FLATLAND96: RAIN Messages: 8 Entries..

Return to Logbook Contents Page
Entry Date Title Site Author #Graphics
54 Mon 17-Jun-1996ETI UnderReportingallMilitzer, John
61 Wed 19-Jun-1996Station Visit: ETI Check, JD1701Militzer, John
92 Sun 23-Jun-1996ETI Always drops Cal. Coef. on Power Cycles2Militzer, John
110 Tue 25-Jun-1996ETI Tests todaynoneMilitzer, John
139 Sat 29-Jun-1996ETI Precip Gauge TestsallMilitzer, John
375 Sun 28-Jul-1996light rain last nightnoneOncley, Steve
381 Mon 29-Jul-1996rained last nightallOncley, Steve
483 Thu 05-Sep-1996Rain gauge test/comparison dataallMichaelis, Matt


54: RAIN, Site all, Mon 17-Jun-1996 19:51:38 GMT, ETI UnderReporting
Heavy Rain last night in the area at about 20-23Z JD168 and again later in
the night through ~1Z JD169, however, stations 1&2 grossly under-reported
the values, showing about .5 and 1.5mm respectively.   Station 3 reported
22mm during the same time.

Today, I checked the gauge at Station 3.
First Test, "44-tip" test set:	15 tips
Second Test,"44-tip" test set:  20 tips
Reset power to logger, then
Third Test, "44-tip" test set:  15 tips

Called ETI.
Jerry Price thinks we're under-reporting in the logger.
Apparently Jeff Cole has seen this in the CR10 and fixed it also.
 
61: RAIN, Site 1, Wed 19-Jun-1996 14:30:30 GMT, Station Visit: ETI Check, JD170
Station Visit: 22:15-23:05 GMT, 96/06/18, JD170

Check Corn height:	23" (top of knee)

Uploaded METPPF Data:

Barometer Fixed:	Power cycle required

ETI Test:
The ETI's are under-reporting rain.
I performed some tests to determine whether this is caused within the ETI, or
else the Campbell Logger:
- Conditions: Warm, West wind ~4-7m/s
- Electrical: OK
	ETI is wired correctly and Campbell P1 port is being held at 4.8V.
- Initial Test:
	Poured '44-tip' water in quickly (~10 seconds)
	heard multiple sets of 5-pulse beeps each second and saw the
	'tipping' on the CR10 using a digital meter observing about 15 'cycles'
	NOTE: that with this meter i could not tell whether or not the pulses
	were actually pulling the P1 port to zero.
	Via EVE, logger reported 10 tips.
	Waited approximately 5 minutes...no more pulses.
- Power Cycled Box while doing other tests/baro/etc.
- Second Test:
	I decided to rerun the test and try to count the 'tips' exactly because
	i may have missed a few on the first go around because the beeps and
	meter 'pulses' didn't coincide exactly and i lost count a bit.
	By my watch....
	17:53:10-20 poured in '44-tipper'
	17:54:28   14 beeps/pulses
	17:55:28    3 beeps..
	17:56:	    nothing
	17:57:28    1 beep..
	17:58       nothing
	17:59       nothing
	EVE Check:  11 tips, not 18
- Third Test:
	simply because....
	18:01:15    poured small remaining amount of water in
	18:02:28    6 beeps/etc.
	18:03:28    4 beeps/5 pulses....odd, but...
	18:04:	    nothing more
	EVE Check:  6 tips, not 10
- Conclusion:
	- I need to re-do the tests using a scope to count exactly, and the ETI
	test software to examine what the processor says (didn't have the right
	cable today).
	- I believe the ETI itself is under-reporting and skipping the second
	and third minute pulses.
	- The Campbell may be missing tips, but cannot tell without seeing
	exact wave-form to determine whether the line gets pulled exactly to
	zero or not.
92: RAIN, Site 2, Sun 23-Jun-1996 19:23:03 GMT, ETI Always drops Cal. Coef. on Power Cycles
ETI Check performed during station visit (see previous note).

Jerry Price said the ETI shouldn't lose it's coefficients unless the
battery voltage goes down slowly.  Also, he wanted to see a higher voltage than
11.2, which i saw at station 1.

Voltage at Campbell and ETI: 13.3
Coefficients were definitely gone.  Programmed them in.  Cycled power by pulling
the 9-pin cable.  Coef. dumped.  Reconnected the ETI to the input power rather
than on the campbell take-off lugs.  Programmed Coef. Cycled power. Same
results.   Repeated by programming coef., cycling power using the main EVE box 
switch, same results.

Conclusion, the ETI will dump it's cal coefficients on a power cycle.
ETI probably needs to put a delay into the decision about reprogramming the
coefficients during 'low-voltage' situations.
110: RAIN, Site none, Tue 25-Jun-1996 23:50:07 GMT, ETI Tests today
ETI Checks today:

Jerry Price could reproduce the dropped coef. only by 'sparking' the
input power, these tests were to check out the possibility of 'dirty'
power causing our problem.

Station 3, s/n51
----------------
When i arrived and checked the eeprom cal coefficients, they were OK.
This station has definitely been power cycled since they were loaded.


Station 1, s/n57
----------------
This is the A/C powered station.  The power system was stabilized yesterday
with the addition of a battery box, in between the DC supply and PAM.
The voltage is around 12.5-13 and is rock (granite, not pumice that is) solid. 

Coef. were dumped when i arrived.
Programmed them in, and cycled power hoping that they'd stay programmed..
....foolish me. 
Test for unusual power up glitching.
The scopemeter was setup to trigger/record the positive slope.  Coef. were 
loaded and verified and power was cycled.  The waveform rose from 0 to 
10V within 250microSec, 12 within 600microSec, and stabilized at 12.5 within
1.2mS: no glitches, signal was very clean.  The Coefficients however, suffered
greatly.
Test for unusual power down glitching.
After confirming the power up looks good, i decided to look at the power down.
It dropped from 12.5 to 4 within 200microSec.  It did take
3mS to drop from 4 to 2Volts, and the scope memory trace didn't have enough
to show the wave all the way to 0, but it was a very clean signal.  However,
the reprogrammed coefficients suffered greatly.

I tried these tests both with and without the PC attached.
It may well be that ETI s/n #57 and #58 are the only ones having this problem,
because #51 did retain it's coefficients after i programmed them in several
days ago.

Jerry Price had recommended to put a small 12V battery in the gauges just to
keep them up until returning to Boulder.  We may do that because frankly this
isn't important enough for this project to continue right now.

139: RAIN, Site all, Sat 29-Jun-1996 23:36:22 GMT, ETI Precip Gauge Tests
With the New Campbell Proms installed, and the power cables for the ETI's
s/n 57,58, and all of the cal. coef loaded the following tip tests were
observed.

44-tip reservoir used:
Sta 1, s/n 58	~14:00Z	 45-tips
Sta 2, s/n 57	~17:40Z	 46-tips
Sta 3, s/n 51	~21:10Z	 45-tips
375: RAIN, Site none, Sun 28-Jul-1996 14:29:54 GMT, light rain last night
We had light rain about 04Z at the Radisson last night, but none of the
stations saw it.
381: RAIN, Site all, Mon 29-Jul-1996 21:09:14 GMT, rained last night
It did indeed rain last night - first just as we were leaving at 23Z (drizzle),
and much later (07Z?) while we were in the motel.  Both had thunder and 
lightning.
483: RAIN, Site all, Thu 05-Sep-1996 20:02:27 GMT, Rain gauge test/comparison data
                                             FLAT96
                                       Rain Gauge Tests
                                       GROUPED BY STATION
                                       [report rain_chk.r]

                         RG       IC      RG      %
Sta#    Date    Time   Type-id#   Qty   Reading Error   Comments
----  --------  -----  --------   ----  ------- -----   -----------------------

  1    6/18/96  14:19   ETI 58     44    11.0   -75.0   Yuk, under-reporting
  1    6/18/96  14:20   ETI 58     44    11.0   -75.0   Yuk, under-reporting
  1    8/04/96  19:00   ETI 58     44    45.0     2.3   Corn leaves may be
                                                        interfering withrain
                                                        catch

  2    7/30/96  19:53   ETI 57     44    46.0     4.5   may have rinsed a few
                                                        bugs in

  3    7/30/96  19:03   ETI 51     44    45.0     2.3   Matt says: darn close
                                                        or something like that
  3    8/04/96  17:39   ETI 51     44    45.0     2.3   no evaporation problem
                                                        here