STORMFEST92: SONIC Messages: 52 Entries..

GILL SONIC ANGLE: 
 
	The GILL sonic was leveled 
and its angle, along boom, was measured. 
The U axis was align along boom. This component 
should be positive with wind blowing in 
toward tower. The height of this sensor is 
roughly 4.25m. 
 
angle: 2.46 
 
 

4m ATI SONIC 
 
Since the new ATI sonics will not arrive until the end  
of the week, we put the ATI sonic loaned to us for the  
sonic intercomparison onto the E-facing flux tower at 4m,  
along with the Campbell Krypton hygrometer.  Prior to  
putting it up, we adjusted the received signal levels  
(2nd wave to ~10V) and calibrated the sonic in the lab,  
using the temperature probe from the fast T chamber  
to enter the temperature. 
 
 

ADJUSTED 4m ATI SONIC  
 
Yesterday and this morning we noted noise on the v and w components 
of the ATI sonic at 4m.  This morning at 1500 I adjusted the receive 
pulses, lowering the amplitude of the first wave to below 1 V, to  
hopefully eliminate mistriggering.  We will monitor the output for 
a while and then recalibrate if necessary. 
 
 

ADJUSTED 4m ATI SONIC 
 
Sonic was spiking again, so at ~18:15 increased amplitude of  
received signals slightly.  Will monitor. 
 
 

GILL SONIC: 
 
	The A/D gain settings for the channels 
used by the GILL were changed to a gain factor of 2 
with a 5 volt offset. At this time we still do not 
know the range of the #4 signal. It is assumed to 
be the speed of sound with 0v = 0 m/s and 
5v = 500 m/s. 
Prep.config has been changed to agree with the 
A/D changes. 
 

GILL SONIC DATA RATE: 
 
	Looking at a data dump of an 
archive file, it appears the GILL sonic 
is running at 20 hz. They analog signals 
are changing every 20 hz sample. 
 

4m ATI SONIC NOISE 
 
I note now that the noise on the 4m ATI sonic appears to be 
common to all channels, suggesting a common noise source, 
rather than inappropriate amplitude of the received signals. 
 
 

Angle of 4M ATI sonic. 
angle = 91.35 degrees. 
 

4m ATI SONIC NOISE 
 
4m ATI was very noisy last night.  The noise currently appears 
to be synchronous on all channels. 
 
 

KAIJO DENKI SONIC: 
 
Research Unit Model 1012R2, Serial #0055 
 
	The KD sonic was added to the 
network at about 16:00 GMT. The data are 
being collected on cosmos analog channels 
100 thru 103. At this time the tower has not 
been aligned with the theodolite. 
 
channel		signal		gain/offset 
 
100		u		5/-- 
101		v		5/-- 
102		w		5/-- 
103		tc		5/-- 
 
calibration factors: 
 
signal		cal factor 
 
u		.000305185,0.0 
v		.000305185,0.0 
w		.000305185,0.0 
tc		.001525925,0.0 
 

KD ARCHIVING: 
 
	The archiving of the KD sonic 
was not turned on yesterday. It is now 
archiving under inter/KD. 
 

4m ATI SONIC 
 
Sonic is still very noisy.  Noise appears to be synchronized  
on all channels.  Looked at received signals on 'scope, but 
saw nothing wrong.  Note that I also looked at the sonic  
several times yesterday afternoon, and thus there will be 
breaks in the data. 
 
 

KD ANGLE: 
 
	The KD was levelled and the angle 
of alignment was taken this afternoon. 
angle along horizontal axis = 60.08 deg. 
 
Assuming proper internal alignment of the 
KD sensing head, the u axis is pointing 
toward 0.08 degrees. 
 
 

KD PREPROCESS CHANGE 
 
Steve changed the sign of the KD v component (to -) 
to correspond to the Gill (where we also changed v to -v) 
and ATI's coordinate system. 
 
 

ATI SPIKING: 
 
	The spiking on the ati has stopped. 
Who knows why. Time is 23:30 GMT. 
 

SONIC WORK: 
 
	Throughout the day the flux tower will 
be up and down for installation of new ati sonics. 
Therefore the data will be bad. Look for a later 
logbook entry stating data are good. 
 

