SHEBA: Rad_System Messages: 52 Entries..

Some thoughts to keep hold of for now on the Radiometer Icing Issue:

Tony is looking into stronger ventillators..

		------------------------------

Subject: Re: Rime icing
To: delany@ucar.edu (Tony Delany)
Date: Tue, 14 Oct 1997 20:20:42 -0800 (AKDT)
Cc: horst@stout.atd.ucar.edu, militzer@spows.spo.ca (John Militzer)

Tony,
> 
> John,
>      we could provide additional ventilation by either:
> ...
> If we heated the air at the ventilator it would cool
> during its passage through the tubing.
> Direct heating would have to be done at the base of each dome. 
> 
> Can you tell me the sort of weather conditions lead to riming?
> Is the riming only due to the transitional conditions which are
> occuring now or will they persists into the winter? 
>                            Tony

	Steve and Chris should be here tomorrow so we can chat about
	all this.  
	I can't answer about the 'transitional' issue, but some folks
	here seem to think that it's more common through the year
	because of the blowing snow conditions.  I saw the most of it
	during the storms we had when the winds were 20-35 knots and
	also when there was fogging.   The fans couldn't stave off the
	ice, and once iced, they can't clean them.  In fact today
	before going out to setup sta2, we had to take the 4-comp into
	the met-hut for warming to free things up and dry.  Having a
	small heater in the schwarz might help but i'm frankly not
	sure.  Judging from the power data, there's quite a bit of
	spare: perhaps 1 Amp, but i want to keep some in reserve for
	battery recharging when/if that becomes necessary.  Menwhile,
	see my note in the status rpt about the excess heat in the
	sleds from the TEGs.  You're probably right that the air would
	cool too much going out to the sawhorse.
	More ventillation should help, but i doubt in 20-30 knot winds
	anything we can realistically put in would overcome that.
	The Scientific Solutions approach was to put the radiometer
	(1-each) on a rotating plate that is programmable to cover-up
	the sensor, heat it and calibrate during non-measurement
	periods.  Their config. is to measure 1 value during ~30-sec
	ea. hour.  (Note they have a motorized auto self-leveling
	system).  Because its programmable, they can set it up to run
	continuously and then cover/clear maybe 1-hour/day for
	example.  Problem = patented, expensive ($15K), and only
	1-component. 
	We might be able to rig something that includes a small 10-ish
	watt heater at the base of the domes as you mentioned to turn
	on for maybe 1-hour every 12/24 hours and hope that would
	remove any buildup in addition to the stronger fans.  The main
	battery can handle the excess and we could use a power
	switching circuit similar to what's in the trh.  In fact,
	we've talked about integrating the whole thing anyway, and
	with its own CPU, we could also build in that motorized self
	leveler.   (PS leveling the sawhorse thingey without screwing
	up the snow is extreme difficult esp, when the legs are frozen
	solid).
	Anyway, I have doubts that the fans we have will last very
	long anyway either in the 4-comp or trh.  Icing will effect
	them too, if not the serious cold.
JM

From yesterday's station visits, I have in my notes that my impression of
the icing amount on station 4 was greater than that observed on stations
1 and 3.  This is significant because Station 4 didn't have the ventillator
plugged in the previous night.

So, the ventillator does appear to be doing some good, but is inadequate.
That might imply that a 'turbo' solution could have some hope of working.

The Kipp-Zonen problems at station 4 have been fixed. The
incoming had its positive and negative signal lines reversed.
The outgoing appeared to be a connector problem. It started
working after we played with the connectors between the Kipp
and the 9 pin amp.
NOTE: This unit now has a ventilation fan with a 25ohm resistor
heater in the cone of the ventilator. Time will tell if this
approach works.

The data from the PYG's on the 4-comp system appear bad.

From the signature, they're following the PSP values.  This
can happen within the Campbell Mux. if a channel which
is being sampled has an open connection.....check for this
during the next visit.

Gordon and I went out to station 2 this afternoon to
work on the 4-component system. Based on what little
testing we could do, the interface from the pygs to the
10X is questionable. Decided to bring in the 4-component
unit to test in the room and install at station 4. So at
the moment station 2 does not have radiation data.
  A new configuration was also loaded. Refer to message
57.

Time of visit about 22:00 GMT.
  The 4-comp system was taken back to station 2 yesterday.
It checked out ok. There were 2 problems found: (1) the
CR10 was not multiplexing correctly and (2) the cables for
the PYGS needed to be reversed (end for end). The spare CR10
with the correct rom was mounted.
  The heating fan from station 4 was moved out to 2. A new
heater will be added to the fan at station 4.

