Blog from October, 2013

Gordon, Oct 5

The Gill windsonic at base was installed with no concern about its orientation.

Today Tom calibrated the data scope, then measured the sonic boom azimuth (with declination=0), getting the following readings:

The magnetic declination here is 10.5 degrees. So the boom azimuth is 36.5 + 10.5 + 180 = 227.0 degrees wrt true north.

Then we removed the sonic and boom from the tower to figure out the orientation of the sonic relative to the boom.

By putting a ruler along the transducers and some advanced trig we estimated that the north arrow on the sonic is 2 degrees clockwise from the boom.

So N on the sonic is at 229 degrees true.

Entered this as the offset angle in $ISFF/projects/METCRAXII/ISFF/cal_files/noQC/dir_base_10m.dat.

Yesterday deployed sonics and added 30W solar panels at SSW2 & SSW4.
Measured azimuths of csat sonics in crater.
Fixed P.rim and csat.15m.rim.
Replaced radiation mote and restarted SPN1 mote at FLR.
Sebastian cleaned radiometers and kh2o at FLR.

Review of cockpit around 11am:

T/RH:       OK
P:          OK, P > 850 mb on crater floor
csat u,v:   OK
csat ldiag: OK, flr was at +1 at a previous time
csat w,tc:  OK, w > +/- 5 m/s at rim
kh2o:       kh2oV.flr = 0; kh2oV.far > 2 V
wetness:    rad.flr, far are RIP
radiation:  rad.flr, far are RIP
Tsoil:      near is RIP
Goil:       flr is < -100 W/m^2, far & near ~ -50 W/m^2
Qsoil:      far & flr = 0, near OK
TP01:       Vheat OK; Vpile on OK, near=1000 uV, others ~ 600 uV; Vpile.off=0; tau OK; lambdasoil are RIP
Spd, Dir:   OK
T/RH:       OK

10/4/13, twh

Sebastian cleaned the kh2o windows, but found out later that the data still near or at zero.  Sebastian will check the wiring on the next trip into the crater.

10/4/13, twh

The radiation mote at FLR was not reporting.  It looked like it was sampling the data okay, but John suggested that the radio might have shaken loose from where it plugs into the main mote board.  I replaced the mote and it began reporting data okay.  I removed the GPS jumper and  brought the original mote back to the base for inspection.

Gordon reported that the SPN1 mote was not sending data.  It appeared to be hung with a solid yellow light in the middle of the board.  I recycled power and that appeared to fix it.

Sebastian cleaned the radiometers and noticed the some require new dessicant.  Something to check for all the radiometers.

10/4/13

Tom measured the orientation of the FLR CSAT:

041.6
041.0
041.3
040.8
041.0

mean: 41.1

Calibrated DataScope and reshot CSAT orientation.

38.3
39.3
39.1
39.1
38.6
39.0

mean: 38.9

I don't know whether this difference of 2 degrees is due to calibrating the DataScope or simply standing in a slightly different spot to align myself with the sonic u-path.

Did not have sufficient time to shoot the orientation of the RMYoung 2D sonic at 10m.

10/4/13

Tom, Sebastian, Matt and Eric.

Installed the two sonics at SSW2 and SSW4 today, after much running up and down the crater sidewall to get the correct cables to plug into the motes (and even up to the rim for a delivery of sonic cables to Matt from Gordon).  Also installed a second 30W(?) solar panel at both sites.  These were scavenged from the solar panel stands at FLR.  FLR is now powered by 4 60W(?) solar panels.

SSW4:  The battery measured 13.87 V.

After installing the sonic, I tried to calibrate the data scope, but was unsuccessful because of the requirement to rotate through eight 45 degree increments within two minutes while standing still on a steep, treacherous slope.  Measured the magnetic orientation of the sonic (no  declination applied):

353.4
352.5
350.3
348.2
350.8
348.8
mean: 350.7

SSW2:  The battery measured 13.66 V.

