From the edge of space a tiny camera captures the dramatic curvature of the Earth, during a test flight that is one small step for Cambridge University students aiming to launch a rocket into space for under £1,000. - The Guardian
Cambridge University Spaceflight is a student run society founded in 2006 comprising undergraduates and postgraduates from many disciplines. We aim to develop the technology needed to reduce the cost of sub-orbital access to space for scientific research, in the form of high altitude balloon launches, designing rockets, and other related experiments.
Current radios on High Altitude Ballons (HABs) are laughably slow, on the order of 50 bits/s. While this can suffice for transmitting simple data strings such as GPS co-ordinates, it’s useless for anything else. There have been several attempts at in-flight image transmission for amateur HABs with existing radios, though these took around 6 minutes to transmit a single small grainy image.
Lynx is an experimental digital radio transmitter for HABs, capable of data rates an order of magnitude above existing amateur systems. The end goal of the project is to have a live video stream coming from a balloon in flight. It features a powerful ARM chip to deal with the error correction codes and signal processing.
This development board will allow a number of the key radio systems to be thoroughly tested before implementation in the first flight model. Currently, we are waiting for some RF components to arrive from the US before the board is assembled.
A while ago, the nice folks at Bronkhorst gave us a helium-compatible gas flow meter on loan. This shiny (literally) bit of kit allows us to deliver precise quantities of helium to our high altitude balloons, hopefully increasing the accuracy of our predictions. The flow meter needs control from a PC (with an RS232 port!), which is inconvenient when in the middle of a field.
Skunk is a project to use an Arduino as the controller for the flow meter, meaning we can put it all in a nice case with a battery and control panel for field-filling. A key part of this is the interface PCB – an Arduino shield with all the random circuitry crammed on. After a few weeks of design, the wonderful, amazing, people at the Cambridge Circuit Company fabricated a beautiful PCB for us – pictures of the bare and made-up boards are below! (Points for spotting the design mistakes I’ve had to correct – thankfully none were show-stoppers). There’s still a lot left to do on Skunk, but for once it’s making solid progress. One of our new members will be developing the firmware in the new year, once the hardware is all hooked up.
The top of a finished Skunk v1 board
The bottom of a finished skunk v1 board
Front of a bare skunk v1 board. Photo by Cambridge Circuit Company
Some of our team have recently been developing a prototype hybrid rocket motor. Our motor was to be ignited by lighting a length of slow-fuse, waiting for the fuse to burn down some of the length of the fuel, then opening a valve to allow the oxidiser to flow through the fuel. The oxidiser valve requires around 10 amps continuous current, which is rather more than most ignition systems are geared up for.
OkGo (Named after the band) is our ignition system, developed to light the hybrid (We also used it to ignite a number of solid fuel rockets). It’s a very simple system based on buttons, switches and relays, but works well and should be highly reliable. We are considering adding features such as continuity testing (This lets you test whether the ematch is intact and connected before activating it).
DorMouse (pictured below), is a flight computer, designed to go in the nose cone of a model rocket. We built DorMouse for the Sunday 2nd Rocket launch at EARS, though plan to continue to use it for low to medium level model rocket launches. It has a large suite of sensors, including 2 accelerometers (16g, 200g), a barometer, and a uBlox GPS, all of which are logged to an SD card, with a polling rate for most of the sensors of at least 500Hz. There’s also an XBEE on board for coarse live data at the highest rate the signal strength allows. DorMouse is controlled by a 72MHz (STMF103RFT6) ARM CPU.
Actual size: 4×7 cm
Headers shown are for an XBEE
We are particularly grateful to the Cambridge Circuit Company for delivering the first revision of the DorMouse PCBs, one of which was partially assembled for the first launch. It was unfortunately afflicted by a software (DMA related) bug which prevented it from flying … though in a rocket whose front section was subsequently lost – a near miss!
We’ll have an update once the full board is assembled, and we plan to do an even smaller revision 2. Other fun future ideas include adding a precise timestamp to all the logged data, using the GPS unit for timing.