APRS - 144.390 MHz
Bob - VE5RGM Bruce - VE5BNC Gus - VE5SPI
Spring has finally come to Saskatoon and like the world outside, our balloon project is about to bloom. Each of us has been working on our parts of the project and things are starting to come together. We even have a name. SABRE, Saskatoon Amateur Balloon Radio Experiments.
This article will describe the design and construction of each of the components of our balloon. They are:
Payload Capsule (VE5BNC)
We did quite a bit of research on what other people were using to carry their payloads. The designs ranged from custom built, fibreglass reinforced capsules to insulated lunch bags and everything in between. We had already decided that we would want something that provided both physical and thermal protection. Rigid Styrofoam seemed to fit the bill and I just happened to have a sheet of pink 2 inch thick R10 insulation in the closet. A big shock came when I weighed it and found that it's about 110 grams per square foot. This meant that the 1 foot cube capsule would weigh over 500 grams once the joints were taken into account. Since this is our first flight, we decided to err on the large size and hope that the extra weight wouldn't be a problem.
The capsule is built from 4 identical panels, each 12 inches high and 11 inches wide. The corners are joined with rabbet joints and secured with epoxy. This creates a box that is 12 inches on each side with an 8 inch square cavity inside. The bottom is simply an 8 inch panel square that slips into the bottom of the capsule. It's held with epoxy and 1 small bamboo dowels. The lid is a bit more complicated. I started with a 12 inch square panel. I then removed a 2 inch wide and 1 inch deep channel around the outside edge. This created an 8 inch square area in the middle that would slip into the cavity in the payload. To provide a bit more protection and to make our payload more visible, the entire box is covered in fluorescent orange nylon.
Once I knew the location of the digital and video cameras, I cut for them. The digital cameras look out through a cone shaped hole in the wall of the payload. The ATV video camera is built right into the wall of the payload and looks down at a 45 degree angle. The surface of the holes is painted flat black to reduce glare. I also mounted the external temperature probe so that it protruded about a half inch through the side wall.
A lifting harness was made from light nylon webbing. The straps cross under the bottom of the payload and one strap runs up each side. The straps wrap over the lid to secure. An additional lifting loop is attached to each strap. These loops will be connected to our lift lines which connect to the parachute shroud lines.
Capsule Lid Drawing
Capsule Showing Camera Ports
Capsule Showing Video Camera Port
Capsule Inside with Modules
Balloon and Parachute (VE5BNC & VE5SPI)
The balloon is a surplus military balloon purchased on eBay. It was made with a large skirt which we removed. This saved over 1 kilogram of weight. In the end the balloon appears to be equivalent to a Kaysam 1200 gram balloon. The parachute is a 72 inch model purchased from Spherachutes. They custom make each parachute and ours is made from 2 panels each of fluorescent green, fluorescent orange, fluorescent pink, and fluorescent yellow. It one ugly parachute but it's sure to be seen.
Gus took on the task of building the filler hose assembly. We acquired
a helium regulator from eBay but we needed to adapt it to fit the neck of our
balloon. A bit of checking revealed that the outside diameter of 3/4 inch
PVC electrical conduit was the prefect size and I just happened to have a piece.
Gus then picked up a 25 foot hose designed for an air compressor. Anyone
that has worked with air compressor fittings, electrical conduit and plumbing
fittings know that none of them seem to use the same size of fittings.
After a bit of mixing and matching at the local hardware store Gus has something
that would connect the compressor hose to a plumbing valve and then to the
conduit. It's a combination of brass, copper, steel and plastic, but it
Camera Module (VE5BNC)
Last year I picked up two Polaroid iZone 330 digital cameras from a local surplus outlet. These cameras are small, reasonably light, take good daylight pictures, have internal Li polymer batteries and are easily modified for remote control. Each camera has a 1GB SD card which will hold about 700 still images or 1 hour of video. The modifications involved removing the flash circuit board and installing a 10 pin connection that connects to various points in the camera. This allows the controller to turn the cameras on and off, switch from digital to video and even supply power to recharge the internal battery.
The controller is built around a Microchip 16F84 microcontroller and two
supporting logic chips. I could accomplish the same function without the
supporting logic chips but that will have to wait for the next generation.
The controller programmed for the sequence below. The end result is that
the cameras will take one still picture each minute and a 5 minute video every
15 minutes. The controller also takes care of the audible siren.
This is a simple 110dB personal panic alarm. The controller is set to turn
on the siren for the 12 seconds it takes to take a picture. It does this
after 3 hours avoid triggering the peizo element in the extreme cold and
possibly damaging it.
Still camera cycle
Turn on still camera and take a picture
Turn off camera (12 seconds has elapsed)
Wait 48 seconds
Video camera cycle
Wait 15 minutes after power-up
Turn on video camera and start taking video
Wait 5 minutes
Stop taking video and turn off camera
Wait 10 minutes (optionally provide charge voltage for battery)
Controller Source Code
Camera Module Front
Camera Module Back
OpenTracker APRS Module (VE5BNC)
When the decision of what kind of tracker to fly came up, we just couldn't decide so we chose to fly two separate tracker. I wanted to use an OpenTracker because it would allow me to measure internal temperature and battery voltage during the flight. It also has the capability of switching between two configurations base on several parameters including the temperature and voltage readings.