Sonic orientation (still no calibration):

337.5
337.0
337.2
336.8
337.2
mean: 337.1

12/4/13

From data files:

ssw2 was SN1121
ssw4 was SN0677

Gordon, Oct 4 04

Typically when the base trailer air conditioner kicks on, the lights brown out and an alarm beeps once on the UPS (APc Back-UPS ES 350).

This afternoon, a brownout caused the ap24 WIFI at the tower to reboot.

Currently these devices are on the UPS:

• /media/backup external UPS drive
• AP24
• 8 port ethernet switch
• netgear router (connected to Hughes modem)
• Hughes satellite modem

May want to get a better UPS. The battery in this one is probably shot.

Gordon, Oct 4

Went to rim tower to deliver some CSAT3 cables to Matt who climbed up from the ssw4 sonic.

While at rim, I screwed in the serial connector on the bottom of the Paro6000 barometer, P.2m.rim. It was not fully screwed in.

After returning to the base I noticed that all data from rim had stopped.

Returned to the site. Console was reporting many messages, something like the following:

"Too much work for IRQ 4"

IRQ4 is the second Emerald serial card.

Disconnected all sensor connectors at the interface panel to the emerald ports (5-20). Error messages didn't stop. Removed the ribbon connectors on the inside of the interface panel. No effect.

Could not get a shell prompt at the console. Power cycled at 4:28. System came up. Reconnected all serial cables. All are working, even the 15 meter CSAT on port 7 and the barometer on port 13.

Will have to watch this system. Perhaps it was due to a loose connection of the barometer on port 13, or a questionable serial card or something amiss on the Titan.

Apparently, the only problem today is that the soil mote at FLR has quit.  Is there a spare in the crater?

Also P.rim

ToDo:

Clean kh2o & radiometers at FLR (done yesterday at FAR and NEAR).

Deploy two sonics on the SW slope in the crater.

Shoot sonic azimuths.

Raise radiation stand at NEAR (done yesterday at FAR).

October 3, 9:50 - 10:05

Replaced the dc power cable on the radiation mote at FAR.  Yesterday, Chris jury-rigged a temporary cable.

Noticed that the local usb flash drive was not mounted at "near".

/var/log/isfs/kernel contains the following:

Oct  1 21:37:53 near kernel: usb 1-2: reset full speed USB device using isp116x-hcd and address 2
Oct  1 21:38:31 near last message repeated 4 times
Oct  1 21:38:31 near kernel: sd 1:0:0:0: Device offlined - not ready after error recovery
Oct  1 21:38:31 near kernel: sd 1:0:0:0: [sdb] Unhandled error code
Oct  1 21:38:31 near kernel: sd 1:0:0:0: [sdb] Result: hostbyte=0x05 driverbyte=0x00
Oct  1 21:38:31 near kernel: sd 1:0:0:0: [sdb] CDB: cdb[0]=0x2a: 2a 00 00 28 3b e1 00 00 18 00
Oct  1 21:38:31 near kernel: end_request: I/O error, dev sdb, sector 2636769
Oct  1 21:38:31 near kernel: Buffer I/O error on device sdb1, logical block 329596
...

On Oct 2, at 22:41 UTC, Chris unplugged/replugged the flash drive. It is a flash drive not a pocketec. After an automatic fsck, the file system was recovered.

We'll watch this flash, and replace it if necessary.

The WIFI has been working very well, with dependable connections between our data systems. The check_ap24.sh script on flux reports the following today, Oct 2 around 14:40:

