First here is a plot straight from the AltAcc GUI. Which I am beginning to hate. It has this habit of puking on the communications with its graphing package. There is no obvious way to cure the problem other than to mouse around in the graph control window until it starts working again. This was a problem in the version I received in 1999 and the updated version from 2003.
This graph shows some obvious problems and some not so obvious. Wow. I just realized something. The deployment event was commanded by the AltAcc. I guess looking at the data again and again is useful.
Deployment was commanded at 15.5 seconds. This is much earlier than apogee but the AltAcc acceleration data has two significant anomalies in it. The first occurs during the motor burn around 3 seconds and the second around 14 seconds. These acceleration anomalies are accompanied by problems in the pressure data as well. I am fairly certain that these anomalies in the data are the result of Electromagnetic Interference (EMI) from the APRS tracker. If you look at the RDAS acceleration data you will see that it does not record these anomalies so they are not the result of anything physical.
To verify that there could be interference I used the live data mode of the AltAcc software to watch the AltAcc in real time. I then placed the Alinco DJ-C5 transceiver next to it. When I keyed up the transmitter, I could see that the pressure and acceleration readings changed.
This EMI did not effect the RDAS nearly as much. It did produce slight changes in the pressure data. These changes are harder to see during the boost and coast phases of flight but I think I see a hint of one at around 14 seconds.
This is pretty bad but not as bad as the EMI directly causing the outputs to fire. I feel much better now.
I have already correlated the RDAS pressure data anomalies with known APRS transmission times. Can I correlate these AltAcc data problems with RDAS data problems?
While the RDAS pressure anomalies really stand out while it is floating down under parachute, it is really hard to see anything while the pressure is changing rapidly as it goes up. I think I can see just a hint but it could just be my imagination. So I tried to do something that would magnify things a bit. I subtracted the measured altitude from the Kalman filtered altitude. If there were a EMI induced pressure anomaly, it would not be accounted for in the Kalman filter model of the system and would appear as an error. This might magnify any EMI induced pressure problems.
The difference has a lot of noise in it but in addition to the noise is mostly smoothly varying. But besides the obvious spike at deployment time, there a couple of brief bumps in the plot at precisely the same time as the acceleration data problems in the AltAcc.
OK so that covers the early part of the flight. But the RDAS data shows several more post apogee hits in its pressure data. Are there any similar hits on the AltAcc acceleration and pressure data?
This plot shows the AltAcc raw acceleration along with the RDAS pressure altitude.
Although there is a slight offset in time, two large acceleration spikes match up with two RDAS pressure spikes. Two smaller acceleration spikes, hardly large enough to pop out of the noise, also match up with RDAS pressure spikes.
Comparing the pressure data shows very good correlation in the data anomalies.
Aside from the slight time offset, there are five pressure anomalies in each data set and they match up exactly. So the EMI induced by the APRS system is effecting the sensor readings of both altimeters. It just happened to effect the AltAcc worse than it did the RDAS.