NEW SONIC STATUS: 
 
	The new sonics are having their 
problems. AT the 4m height, the v component 
is off scale. The same case exist for the w 
component at the 10m height. 
 
 

4m ATI SONIC 
 
Steve put the new ATI sonics up yesterday at 4m and 10m on the  
flux tower.  The output signals from both were poor, so today  
we looked at the received pulses with the 'scope and found  
that they were quite noisy.  We brought the 4m sonic into the 
trailers and it performed good both within the trailer and  
outside on the tongue.  When we remounted it on the flux tower, 
after recalibrating it on the trailer tongue, 
we found that the noise was apparently being generated by the 
fast T.  Steve tried various substitutions for cables and also 
the bridge itself with no improvement.  At that point we simply 
disconnected the 4m fast T in order to get good signals from 
the sonic.  I think that this also improved the 10m ATI signals, 
but they are nevertheless still poor, and disconnecting the 10m 
fast T only made its signals worse!  We have now turned off the 
10m ATI since its output was obviously still bad.  Tomorrow,  
weather permitting, we will see if grounding the sonic booms 
reduces the noise on the received pulses. 
 
 

ATI FLUX SONICS 
 
We tried grounding the sonic booms to the tower this morning, 
but that only made the noise worse.  Then tried grounding the 
sonic electronics box and this eliminated the noise, even  
with the fast T's connected.  We then tried grounding the booms 
again and this brought back the noise, so we now have an  
external ground only between the electronics case and earth  
ground. 
 
 

10m ATI SONIC 
 
We hooked up the 4m sonic cable to the 10m sonic 
electronics and got good data from the 10m sonic, 
indicating that the problem with the 10m sonic is 
with the cable or with the array. 
 
Looking at the 10m ATI sonic received pulses with the  
'scope, the v axis has no received pulse.  Allen  
tightened the crystals on the v axis and this brought 
the signal back.  He  also found most other crystals 
loose and tightened them as well.  We then redid the 
calibration procedure on the tower with the exception 
of the vertical axis.  However, the 10m sonic is now again 
spiking simultaneously on all channels. 
 
 

10m ATI SONIC 
 
Tried switching the 4m probe to the 10m electronics 
again to verify that the synchronous spiking is 
a problem with the array or the cable.  This is the 
case. 
 
 

TESTING FOR NOISE SOURCE 10m ATI 
 
I am searching for the source of the noise that 
is causing synchronous spiking on the 10m ATI. 
I turned off the fast T's and Krypton Hygrometers, 
checking each time for noise on the 10m ATI.  Then 
I turned off the other sonics.  The noise persisted 
when I first turned off the Gill, and then quit 
when I turned off both other ATI's.  The noise did 
not return when I turned on the comp.ATI, but did 
return when I turned on the 4m ATI.  I will monitor 
this for a period and then turn it off again. 
 
 

TESTING FOR SONIC NOISE SOURCE 
 
Have spent the afternoon searching for the source 
of the noise on the 10m ATI sonic.  With only the 
10m ATI sonic running on marigold, there were 
still synchronous spikes on the 10m ATI.   
 
Then turned the other sonics, fast T's and hygrometers 
back on.  The Gill output is now very noisy and 
there is no output from the 4m ATI v component 
(just 9999).   
 
In order to get the 4m ATI running we needed to  
'goose' it with a  while logged 
in through rserial.  In addition, when it came 
back on the first time there was no data from the 
level sensors.  We tried communicating with rserial 
to turn the ATI A/D channels back on (the level 
sensor channels), but only managed to put the ATI  
in verbose mode.  We suspect some communications 
protocol mismatch.  We hooked the Toshiba directly 
to the ATI and easily got it back into terse mode 
with the A/D channels turned on. 
 
 

GILL POWER CORD (?) CAUSED MARIGOLD CRASH 
 
Marigold quit this morning apparently when I 
wiggled the power cord for the Gill sonic. 
Restarted marigold and will check out the 
power cord.  This was likely the cause for 
marigold going down late yesterday afternoon, 
because it corresponded with pluggin the 
Gill back in after our sonic tests.  It is  
also hopefully the reason for the noisy data 
on the Gill. 
 