  The 4-comp ventilation fan was removed so it could
be placed out at station 2. This took place at about
21:00 gmt.

  Built up another heating unit and added it to station 4.
I did some tests last night with the ventilator using the
Campbell and the spare Tsoil probe. There was a 5.5C temperature
difference between the intake air and the outlet air.

1/25/99 Changed Logbook entry type from Hygrothermometer to Rad. System

  The dome temps at station 2, Cleveland, have taken
a step change of about 15C. This happened at 10:00 GMT.
We may still have a problem with the Campbell or else the
heater unit is working overtime!

  Jeff and I went out to station 2 this morning to remove the
Campbell. We brought it back to the ship to try and find
any problem with it. I decided to send it back out with the
rnet and licor channels shorted together since they were floating
and may have had an effect on the multiplexer relays. If this 
works then we will do the same thing at all other sites. If it
dosn't work then I will reprogram the CR10X with a longer delay
after the relay switching.

  The Tdomes at Cleveland have gone high again. This occurred
at about 17:00 local time. There was alot of spiking before
they made a final transaction top the high state. Based on this
response, the shorts across net rad and licor may have helped.
Our next test will be to reprogram the logger for a longer
delay time between relay changes.
NOTE: PYGS still tracking PYRS

  Chris and I went out to station 2 to switch the
Campbell and the cables going to the PYGS. Based on
past test, this Campbell gave good PYG.raw data but
bad T.domes. The original Campbell was bad on both.
Hopefully the cable swap will help with the problem.
Time will tell.
NOTE: we had to cross 3 new leads to get to the site.

NO COMMENT!

  The latest plan for station 2 is to move the campbell
and the 4-comp system fromstation 4 out to station 2.
The campbell will be reprogrammed to handle the new
coefficients fro Tsoil and HFT. Persently, we can go out
yet because of visibility.

12/13 22:30 - 12/14 01:00gmt
	Replaced Florida's 4-port ventilation fan for its radiometers 
	with 4 individual "turbo" fans.

	First, brought in 4-component and thawed it out.  Both pyranometers 
	and the upwelling pyrgeometer had snow packed around the radiometers, 
	under the sun shields.

	Thoroughly thawed and dried all the radiometers, installed the fans, 
	and replaced the unit at Florida.  The fans did not go on, so I 
	removed the fans and brought them inside.  There, I moved pin 4 to 
	pin 2 in the new fan connector so it matched the 4-port fan 
	connector.

	I replaced the fans at Florida and this time they worked.  They 
	have a loud high pitched whine compared to the earlier fans.

	I also brushed frost off the sonic with the feather duster.

	Found that Baltimore CSI data has not been recorded since 1/1 site 
visit.  During site visit Eve "POWR" command showed realistic values, but it 
turns out these are not current values, but the last recorded values.
	Had spare RJ45 cable with me at Baltimore, but vinyl cables in Eve 
box were too cold and brittle by that time to risk moving them.  POWR command 
output made it appear that data was being recorded correctly.
	On next Baltimore site visit will replace RJ45 cable and check 
communications using "talk" function.

  The 4-comp system was brought to the ship to modify it
with the heaters mounted on the radiation shield.

  The 4-component system is back on station #4. There are
3 heater resistors mounted on the radiation shield of each
sensor. Now we wait and see what happens!

NOTE: forgot to remove the large lockdown washer/bolt. Level
      of sensor still looks ok.

Visited Florida
At ~1300 AST disconnected the fan and ventilator power (NB the ventilator fan
was running until disconnected). The four radiometer cables were disconnected.
In the Met Bldng inspected rad-array and removed rad shields. The entire
plenum spaces were completely packed with snow. Thus there was no ventilation
provided by the system.
Any deicing would have been solely due to the effect of the heaters under the
inner edge of the ventilation slot.
I fitted the array into the carrying case and returned with the case to the 
PAM room.

POWER BUDGET:
Single fan in fourplex ventilator: 12 vdc X 0.22 amp = 2.84 Watt
Each rad heater:3 x 33 ohm resisters in series = 100 ohm,
      4 x (12 x 12)/100 = 4 x 1.44 Watt = 5.76 Watt
Total ventilation + heating power = 2.84 + 5.76 = 8.6 Watt.

