Just grabbed and plotted 8 hours of data from today to start looking at things. Got data from stations 2,3,4,9,10,12,13. (Only a partial file from 12, since it apparently was restarted in the middle of the day.)
In general, everything looked okay (see below). I looked at the variables: spd, dir, w, tc.5m (should be changed to .3m), T, RH.2m, Spd, Dir.10m. The only issue was that birds definitely interfere with the Gill at stations 10 & 13 – perhaps at 2 as well. I entered the boom directions for the csats, but not the gills, so I see the Gill at s4 different (but close to correct with respect to north) from the others by about 180 deg, which is expected.
I also note that the vertical velocities are quite small and don't vary, indicating both that the set-up crew did a great job of leveling the towers and that a turbine wake is not visible (expected since the winds were generally from the wrong direction).
HOWEVER, the Gill at s12 is bad. When I rserial to it, I get binary gibberish. I have now checked (using minicom) that this isn't a baud rate issue. I've also verified that "tp4" shows the same settings (120ohm termination, normal slew) as for s13, which is working. (I also played with these settings and didn't noticably change the signal.) I conclude that this is not a problem with DSM settings, but a cabling issue or something in the Gill itself.
2 Comments
Gary Granger
There is one site, s8, where the pre_dsm.sh script sets tp4 422 nt ns. During staging we couldn't figure out why it was reporting gibberish, eg, samples were 8191 bytes long because the carriage return terminator is never seen. Then Gordon discovered that if he set no-termination it worked. So I tried that adjustment on s12 and it didn't seem to have any effect. And I don't think we ever had an explanation of why the Gill on s8 was different.
Do the Gills get setup in the field with the same DSM they were tested with? Maybe the problem on s8 during staging was not with the DSM end of the wiring but with the Gill itself, and maybe that flaky Gill is now attached to s12?
Steve Oncley AUTHOR
John called me with the same problem as they were setting up site 8. Changing from Titan port 4 to Emerald port 6 got the Gill working. We suspect that the same change will fix site 12. On site 8, I changed the port in the local copy of vertex.xml and added an "emode 6 422" at the beginning of ..../scripts/dsm_pre.sh. (Later noted that there probably was a better place in this script for this command.) I haven't yet made the port change on site 12. I did notice that site 12 apparently has an old Emerald board, so "emode" doesn't work – apparently we'll have to change jumpers to enable 422?