 

ATI REP WORKING ON SONICS TODAY 
 
Gary Zimmerman was here today, working on the 
noise problem on the ATI sonics.  The problem 
with the v channel on the 4m sonic was that 
the calibration parameters in the eeprom were 
corrupted.  He recalibrated and data were  
again reasonable.  He searched for the cause 
of the synchronous noise on the 10m ATI and on  
the comparison ATI, but had little luck.  We  
grounded the sonic booms to the tower and this 
apparently had no effect on the received  
signals. 
 
 

Gill-sonic grounded. 
High frequency noise (above 2 hz) in Gill sonic. 
Oscilloscope traces at output of sonic showed large noise content. 
Noise disappeared when sonic was disconnected from data-acquisition system. 
All amplifiers removed from amplifier board did not solve the problem.  
Signal-ground of Gill-sonic grounded to base of mast => noise reduced  
but not disappeared. Data disturbed approx. 13:45 - 15:15. Jorgen Hojstrup. 
 
 
 

Moved ati.10m back up to the 10m level. Now seeing  
an ocasional spike. 
 
Lots of activity around the flux tower for the last 
half hour. 
 
Replaced fast T sensor. Calibration seems to be 
about 4.9 deg C, although because it was so windy  
the signal would not stabilize, even with the calibration 
chamber on. 
 

ati.4m is spiking very badly. u,v,w and tc 
show simultaneous spikes. Winds from propvane show: 
 
        channel   height(m)       speed(m/s)      dir(deg) 
         200        1.0              5.96           169.0 
         201        2.0              5.90           163.4 
         202        4.0              6.94           155.1 
         203        7.0              8.10           169.2 
         204       10.0              0.00           168.3 
 
 
 

both ati.4m and ati.10m have been spiking badly 
for most of the day. Later in the afternoon they 
improved considerably, and now ati.10m is clean, 
ati.4m shows an occasional spike. 
 
These spikes are definitley of the mode where  
all channels spike at once. Perhaps moisture  
is the problem; we found water on the barrel 
of the connector on the sonic head for ati.10m  
when we moved it to the top of the tower. This moisture 
was down in the part where the connector slides 
into the sonic head; however the pins looked dry. 
 
Also had a heavy frost this morning. 
 
 

changed the calibration coefficients  
for the speed of sound channel on the 
Gill sonic, after Jorgen had obtained the 
proper specs from IN/USA. Restarted 
the covar for marigold. 
 
Gill specs are: 
 
290 m/s     0V 
370 m/s    +5V 
 
 
 

prep.config changed for Gill sonic, to compute 
temperature from the speed of sound channel, 
using sound_vt_1 calibration function.  
 
New did added: tc.comp.GILL.4m, and marigold covar set up 
to use this instead of c.comp.GILL.4m. 
 
From now on the covars will produce tc.comp.GILL.4m,  
instead of c.comp.GILL.4m 
 
 

Installed 4 meter ATI comparison sonic.  u,v,w,tc look ok. 
 
alev and blev are off scale - serial channels read 0000.   
Tried adjusting level by jostling 4 meter tower but serial 
channels didn't respond.   
 
Tower was knocked around from the time the sonic came online, 
approx 19:30, to about 20:00, jday 053. 
 

17:45 gmt, Feb 25. 
Tried to perform calibration of 4 meter ati comparison 
sonic.  Decided the wind was too high ~ 6 meters/second. 
 

ZEROING OF INTERCOMP 4m ATI SONIC 
 At 1720-1740 (2320-2340z) 26 Feb the three axes of the ati 
 intercomparison sonic were enclosed in paper-wrapped foam chambers 
 and the zero set. The paper was necessary because of the wind ~ 4 ms-1. 
   