POSSIBLE FUTURE ACTIONS
I will determine how much wattage is available and will consider whether
to abandom ventilation and to put all available power into heating

Between 1130 and 1200 AKST I visited Florida and made changes.
TEG box:
Briefly opened box and disconnected internal ventilation fan
.
Radiometer array
The Rad Array had been modified in the PAM room. The previous ventilation/heater
consisted of a single fan in a fourplex ducted housing with only ducts to the
pyrgeometers and a daisy chain of three x 33 ohm resisters (1.4 W / radiometer)
under the rad shields. The system was completely non-effective. The array was
modified to have a daisy chain of 6 x 10 ohm resister (2.3 W / radiometer) atop
the rad shield fairly near the base of the domes. Larger holes were bored in
the housings to accomodate the new singleton ventilators but the new ventilators
were not yet installed. (These tests will be done later this week)
The rad array was reinstalled and cabled. (The level
lock had previously been left fastened. This time it was removed)

TRH:
I removed the coaxial intake, knocked out the frost, removed the Goretex
membrane cap, and then reinstalled the coaxial intake.

1/28/99 Changed Site number from 3 to 4 - twh

Between 1115 and 1200 AKST visited Florida to check rad array.

Checked cable connection for correctalignment and for tightness. OK

There was a light frosting on the top domes. Single lacey crystals, 
millimeters apart,and the size of sugar crystels. The lower dome were 
less frosted. There was no indication that the resister heaters were
having any effect. There was no absence of crystals near the individual
resisters.I did not remove the frost from the domes.

The sun was only two fingers above the horizon and I could not
perceive any shadowing by the plastic stand-offs and the daisy chain 

I measured the dimensions of the TEG box opening:
           5 3/4" high and 8 1/4" long

Later the TRH was replaced with the spare. The Goretex cap was on the sensor 
element and was not removed.

1100 - 1200 AKST took Florida rad array to the Met hut and working atop
wooden crate outside the door, removed the resister daisy chain rad shield.
There was a light covering of snow on the top of the radiometers. The snow
just next to the resisters showed a slight indication of the effect of the
heating. There was some snow inside the housings.
Attached the 1 3/8" fan ventilators and routed the wires out through small
holes in the metal tape sealing the nipples. Replaced the wide ventilation
slot radiation shields. Reinstalled the Array on the end of the beam and 
recabled. The fans made a screeching noise. 

1300 AKST went to check ventilation. All fans stopped. TRH fan still running.
Removed Micronel ventilatops and returned to lab. Rebuilt ventilator using
2 Rotron and 2 Nidec fans. At 1600 AKST returned to Florida and installed
ventilators. The fans didnt start. Checked voltage on fan plug. Zilch

Did iplot of load. The load ihad ncreased after the Micronel fan ventilators
were installed and then a fuse blew. The Campbell data logger also quit.
At 1645 AKST replaced fuse in Eve box. The fans started and the Campbell
also came back online... 

At ~ 1400 AKST on Sunday 01 Mar the Fla rad arry was taken down for
modification work on the ventilation. 
Also for the last day and a half, the Rsw.out had shown a negative offset
and a ragged signal. A check of connectors at 1000AKST had had no effect.

The ventilation system for the Fla rad array was modified and the array was
reinstalled at 1500 - 1530 AKST 4 Mar.
In each housing, a large (2 in" high and 2 1/2" wide) hole was milled. (This
was done by the Des Grosseilliers' shop). A  2 1/2" x 2 1/2" square Nidec fan
(50 ma / 0.6 W startup, 40 ma / 0.5 W run ) was fitted externally to the hole
using metal/cardboard from a juice container and duct tape. A 5" long, 2 1/2"
cowl made from the same material was attached to the intake of the fan.
These Nidec fans have a good history of cold weather endurance. They have been 
operating in the old ventilator housings continuously for several months.
Similar Rotron fans are used in the TRH system (the 7 bladed Nidec have a 
slightly greater power requirement than the 5 bladed Rotron and has
a slightly greater volume flow)
The modified ventilated radiometer array was reassembled and put back on Fla.
The iload increased by 0.6 amps ~ 7.2 W

Cigarette smoke showed that there was no back-flow and, on under the calm
wind condition, air flow could be sensed on the palm of the hand up to 5" 
above the dome. For comparison, the line power ventilator for the ARM
PSP and pyrgeometer yielded ventilation which could be sensed 5-6" above the
domes.

Checked the fans at 1800 AKST. Still operating.

Florida is station 4. However the radiometer array at Florida is 
designated Station 3. It has been thus, at least, since I arrived on 19 Feb 98

Atlanta is station 1. The Atlanta radiometer array is designated Station 1

At 1300 - 1320 AKST, 7 Mar I visited Fla.
All fans running dispite frost on cowls and insides of cowls. 
Three domes free of frost. The sw.in dome was clear on the sides but had
a cap of frost.
Defrosted sw.in using heat from hand and back of fleece glove.
Attempted to level radarray using horizon.
NOTE that rad array and Campbell are both Station 3 dispite Florida's 
designation as station 4

Result of leveling: level x changed from -0.5 to +1.0
                    level y changed from +2.0 to +1.0

Result of defrosting cap of dome: R.sw.in decreased from 235 Wm-2 to 160 Wm-2
 This is because when there is a frost cap, the low sun illuminates the cap
from the side and effectively increases the perceived down-welling flux of sw 
radiation.
 The implication is that ventilation that is not sucessful in completely
preventing/removing frost is possibly worse than no defrosting at all.                

Test of the effect of frost cap 
An attempt to gain more information as to the effect of the frost
cap which the better ventilation sometimes leaves residual.

The sunshine conditions were not uniform today, Sun, 8 Mar,and there were
periods of cloudiness as well as a period of hazy sunshine.
1240 AKST lw.out and lw.in domes frost free. 
          Sw.out dome with very light sprinkle of frost particles
          Sw.in dome has a cap of frost above 50 degrees, 
      Cleaned sw.in dome with EtOH below 45 degress
1315 AKST Cleaned sw.in to 60 degrees
          sun just emerging from cloud, hazy sunshine
1415 AKST Cleaned awy the rest of the cap (also cleaned very light sprinkle
           of frost particles on sw.out dome.
1500 AKST  Hazier conditions
1515 AKST  leveled rad array with bubble level

3/17 19:39 gmt
	Changed level coefficients using *4 mode of CR10 for Baltimore.  
Installed coefficients for level sensor s/n 0383.

Visit / Rad Systems at Clev. and Florida

Morning of 4/4/98, both systems had rime on both sets of domes although
the SW had considerably more (dime size on LW, 50% of dome on SW).
Cleveland was operating without any ventillators and it did have more rime
than the 'juice-can' aspirated Florida 4-comp but only slightly more.
Note the Cleveland system had the new 4-comp ventillator installed but
not turned on.  There could be an effect of natural ventillation through
the straight-in fan ducts.

By afternoon the LW domes on both Cleveland and Florida had cleared off
(sunny day, warm with the wind at ~10-20knots) and most of the rime on
all structures had gone as well.  The SW were still rimed though with
perhaps 30% more coverage on the unaspirated system.

19:00 AKST 4/4 (4:00 4/5 UTC)
	Domes cleaned.

Sa. 4/4, ~19:15 AKST, or 4/5, 4:15 UTC
Both SW domes were removed from the Kipp and Zonnens to begin the
dome vs no-dome intercomparison test with Florida.

The new 4-component individual fan technique similar to the 'juice-can'
system was powered up.

4/5/98 ~9:00-11:30 AKDT  Visit Florida / Cleveland

CONDITIONS:
	~-18, similar winds as yesterday 6-10 knots from NE, overcast
	over night, beginning to clear in later morning also slightly
	drier than yesterday at low ~80's RH.

RADIOMETER Systems:
	Florida double dome: SW covered lightly w/hardish ice crystals all
		around the dome. This was slightly worse (but debatably) than...
	Cleveland single dome: SW also covered lightly w/hardish ice crystals
		all around the dome.
	ETL system: SW had much less icing, perhaps 40% of the dome to windward.
	All LW domes were clear.

	The white radiation shields and the PVC cases on the NCAR system also
	had light ice but the ETL shields were clean so the excess heating
	appears to help.

	Reviewing differences: ETL systems are insulated with aluminized bubble
	wrap, reduced ventillation rate, more heat, and shield ring's outter
	edge has ~3/16" opening for air flow.

	Conclusions: From what I've seen during recent conditions, the
	'new' 4-component ventillation system is of little value.  Possibly the
	only benefit is if it helps clean the domes more quickly as conditions
	permit.

	Old Ideas re-discussed w/Ola: 
	Add some insulation to the NCAR plenums
	Try heating elements inside plenums again, only for the SW
	Try hydro-phobic solution (anti-fog stuff used for glasses on SW domes).

