I've been testing our 10 infrasound sensors today (all day!)
Cassandra reported that sensor 295 appeared broken (in Sakurajima?). It seems to perform okay now.
I tested infrasounds in sets of three, but keeping red channel (in plots) constant (sensor 282) for comparison.
Tests:
1. Sine wave: By having Alex or Seyi raise and lower all three about 1 metre for about 15-20 seconds. Thanks to my sinusoid signal generators!
2. Door test: Close my office door. Wait a few seconds. Then rapidly open it. Gives a really big signal.
Then from the Maintenance page of the digitizer website, download the windows of test data as seed files. Unix "mv" them to shorter names, then run seed2db. Then use GISMO to load them into MATLAB, e.g.
ds = datasource('antelope', '/Users/thompsong/test/dbinfrabsu')
scnl = scnlobject('USF1', 'BD*');
w = waveform(ds, scnl, datenum(2015,12,7,22,20,48), datenum(2015,12,7,22,22,51))
w2 = detrend(w)
plot(w2)
Consider this an outline procedure for testing our infrasound sensors in the future. Also, it could be really fun to huddle test them sometime by taking them to Tampa airport, or even better, a rocket launch at Kennedy. I might even do that as part of my time series class. We could get some good ground-coupled airwaves too on the seismometers.
Main result - door test works best, and sensor 295 has a response that is too small by a factor of ~15.
More details in this report