 
 

CHECKED LEVEL OF ATI.4m.INTERCOMP SONIC 
At 1400-1430 cst the roll and pitch were checked with the theodolite. 
The roll did not need to be adjusted but the pitch was adjusted 
slightly, 
6 turnbuckle turns pulling the tower top backwards 
 
 

Toggled gain state on KD sonic from 10m/s to 25m/sec at 
16:00Z, jd=061 
 
The first linear calibration parameter for the u & v components 
(cosmos:100 and cosmos:101) were then multiplied by 2.5. 
 
Covar was restarted. 
 
Note! When post-processing the KD sonic data, the calibration 
	parameters must be changed at this time! 
 
Here is an excerpt from the prep.config file, showing entries 
appropriate for each gain state: 
 
# 
# KD sonic u,v in 10m/s gain state 
# 
s=cosmos:100	c=linear(0.000305185,0.0)	d=(u).comp.KD.4m	r=none	m=(-10:10) 
s=cosmos:101	c=linear(-0.000305185,0.0)	d=(v).comp.KD.4m	r=none	m=(-10:10) 
# 
# KD sonic u,v in 25m/s gain state 
# KD was switched to this state from 10m/s during STORMFEST92 on 
# jd=61 16:00Z 
# 
s=cosmos:100	c=linear(0.0007629625,0.0)	d=(u).comp.KD.4m	r=none	m=(-10:10) 
s=cosmos:101	c=linear(-0.0007629625,0.0)	d=(v).comp.KD.4m	r=none	m=(-10:10) 
 

Wrong calibration parameters in KD sonic covars. 
 
As it turns out, the covar on cosmos was NOT restarted 
at the time of the KD gain state switch yesterday, 
so the covars are wrong from 16:00 gmt jd=061, to  
01:45 gmt jd=062. 
 

ATI 4m FLUX SONIC: 
 
	Decided to adjust the gain pots on 
the 4m ATI sonic. The gains were decrease such 
that the first receive pulse was below .5 volts. This took 
place about 30 minutes ago; 21:50 GMT. 
Since then there has been no spikes on all ATI units. 
The wind speed has also decreased. 
 

4M ATI COMP GONE DOWN: 
 
	The 4m ati comparison has gone down. 
Recycling power had no affect on it. Heavy rain 
conditions prevent us from doing much more. 
 

ATI SONICS LOOKING BAD: 
 
	All 3 ati sonics are looking bad. 
They have gone beyond spiking; just bad data. 
It may be due to moisture getting to the 
transducers. 
 

SONIC WORK: 
 
	I talked to Gary Zimmerman about the 
sonic problems. He said the outages where due 
to water attenuating the signal from the transducers. 
Since it stopped raining, I went out and cleaned 
off the transducers and the krypton windows. 
Everything is working again. The problem 
with the 4m ati comparison was a loose serial 
cable at the breakout box. 
This work occurred at about 18:50Z. 
 

Adjusted ATI Sonics 
 
I checked the received signal levels on all ATI 
sonics during a low wind period today from 2130 
to 2230.  The 4m ATI looked okay.  I reduced the 
gain on the 10m v and w axes and adjusted the 
capacitor on the w axis.  I also reduced the 
gain on the v and w axes of the 4m intercomparison 
ATI.  I recalibrated the v and w axes of the 10m 
ATI and all axes of the 4m intercomparison ATI. 
 
 

KD SHUTDOWN: 
 
	The KD has been removed from the 
system at 16:45Z. 
 

KD TOWER DIMENSIONS: 
 
While taking the KD down measurements 
were made with respect to the tower. 
height from ground = 3.96 m 
boom length = 1.15m 
The theodolite was used to check the 
pitch and roll. There was no obvious 
error in either dimension. 
 

ALIGNMENT ANGLES FOR COMP SONICS: 
 
	New alignment angles for 
the GILL and ATI comparison sonics. 
 
sonic		angle (degrees) 
 
ATI, n:		2.05 
ATI, e:		1.62 
 
GILL, n-s	4.13 
GILL, ne-sw	244.4 
 
Also measured the heights and boom lengths of 
these sonics.  The center of the ATI u and v 
axes was 115 cm in front of the tower, and the 
center of the GILL sonic was 115 cm in front 
of its tower.  The ATI was at a height of 4m 
on the tower and the GILL was at a height of 
4.25 m on its tower. 
 
 

SONIC LEVELS CHECKED 
 
After lowering the flux tower, we checked 
the signs of the sonic levels.  With the 
booms pointing to the east, the level signs 
are negative for the boom tilted up and for 
a rotation to the north, that is the level 
output is positive for a counter-clockwise  
rotation as seen when looking at the sonic  
from the front (the direction from which the 
wind blows for a positive u) and from the 
right-hand-side (as seen from the front, i.e. 
the direction from which the wind blows for  
a positive v). 
 
 

Logbook summary for Gill sonic: 
 
Feb 1: 
Mounted at roughly 4.25m with U axis along boom (+u toward tower). 
Leveled. 
Azimuth measured at 2.46 deg 
 
Feb 2: 
A/D had gain of 2 w/ 5V offset. 
Assummed channel #4 to be the speed of sound w/ 5V = 500 m/s. 
Data dump shows 20hz data is changing every sample. 
 
Feb 13: 
Wiggling power cord appeared to cause marigold crash 
 
Feb 15: 
High frequency noise (above 2 hz) detected in Gill data.  Noise 
disappeared when sonic was disconnected from data-acquisition system. 
Signal-ground of Gill grounded to base of mast => noise reduced 
but not disappeared. 
 
Feb 20: 
Changed calibration coefficients for Gill speed-of-sound to 
0 V = 290 m/s, 5 V = 370 m/s. 
Changed prep.config to calculate temperature from speed-of-sound; 
covars will now output tc.comp.GILL.4m instead of c.comp.GILL.4m. 
 
Mar 14: 
Remeasured azimuth alignment to be 4.13 deg. 
Confirmed the height of 4.25m and the center of the Gill was 
115 cm in front of the tower. 
 
Data archiving stopped at 20:29 gmt. 
 
April 24: 
Noted that the Gill sonic was model 1012R2, S/N 0055. 
 
 

Logbook summary for Kaijo Denki sonic: 
 
4 Feb: 
Added to network ~ 1600 GMT 
 
5 Feb 
Began archiving ~ 1400 CST 
Leveled and alignment angle measured to be 0.08 deg. 
 
6 Feb 
Changed sign of v component in calibration function. 
 
1 Mar 
Changed full scale range on u & v from 10 m/s to 25 m/s 
and also changed calibration function.  Time: 1600 GMT. 
Unfortunately, covars were not restarted until 0145 on  
Julian day 62, so covars are in error during this interval. 
 
11 Mar 
Removed from system at 1645 GMT 
Height of KD was 3.96m and it was 1.15m in front of the 
tower.  Theodolite used to check pitch and roll and looked okay. 
 
 

Logbook summary for comp.ATI sonic: 
 
22 Feb: 
Installed 4m ATI intercomparison sonic.  Data good after 
2000 GMT. 
 
25 Feb: 
Tried to zero sonic, but winds too strong. 
 
26 Feb: 
Zeroed sonic at about 2330. 
 
27 Feb: 
Checked sonic level with theodolite at 1400-1430 and  
adjusted pitch by pulling tower top away from boom and 
thus raising the sonic end of the boom. 
 
5 Mar: 
No data from ati at 1700 
 
7 Mar: 
Adjusted gains and recalibrated all axes 
 
8 Mar: 
Alignment measured to be 1.88 deg. 
 
14 Mar: 
Alignment from north is 2.05 deg, from east is 1.62 deg.  The 
center of the ATI u and v axes is 115 cm in front of tower and 
the height is 4m.  Stopped data collection at 2030. 
 
 

Logbook summary for ati.4m and ati.10m (flux) sonics 
 
Feb 1: 
Loaner ati installed at 4m on flux tower. 
 
Feb 2: 
Readjusted receive pulses at 1500 GMT 
Readjusted receive pulses at 1815 GMT 
 
Measured azimuth to be 91.35 deg. 
 
Feb 5: 
Examined sonic signals with 'scope on both Feb 4 & 5. 
 
Feb 6: 
Noted data cleaner at 2330. 
 
Feb 8: 
New ati sonics installed on flux tower. 
Note that 4m v component and 10m w component off scale. 
 