OTHER: 	Gill transducers and TRH were clear but there was a little rime on
	all of the structures.

4/8 14:00 AKDT Removed Fla 4-comp. and replace with modified 4-comp from Bal.
	Reporting reasonable considering wrong calib.
    15:00 AKDT Removed Bal 4-comp. and packed it ready to go out.

    21:15 AKDT Installed Fla 4-comp after adding a heater.
	Cannot run heating elements until bigger fuse put into EVE...
	tomorrow.

NOTE: Discovered a fair amount of snow inside the 4-comp plenums esp. from the
	upwind components.

Checked level and made slight correction per bubble level.

4/9, ~11:15 AKDT
	Installed 2A fuse in EVE.
	Turned on Minco heater installed under outter dome cap
	on the uplooking SW K&Z.
	Ventillator going full speed.
	The LWout, downlooking Pyg's ventillation fan seems to be
	complaining about the cold.

Cleveland Logger #4: 

4/10 ~14:30 Installed HFT, and Tsoil sensor at snow/ice interface.
		This is roughly 1.5' under the snow level.

4/10  18:45 AKDT, Programmed CR10 Coefficients
	HFT   = 943149	40.1 Wm**2/mV
	Tsoil = NCAR002 a0 =-236.8619
			a1 = 223.6343
			a2 =  14.91284

NOTE: Good opportunity to intercompare with Fla.
	Both at roughly same depth and ice condition and are within
	~20 feet of each other.
NOTE: Hopefully we will move Cleveland to near a lead tomorrow or
	shortly thereafter.

NOTE: Quick check of the site showing that the Tsoil is not
	working and needs to be rechecked.

Florida:  Logger #3

4/10 ~14:30 Installed HFT, and Tsoil sensor at snow/ice interface.
		This is roughly 1.5' under the snow level.

4/10  18:45 AKDT, Programmed CR10 Coefficients
	HFT   = 943148	43.8 Wm**2/mV
	Tsoil = N001    a0 =-239.1288
			a1 = 227.3445
			a2 =  11.91465


NOTE: Good opportunity to intercompare with Clev.
	Both at roughly same depth and ice condition and are within
	~20 feet of each other.
NOTE: Hopefully we will move Florida to a better (but still nearby)
	location within a week.

4/10 ~15:00 AKDT Removed Clev. 4-comp.
		Reinstalled outter domes
		Left in met-hut while refreshing shot desiccant that's
		used for between the domes.
     ~18:15 AKDT Reinstalled Clev 4-comp.

NOTE: Once again, found snow inside the 4-comp plenums esp. from the
	upwind components.

4/10, 9:30 AKDT
	Turned on Minco heater installed on case of
	the downlooking SW K&Z.

4/11 9:30 AKDT
	Tsoil sensor connector at Cleveland had come loose
	after installation.  Reinstalled.  Working.

NOTE:   Fla = -8.5
	Cle = -6.5
	These are very close together, approximately 15-20, both
	were at the ice/snow interface.
	Both had roughly 1.5' of snow over them.
	Coefficients in loggers verified.
	Gsoil for Clev. showing ~2w/m**2 higher (~1 vs 3) positive flux
	so perhaps the ice is thinner and there it actually is warmer.
	Otherwise these numbers seem suspiciously different.

Clev. outter domes reinstalled on 4/11 2:30Z

SWin very consistent between Cle. and Fla.
SWout of Clev. is ~25w/m**2 low.  This could be due to either the
	dome scratch which occured during the station crunch, or
	some antifog we tried to determine whether that would inhibit
	icing (it didn't).

4/12 2:15Z SWout dome on Cle. cleaned with alcohol to try to
	improve the reading vs Fla.
	All domes Fla,Cle,Etl were clean.

The Fla Station load without the wind generator was draining the battery
to 12.7 by this evening.  On the off chance that there are riming conditions
tonight I disconnected the extra load so that the station will keep running
and we can observe the SWin heater performance.

4/14 18Z
	Observed Domes: All clear but the SWin at Clev.
	It had light crusty ice crystals over ~40% on top.

	Fla. SWin was clean.  It has 30-ohm minco heater
	under dome and 'juice-can' vent vs 'jet-intake'
	I prefer the juice-can configuration but this may
	be a case where the heater in light rime/icing
	conditions kept the dome clear.

