In all of these sets of data, I record at 100 samples/second and save 50 samples of data (but only the 8 MSBs of the A/D converter) before launch is detected.
I use Gnuplot to make the graphs. It gives me 2 X axes and 2 Y axes, but I only use 1 X axis.
The gizmo's first flight was last-minute situation. I had finished enough hardware and software to record data, but I hadn't put together a payload compartment for it. I discovered that it fit well in the payload compartment of my scratchbuilt Aerobee 350 (a BT-60 4FNC rocket). I flew it on an Aerotech 24mm F24-4. Ejection was before apogee (as I had expected).
I didn't know exactly what a launch would look like, so I used a mechanical switch (like I'd used before) to trigger recording. I was very happy to see good data, allthough the accelerometer clipped the op-amp's output.
The values on the Y axis are the raw values from the A/D converter. Later I'll convert these to Volts and Gs/kPa.
After one good flight, I got rid of the mechanical switch and watched the accelerometer to see when launch occurs. I also changed the accelerometer's op-amp from an LM348 and the resistor values I got from a Circuit Cellar article to the current LMC5842 and set the scale to +- 20G.
I now require .2 seconds of 1G+ acceleration to detect launch. Thus, I get 30 samples of pre-flight data and the first 20 samples of flight data have only 8 MSBs (most significant bits). I added a second SEEPROM and discovered that I was only using half of the first SEEPROM. I now record 27 seconds of flight data.
This was flown in a scratchbuilt Aerobee 150 (a 3" cardboard tube 3FNC rocket) on an Aerotech H238T-M.
My third flight was a matter-of-fact sort of flight; I hadn't made any changes, but a club launch was upon me. I flew it in the Aerobee 350 again, on another F24-4.
All these charts are pretty, but I'd like some real information about the flight.
In this chart, I convert the accelerometer's raw reading into Gs and use that to calculate the rocket's velocity. From the velocity, I can calculate altitude. This is where I really wish Gnuplot had a 3rd Y axis scale. Instead, I'm making two plots with the same X scale.
On the lower chart is acceleration and velocity. The upper chart shows distance, as calculated from the accelerometer data and two altitude graphs: one using a pressure/altitude table that uses linearization to calculate altitude between table points. The second pressure-based graph uses the calculations described on the Hardware page for each point.
I cut this graph off at ejection. The accelerometer data isn't useful after ejection, other than to watch all the events.
Finally, a plot of the pressure sensor readings for the entire flight. The first two are the same as above (pressure/table and pressure/formula). The third plot uses the altitude calculation formula from the Programmin' by Pete web site.
Note that this launch site is at 900' above sea level. These graphs show it is lower, which I assume is due to local barometric pressure; at the time this flight was flown, local pressure was 29.93 in Hg or 1013 hPa.
The last flight was on 8/6/2002. Here is the pressure and acceleration from that flight.
Here are the Gnuplot scripts I used to create these charts:
Next: References
YAFC Home
Hardware
On-Board Software
Flight Results
References
Home
YAFC