check_ap24.sh 
local  remote      (         mac-addr)    uptime  ccq  txrate  rxrate  rxsig  snr #txpkts  #rxpkts  txretry  rxretry txB/sec rxB/sec
ap24   ap24-3@near (00:15:6D:20:01:90)  50:30:07 100% 54.0Mbs 54.0Mbs -63dBm 40dB 2337026 2169508       0%       0%     829   12328
ap24   ap24-2@rim  (00:15:6D:10:1C:0D)  22:45:04  64% 18.0Mbs 18.0Mbs -87dBm 16dB 1459062 1446472       3%       0%    2107   14582
ap24   eanth@sodar (00:20:F6:05:24:85)   0:47:21 100% 11.0Mbs  5.5Mbs -57dBm 46dB    4124    4333       0%       0%     887    1031
ap24   eanti@far   (00:20:F6:05:24:56)   0:00:47 100% 11.0Mbs  2.0Mbs -66dBm 37dB     170     170       0%       0%     217    1186
ap24-2 ap24@base   (00:15:6D:10:3C:BC)  22:45:05  47% 18.0Mbs 36.0Mbs -79dBm 19dB 1465417 1440073       6%       0%   14582    2107
ap24-2 eantf@flr   (00:20:F6:05:24:5A)   0:14:40  94% 11.0Mbs  5.5Mbs -46dBm 54dB    3251    3271       0%       0%     222    1493
ap24-3 ap24@base   (00:15:6D:10:3C:BC)  50:30:09  97% 54.0Mbs 54.0Mbs -63dBm 36dB 2205848 2300696       0%       0%   12328     829

ap24 is as the base trailer. ap24-2 is at rim, ap24-3 is at near.

The connection between base (ap24) and near (ap24-3) looks very good, up 50 hours, with 40 and 36 dB SNR.

The connection between far (eanti) and base has also been solid, but the above shows only 47 seconds since the last association. Etherants seem to drop and re-associate fairly frequently.

eanth@sodar has been a bit problematic, requiring power-cycling of the etherant from time to time. Gary re-oriented it yesterday, and I think it has been more solid since then. The SNR looks good.

The connection between rim (ap24-2) and base has a low SNR of 16/19 dB, but has been up for 22 hours, at a lower tx/rxrate than the connection from near to base. This is probably low because the yagi on the rim tower is connecting to the back of the 180 deg yagi at the base trailer. To see if it improves things I just enabled the internal patch antenna on the ap24, which is pointing at the rim tower. I'll wait until the radios re-associate and see what the signal levels are. Both the internal patch and the eternal 180 degree yagi are access points for the SSID of ISFSext. I am also curious to see if that works.

10/2/13, TWH:

Cleaned krypton and radiometers.

Vmote.rad reported discharging, but battery properly wired and measures 12.2 V both at the battery terminals and the battery connector on the mote.

Unfortunately, there were a couple of wisps of smoke when I tested the voltage in the mote and afterword the mote was not sampling properly (based on lights).

Left site at 13:25.

Comment by Gordon: The radiation mote at far is ID16. Looking at the raw data I see that the data quit at 19:00:43 UTC (13:00 local) (with an Xbee-reset at 18:59), resumed for 2 minutes at 19:11 and then quit.

Later at 15:10 to 15:35 Chris and Tom raised the radiation dark horse to 2.02 meters, measured to the top of the horizontal rail. Radiation data should be discarded during this period.

Chris removed the GPS jumper from the radiation mote and it appeared to be sampling the data okay.  When we got into radio contact with Gordon, however, he said that the radiation data were not coming in.  We returned to FAR to replace the mote, guessing that the its radio had failed.  We found that the power connecter on the mote was fused.  Chris cut off the power cable at the mote and hard-wired it to replacement mote.  Seems to be working now.

Tom at near tower, Oct 2, 2013

Arrived around 12:35.

Cleaned radiometers, which were pretty clean, with fingerprints on SPN1. See an upward spike in Rpile.in at 12:44.

Cleaned krypton, signal rose from 0.12 to 1.2 V by 12:48.

Installed pressure port. All P.2m.near data up to this time should be discarded.

Left at 12:50

Returned to NEAR to raise sawhorse, but Chris could not remove one of the stakes without a crowbar.

Gordon, Sep 30

Laura, Chris, Kurt and I instrumented the rim tower

The 0 end of the measuring tape was at the top of the tower, with the 40 meter indicator at the base plate. The first column contains the values read from the tape, the second column is then (40-x) or the height above the ground: