FLOSS01: CO2 Messages: 16 Entries..

Cal gas amounts:
#5: 1050
other: 710

Between 14:00 and 14:15 we removed the Licor 7500, brought it back to the 
trailer and packed it into its box

At 10:15 Thu, 21 Feb, checked cal gas pressures
Lo-cal  cyl #6 1700/20
Hi-cal  cyl #5 1600/20
 
compared to 1750/20
            1690/20
 when installed yesterday, Wed, 20 Feb afternoon

On Fri at 09:30
Lo-cal Cyl # 6 1670/20
Hi-cal cyl # 5 1600/20

On Mon,25 Feb, at 14:20
Lo-cal cyl # 6 1480/20
Hi-cal cyl # 5 1480/20

On Tue, 26 Feb at 14:00
Lo-cal cyl # 6 1400/20
Hi-cal cyl # 5 1400/20

On Wed, 27 Feb at 14:30
Lo-cal cyl # 6 1400/20
Hi-cal cyl # 5 1400/20                 check whether solenoids are operating

At 14:00 checked intake flow of CO2 flux system = 4 lpm. 
This low value prbably because of the progressive loading of the intake filter.

Changed cylinders to
 Hi-cal cylinder # 5 Start at 1690/20
 Lo-cal cylinder # 6 Start at 1750/20

Adjusted bypass valve to make intake flow = 7.5 lpm

Cmmunication with PAM base, ttl_on/off cosmos 0/1 then ttl_return cosmos 0/1
Adjusted lo-cal flow valve to yield an excess lo-cal flow of 1 lpm
Adjusted hi-cal flow valve to yield an excess hi-cal flow of 1 lpm

I just swapped the nearest cylinder on the closed-path CO2 system, since it was
out.  We're now using cylinder #5, which had about 1900 lbs of pressure 
initially.  The other cylinder still has about 500 lbs in it, so I didn't swap
it.

P.S. The cylinder status now is:
cyl #2 - now empty
cyl #6 - not yet used -- ready to swap with #4 when it dies.
cyl #4 - still had 410 lbs as of 4pm yesterday, connected to system
cyl #5 - lost 200 lbs yesterday, so only 1700 lbs as of 4 pm yesterday.
	it is now shut off, but connected to the system.

Here is what I think the Li7500 is reporting (I guess they are using g-moles): 

H2OD - mmol/m3

convert to mass mixing ratio by:

H2OD (mmol/m3) * 0.001 (mol/mmol) * 18 (g_H2O/mol_H2O) / rho.air (kg_air/m3)

CO2D - umol/m3???

convert to ppmV by:

CO2D (mmol/m3) * 1000 (umol/mmol) * 28.97 (g_air/mol_air) * 0.001 (kg_air/g_air)
 / rho.air (kg_air/m3)

A plot of values converted this way is attached.  H2OD is compared to Q and 
CO2D is compared to co2mr.  Everything seems in the right ballpark.

On Dec 10, 09:36 MST cosmos was rebooted to install a
new sio_converter for the licor 7500 on channel 212.

Up until 09:36 we were using noconv_nl as the sio_converter
for the licor, which does no conversion and passes the
ASCII to the cal routine (licor_7500) to parse.

For efficency sake, were now using this sio_converter:

licor_7500 9 "(Data (Ndx %f)(CO2Raw %f)(CO2D %f)(H2ORaw %f)(H2OD %f)(Temp %f)(Pres
%f)(Aux %f)(Cooler %f))"

The cal routine is then the generic serial_float, though I imagine
we may have to create a special one if we want to convert the
units.

The licor periodically (once a second?) puts out a diagnostic
record, which will be ignored (and not archived).  To see
the diagnostic record in real-time use rserial.

I mounted OSU's LI-7500 yesterday about 2pm and connected it to cosmos:212.
We are archiving the data in unconverted ASCII until we can figure out a
different format.  

I don't like the mounting for the LiCor's box.  Currently, I have it just
sitting on a breakout box stand.  At the very least, it should be bolted to
this stand....

I also did some CONFUSING cabling for this sensor.  I have connected it to
MARIGOLD's Serial/Analog breakout box, to avoid adding another BOB.  My
primary reason was to avoid running more DC power cable, since the LiCor takes
2 AMPS.  A secondary reason was that we might connect it to marigold later
anyway, if Larry decides to move this sensor to 10m.

We still need to parse the message from this sensor.  I've just put a generic
prep.config entry in for this sensor, commented out.

I just did the flow rate checks that Tony suggested.  Each cal gas was turned
on for about 30 seconds at about 16:25.  Note that I was standing near the
sonic during this period, so all fluxes at 2m are suspect.

nominal flow rate is ~5.5 lpm
flow rate with cylinder 1 is ~-2 lpm
flow rate with cylinder 3 is ~-2.5 lpm

These readings were so close to what Tony intended that I didn't change
anything.  

But we still don't know why his gas is being used up relatively fast.  I can
think of several possibilities:
- The flow rate out of the pressurized cylinders is more than through the
scintered inlet.  I could check this by metering the analyzer outflow.
- There is a leak in the cal gas line.  This strikes me as unlikely since both
cylinders are going down at about the same rate.
 

The Licor aspirated CO2 system was installed on 30 Nov''01
The configuration is shown in ...../doc/co2licor.idr (to be copied from
~/deploymnets/floss02/co2licor.idr)

lico2.2m cosmos 102
li.2m    cosmos 103

         Channel 0   Channel 1
Ambient      off         off
Lo-cal       on          Off
Hi-cal       off         on

Sequence of 2 minutes Lo-cal, 2 minutes of Hi-cal, 4 Hr 56 minutes of Ambient,
repeating every 5 hrs
was initiated.

The system was tested on 1 dec''01.
Lo-cal and Hi-cal were activated and a Zero can period was run.
The Flows were checked with a floating ball meter.
5 lpm intake during ambient flow.
    Adjusted using bypass valve on pump suction manifold.
    
~2 lpm backflow during both Lo-cal and Hi-cal periods.
    Adjusted using cross handle, very sensitive, for lo-cal
         and using knob, for hi-cal

Set a default calibration factor of 180 ppmv/volt to generate an approximately
correct mixing ratio
Keeping track of cal gas usage

On 8 Nov''01 six cylinders of synthetic air/CO2 were cross calibrated with a CMDL
standard at CUFF.

Cal1 AK-170118 317 ppmv

Cal2 220K-0017 315 ppm

Cal3 LK58629   484 ppmv

Cal4 LK285067  495 ppmv

Cal5 291120Y   493 ppmv

Cal6 220K-0002 324 ppmv

Cal5 and Cal6 remain at FLAB
Cal1, Cal2 , the Lo-cal cylinders, and Cal3, Cal4. the Hi-cal cylinders are at the
tower base.
Cal1=Lo-cal and Cal3=Hi-cal were connected Sat 01 Dec''01
Cal2=Lo-cal and Cal4=Hi-cal are in reserve, laying next to the active cylinders.

I just looked at the CO2 calibration on 5 Dec 02:00.  From an eyeball average
of values from prep, using a default gain of 190 and bias of 0:
cyl #1 - 317 => 334
cyl #3 - 484 => 472

Thus, the gain would be: 190 * (484-317)/(472-334) = 230
and the offset would be: 317 - 334*230/190 = -87

I'll go ahead and put these in, though we expect to generate a time-varying
calibration eventually.

calibrations from Niwot Ridge site:

cal 1 317
cal 2 315
cal 3 484
cal 4 495
cal 5 493
cal 6 324 ppmV

We are currently using cylinders 1 and 3.

Bill Massman and John ? are now sniffing these as well.

We'll start to keep track of cal gas amounts here:

	Cal 1(Lo)	Cal 3(Hi)
01 Dec	20/1750		20/1630
03 Dec	 ?/1710		 ?/1610
04 Dec  22/1610         19/1595 (~10:30; before Massman)
04 Dec  22/1590         20/1480 (~16:30; well after Massman)
06 Dec  22/1400         20/1410
07 Dec  22/1360         20/1400
08 Dec  22/1250         20/1330
09 Dec  22/1200         20/1300
10 Dec  22/1120         20/1290

The co2 calibration is controlled by the ttl program in channel config:

# Tony wants a 2 minute calibration every 5 hours
#   17880 secs = 4 hours, 58 minutes
#
# adam          datel_ttl_output unit chan start        on       off
#
cosmos          datel_ttl_output 0 0    00:00:00        120     17880
cosmos          datel_ttl_output 0 1    00:02:00        120     17880

TTL channel 0 controls the low cal gas valve, channel 1 controls
the high cal gas valve.

The low cal gas is 317 ppmV CO2, the high is 484 ppmV.

After cosmos is booted, it performs its first calibration at the
next multiple of 5 hours from 00:00 GMT of the current day, which
is at 05:00, 10:00, 15:00, 20:00 GMT or 01:00 the next day.

The sequence then repeats every 5 hours.

So assuming cosmos was booted at 17:00 GMT, the following sequence
would occur:
	20:00	low cal gas
	20:02	high cal gas
	20:04	ambient air until 01:00 next day.

The next day the cal times will be an hour later:
	01:00	same cal sequence
	06:00
	etc

Cosmos was last booted yesterday, Nov 30.
Tony connected the licor this morning, Dec 1.

I verified in the high rate co2mr.2m data that the calibration 
sequence was done at 21:00 GMT today.