoyed themselves and learned something.

Return to the UND HABP webpage.