	Plots indicate Clev. reading higher when sun up
	by 25-30W/m**2 (clear morning up to 400w/m**2)

John, Jeff:

  Kurt ran some tests on the level sensor from Baltimore. The error we
see agrees with the data from Baltimore when the sensor was installed
there. We talked to Geomechanics about re-calibrating the sensor and
found out the source of our error. It is in the calibration data set
they provide us. After they calibrate the sensor, they remove the offset
from the fit. As an example, their data shows the y axis of the
Baltimore sensor was off by -2.46 degrees when the output signal was 0.0
volts. Kurt's test data gives an offset of -2.39 degrees. These numbers
are in agreemnet with the logbook entries. I have asked Geomechanics to
send us the offset values for all the level sensors so we can apply
corrections. One of you should add this comment to logbook so it is
documented.
Thanks!

Steve

PS - We did let Geomechanics know that their data sheets do not mention
anything about this offset!

4/18 Freezing supercooled fog overnight

Atl Domes iced up, cleaned ~18Z or before

Florida: SWin with 30-ohm minco heater was completely clean in AM.
	All others iced over.
	For this case the aspirated heater under the outer dome worked 
	very well.
ETL domes: Had some dime-quarter sized ice on top

Visit_Log

Radiation System
	4/19 22:45Z Cleaned 1/2 dollar size ice off SWin
	4/19 23:30Z - 4/20 4:30	Worked on SW sensors,
		Unable to install Minco heaters on site
		in cold.  Returned to Ship with sensors,
		installed heaters, and got them back
		online by 20:30 AKDT.

4/22 ~18:00 - 23  Seattle

	SWout outter dome, which was slightly scratched when the Cleveland
	site was crunched, was replaced with new dome.

Extended the feet on the boom to relevel  on May  31, 1998

BEFORE
x = -0.28	y = 4.07

AFTER
x = -0.10	y = 0.22

Everything else looked good.

Mr. Costa

Showed up and checked the logr data, all bad. Unplugged the rad fans & heaters
from the connector on the boom ( near the radiometers ). No change, then 
plugged them back in and started to get good data.  Is there a delay between 
Rocky and the instrument?  I used the 'talk` command to see if it was working,
but didn't use 'talk' when we played with the connector.  Brain fuzz. 
I will try to confirm using the test using the `talk' command.  In any case we 
left it unplugged and have been getting clean data since then.  

John, I didn`t get your email until tonight and will read it over for the next
visit.

NEW heights from Fla after we moved it a little higher on the knoll and put
the white ply under the legs.

289 cm	middle of the sonic boom to surface
190 cm	middle of the Rad boom to surface
187 cm	end of the TRH probe shield to surface

compass 11 degrees  at 1100

DMC

Plugged the upper heater on to eliminate frosting. We need to watch for noise
in the logger data.  

batt	14.34
teg	16.03
iload	1.478
icharge	1.865

The currents looks OK and still MUCH better than before.

The 6 sec pulse is not gone.  I noticed the same pulse at Maui today.

We also anchored the tripod down.

Daver

Had a little visit to Fla today to figure out what happened overnight.  The 
station went down around 0100.  I found the F1 fuse had blown,  a 5A slow blow.
About 1400 June 15 iload jumped to 1.5A then at 2300 to 4.5A. After an hour and
a half at 4.5A the fuse probably blew. I replaced the fuse and plugged the lower
heater ON and unplugged the upper heater.  Both heaters measured 30.9 ohms.
I pulled the radiometer shield off to check inside things and it looked ok. I
didn't pull the dome off or see any desiccant capsules.

I noticed tonight after about four hours the current jumped from ~1.2A to ~1.45A
We'll see if the lower heater blows the fuse and/or is noisey.

The six sec blip `down` is back.  I will see what happens when I plug the 
auto-ignition back on.

Leveled the radiometers
x = 0.09	y = 0.04

I. M. Searching

The electronic levels were calibrated with the bubble level on 
June 27 & 29, 1998.

at a level bubble

station		  x       y
1 Atl		-0.90	 0.22
2 Mau		-0.15	 0.86
3 Bal		 0.23	-0.10
4 Fla		 0.27	 0.04

Fla needs to be rechecked when the "frozen" leveling screw is freed.  We suggest
that these be the new zero values when checking the electronic x & y data in 
iplot() or from eve_console.

sheba

I tried the "new" Tsoil probe at Fla and it also read -239.10 C. I checked the
the probe connector ( nice RTV job ) and didn't see anything obviously wrong. 
I also looked around for a pin output or some help in the manuals up here, 
nothing more than the calibration report.  Any clues?

I M Soiled