Feb 9: 
Noise on 4m and 10m ati sonics appears to be associated with fast T 
at 4m, so disconnected it.  Turned off 10m ati since data quality  
very poor. 
 
Feb 10: 
Found that grounding the electronics box to the tower eliminated the 
noise.  Reconnected fast T at 4m. (1900) 
 
Feb 11: 
Apparently the 10m sonic still bad, since used 4m electronics to isolate 
the problem to the 10 array or cable.  Tightening the crystals improved 
things, but still have simultaneous spiking on all channels. 
 
Feb 12: 
Searching for interference by switching off fast T's, hygrometers and 
other sonics. 
 
Feb 13: 
Still searching for noise source.  Had problems turning the 4m ATI back 
on again.  Gary Zimmerman later found that the calibration parameters in 
the eeprom were corrupted. 
 
Feb 16: 
Moved 10m ati back to 10m level.  (1745)  I recall lowering it to the  
surface for Gary to work on it.  Note that both 4m and 10m ati's are 
spiking badly, but improved later in the afternoon (2400). 
 
Mar 3: 
Adjusted receive pulses on 4m ati sonic at 2150. 
 
Mar 5:  
All 3 ati sonics spiking badly.  After talking to Gary Zimmerman, dried 
off transducers from recent rain and signals improved. 
 
Mar 7: 
Reduced gain on v and w axes of 10m ati and recalibrated them. 
 
Mar 8: 
Alignment of both sonics measured to be 91.03 deg. 
 
 

Corrections for sonic offsets and tilt 
 
The following values were determined for the sonic offsets 
and tilt fit: 
 
Sonic		Days		u.off	v.off	w.off	 b2	 b3 
 
ati.4m 		32-39(1200)	 0	 0	 0	 0	 0 
ati.4m 		40-end		-0.18	 0.02	 0.01	 0.000	-0.014 
 
ati.10m 	40-67(2200)	-0.11	-0.02	-0.01	-0.018	-0.006 
ati.10m 	67(2200)-end	-0.11	 0.15	-0.01	-0.018	-0.006 
 
comp.ati.4m	53(2000)- 
            	   67(2200)	 0.08	 0.05	 0.01	 0.003	-0.011 
comp.ati.4m	67(2200)-end	 0.10	-0.07	 0.01	 0.003	-0.011 
 
comp.GILL.4m	all		-0.01	 0.03	-0.03	-0.007	 0.031 
 
comp.KD.4m	36-61(1600)	 0.06	-0.03	 0.06	 0.012	-0.006 
comp.KD.4m	61(1600)-end	 0.07	 0.00	 0.06	 0.012	-0.006 
 

 
The offsets and tilts determined in comment 230 were found to be slightly 
in error, the result of some out-of-sequence covariances, and some 
mis-timing of sonic and prop data.  These bugs were corrected and the 
data re-analyzed.  In the meantime, covars were run with the offsets 
and tilts listed in comment 230 for the entire project.  These new 
covars were also analyzed and provided a validation of the following 
new biases and tilts: 
 
Sonic		Days		u.off	v.off	w.off	 b2	 b3 
 
ati.4m 		32-39(1200)	 0	 0	 0	 0	 0 
ati.4m 		40-end		-0.16	 0.04	 0.01	-0.001	-0.014 
 
ati.10m 	40-67(2200)	-0.09	-0.02	-0.01	-0.017	-0.006 
ati.10m 	67(2200)-end	-0.09	 0.15	 0.01	-0.022	-0.015 
 
comp.ati.4m	53(2000)- 
            	   67(2200)	 0.09	 0.03	 0.00	 0.001	-0.012 
comp.ati.4m	67(2200)-end	 0.11	-0.09	 0.02	 0.001	-0.012 
 
comp.GILL.4m	all		 0.01	 0.02	-0.02	-0.010	 0.031 
 
comp.KD.4m	36-61(1600)	 0.09	-0.03	 0.07	 0.005	-0.004 
comp.KD.4m	61(1600)-end	 0.02	 0.08	 0.07	 0.018	-0.011