Next I needed to build the transmitter. Like any good ham I have lots of "spare" equipment laying around. In my heap I found several Motorola MX300 series hand helds. I started with a 6 watt MX350. I was able to remove all of the unnecessary parts like the case, speaker, and so on but the resulting radio was still heavier than I wanted. I then turned my attention to a 2.5 watt MX330. Once it was stripped down it was significantly lighter. Some basic calculations showed that the extra range of the 6 watt radio would be marginal and wasn't worth the extra weight and power consumption so I stayed with the 2.5 watt radio.
Now to decide on the GPS. One option was a Rockwell Jupiter board and external antenna. I also have a couple of Rand McNally "mouse" style GPSs. Research showed that both of these would work above 60,000 feet. A quick test of current consumption showed the Rand McNally GPS was the clear winner. Not long after Christmas, I stumbled onto a Garmin GPS18. This GPS will also work about 60,000 and consumed about half the power of the Rand McNally unit.
I also wanted to track the external temperature during the flight. The OpenTracker has one unused analog input. By purchasing another temperature sensor and duplicating the existing biasing circuit I was able to have a second, external temperature sensor. The sensor was glued into the tip of a piece of 3/16 hobby tubing. The tubing was then mounted so the tip protruded through the wall of the capsule.
Now it was time to decide how to power it. I wanted to use lithium batteries but wasn't sure about which type. I finally settled on 3 standard CR123 camera cells in series. They are rated at 1300 mAh and can provide as much as 3 amps for short periods. My power budget showed that these would provide over 6 hour of run time for my tracker.
Now the only piece left was the antenna. I first considered some type of dipole. Mounting options included a sleeve dipole above or below the capsule, side mounting ad even a horizontal dipole. I couldn't come up with a mounting system that I was happy with so I turned my attention to a 1/4 wave vertical. The big drawback of this was how to build the ground plane. Standard 20 inch radials or ones draped over the sides would almost certainly get tangled in something. Although I had never seen it, I wondered about folding the radials in such a way that they would fit into the area of the lid. I ran some designs for folded radials through antenna analyzing programs and they all showed that it should work fine. After a bit of design work I was able to get them to fit into the payload lid. The Capsule Lid Drawing above shows the details. The radiator is simply a 19 inch piece of spring steel from a hobby shop.
PocketTracker APRS Module (VE5SPI)
Gus decided to follow in the footsteps of many other balloonists when it came to his tracker. When you're 30 kilometres above the ground a little bit of signal goes a long way. To him, this meant that the Byonics PocketTracker would be the ideal APRS tracker. It's no longer available but this is a tracker kit built around the Byonics TinyTrak III. It puts out about 300 mW and is designed to fit into an Altoids mint can.
When it came to the GPS, Gus also has a Rand McNally GPS that he uses in his vehicle. He will be retrofitting this to connect to the PocketTracker.
Gus chose yet another antenna design. He will be using a 2m dipole made from RG174 coax. The antenna is made by folding about 20 inches of the braid back along the cable and leaving 19 inches of the center conductor exposed. This antenna will be mounted on the bottom of the payload and will trail below the payload.
Gus had settled on two power sources for his tracker. Since the PocketTracker was designed to have a standard 9v battery inside the Altoids can, he will using a lithium 9v to power the tracker. Rather than build his own battery pack, he decided to take advantage of the extra capacity in Bob's pack and will use this to power his GPS.
ATV Module (VE5RGM)
Bob has been involved in ATV for over 10 years. Although he has a portable ATV station, retrofitting this to work in our balloon wasn't practical so he purchased a PC Electronics transmitter from a local ham. The transmitter hadn't been used for several years so we decided to give it complete tune-up. To our surprise, it worked perfectly the first time.
Of course an ATV transmitter isn't much good without a video source. Bob scoured the Internet and eBay for cameras. He settled on one designed for an automobile rear view system. He chose this for several reasons. The camera he chose is waterproof, designed for cold temperatures, and of course, cheap. The only drawback is that rear view cameras produce an image that is backwards. This can be bit confusing down here on the ground but shouldn't cause any problems during the flight.
Bob is an inventor in the truest sense. He wanted to build a 1/4 wave ground plane antenna but found that the ground plane was the most troublesome part. He needed to find something that was light, a good conductor and just the right size. The usual things like pie plates were ruled out because they were simply too mundane. Something like an aluminum pop can would be prefect but they were too small. Surely something came in larger cans. Then it occurred to him. He had an idea but only rigorous research would prove his design. After a few prototypes he had what may just be the prefect antenna. You see, not only is a tall boy Pilsner beer very refreshing, the can it comes in is the perfect dimensions for a 439 MHz ground plane. Just mount a BNC connector in the center of the bottom, remove the top and cut the sides in 4 places. Once you add about 6 inches of brass rod to the center conductor of the connector, you have an antenna that is sure to turn a few heads. The ATV antenna will be secured to the bottom of the capsule.
Bob also did his homework when it came to batteries. The transmitter and camera can draw as much as 400 mA and require at least 10 volts to operate. He wanted to use Lithium cells but the CR123 cells didn't have enough capacity. Lithium AA cells looked promising but most are limited to between 100 and 200 mA. He found some Lithium AA cells that are rated at 2900 mAh and were capable of sustained loads of over 1000 mA. One disadvantage is that AA cells only provide 1.5 volts so he had to use 9 cells.
To make things even more interesting, Bob chose to use the audio signal from the OpenTracker for his ATV audio. We don't know how well this will work, but the experiment promises to be interesting.
73, and happy ballooning.
Bruce - VE5BNC