Home

remote data transmision system - Worcester Polytechnic Institute

Contents

1. Connected 0 00 07 Auto detect 9600 8 N 1 ROI APS apture Figure 50 Hyper Terminal Hello World Conclusion By sending out a hello world string over an RS232 com port one of the most important functions in the end design has been demonstrated Although this application is trivial the ability of the CR1000 datalogger to interface with an Iridium transceiver required serial communication with strings 6 1 3 Serial In Purpose The goal of this program was to experiment with reading data through the datalogger s serial port and then performing additional processing based on what was read Setup Power on the CR1000 Connect the serial port from a PC to the datalogger s RS232 port Open a hyper terminal session on the PC Run the program entitled Confirmation after the start prompt is displayed type confirmation in the hyper terminal window immediately after this Confirmation_Received should be seen in the hyper terminal window Solution 61 Figure 51 Timing control ports program This confirmation program defines 2 strings localconf and confirm The program calls a subroutine to read in data through the RS232 port and store that data in the string localconf The program then compares that string to the string confirm If the two strings are equal then the program will output Confirmation Recieved otherwise it will output Try Again The command Serialopen ComPort Baud
2. Return Data Post data to Transport Network Data setupLogging Set up a debugging system E init Initializes settings i getReadings Obtains Data Return receive Calls functions to format and retrieve data Figure 38 Overview of Figure 37 Overview of Receive s Data Datalogger Figure 38 shows a basic diagram of how the data is received by the python program First the logging is set up which will help to debu errors Next the program initializes the settings u g code and display any warnings or sed throughout the code To obtain the 46 data getReadings is called which will retrieve the data and return it to Receive to be posted to the Data Transport Network Log and Initialize The first steps of this program are to set up the logging and to do the initialization The logging has provided the ability to set a severity of the logging statements that were in the code they ranged from debug to error The logging could also be set so that only messages of a certain level of severity or higher could be seen While debugging the program it was important to see all of the messages that appeared However once the program was working correctly only those messages that signaled an error needed to be viewed Next all of the settings needed to be initialized To initialize the program the modem was set to au
3. 116 Datalogger C OBBECUODS isso aan a en aE 117 CSI Voltage Divider c 119 Mounting Panel Comme ction a eee epos e erp S rn EEEE EEES 119 PV panel 3nd ACD Nace ost iseen ie di aient I opua iaciunt 121 rang M 121 CRIDUU Code User Mantal so iinacitg octo dt qee hpdenditin naea denen adde a atn vk tha den igne Indi 122 I Mi TN aac igs pace acs cee peace M 127 iv Table of Figures Figure 1 NOAA PMEL Webcam SIO cas ae fees sadecaspaseadtietsie aie cceenssdeadceestaueaadencadenaencced 3 Figure 2 GEOSummit webpage WWW geosummit org scceeiaaiessccvescorecntenctdccaduneeiaaaecareracin 4 Fig re 5 Location of KOZeDUG eee ve caasenetadeaetanueas beni inesi oiae Ei ensia a aiina 5 Figure 4 Average and Extreme Temperatures for Kotzebue AK esses 5 Figure 5 Average Monthly Precipitation in Kotzebue AK eee 6 Figure 6 Cloudi ess in Kotzebue AK oiii riter in bn I an ie ipie UA UNA 6 Figure 7 Insolation in Kotzebue AK o eode dati disais td odit ded uet ote Da uds 7 Figure 82 Wand Speeds 1n Kotzebue AN betreten rr RO PIE IS SU rua ER enoii ieie 7 Figure 9 Standard components of a CSI Datalogger aite ettet er stein e daas 8 Fig re 10 CR TODO Datalog cea sen es acscnadis ui cttntiectr ta terss aeg adco er Mum eoii UuE 9 Figure 11 A CR1000 used in a weather station
4. Add a plot axes figure add subplot 111 row 1 coll subplot 1 Raise subplot up a little figure subplots_adjust bottome 0 2 Title title axes set title line getY getMeasurement YLabel ylabel axes set ylabel line getY getUnits Grid axes grid True 114 l b w h awex get position axes set position 1 4 0 b w 1 w 2 0 h 1 h 4 0 Plot the lines 11 axes plot date line getX getData line getY getData fmt Set the colors 4 ll set color Color brown 4 X limits 1 Mlab min series x u Mlab max series x 1 matplotlib pylab date2num date_range fm u matpltlib pylab date2num date_range to axes set xlim 1 u Y limits Do this following all plots so ylim is autoscaled to be used l u axes get ylim if line getY getRange range line getY getRange axes set ylim range date formatter matplotlib pylab dates DateFormatter m d n SH axes xaxis set major formatter date formatter Write out the png canvas print figure png path dpi 80 dpi 150 is default def setupLogging logging format asctime s name s filename s levelnames s message s formatter logging Formatter logging format log file logging FileHandler var tmp sri Datalogger Plotgenerator log log file setFormatter formatter log file setLevel logging DEBUG w console loggin
5. Information QU ERY Home Contact er Feb zr axe wi soo iurc Situs To Feb zr vao 1000 m ure O File csv 9 Browser HTML Note The tables take a LONG time to render Output m Health Data E Query Updates m Release Notes Figure 47 Website s Query page Viewing the data through a web browser can be seen in Figure 48 58 query results Information QUERY Query Results diues From 02 27 06 09 00 UTC pots E To 02 27 06 10 00 UTC m Health lo2727 06 09 00 00 22 4183 12 0528 12 5283 0 874853 21 9975 NA 02 27 06 09 01 00 22 4183 12 0527 12 5283 0 874853 21 9975 Data 02 27 06 09 02 00 22 4183 12 0532 12 5285 0 974868 21 9976 E 02 27 06 09 03 00 22 4183 12 0477 12 5285 0 874868 21 9976 Mone 02 27 06 09 04 00 22 4183 12 0476 12 5285 0 874865 22 0305 NA 02 27 06 09 05 00 22 4477 12 0475 12 5284 17 1291 21 9981 Lo 02 27 06 09 06 00 22 4477 12 0529 12 5284 17 1493 21 8348 E UTEM 02 27 06 09 07 00 22 4477 12 0529 12 5284 16 9196 21 7372 02 27 06 09 08 00 22 4183 12 0533 12 5286 16 8928 21 7697 02 27 06 09 09 00 22 4183 12 0536 12 5287 17 0011 21 7699 02 27 06 09 10 00 22 4183 12 0534 12 5289 17 0114 21 7701 02 27 06 09 11 00 22 4183 12 0533 12 529 171163 21 5757 02 27 06 09 12 00 22 3912 12 0833 12 529 16 792 214141 02 27 06 09 13 00 22 3641 12 048 12 529 16 7786 21 3498 02 27 06 09 14 00 22 3641 12 0534 12 5293
6. http www campbellsci com downloads Campbell s website for OS upgrades software downloads updates and various datalogging resources 127 References SRI International December 2005 http sri com esd cgs index html Arctic Theme Pade February 2006 http www arctic noaa gov gallery_np html Kotzebue AK City Profile December 2005 http www epodunk com cgi bin genInfo php locIndex 27971 Kotzebue Alaska December 2005 http www wrcc dri edu cgi bin cliMAIN pl akkotz Alaska com Weather and Climate December 2005 http www alaska com about weather v page2 story 448 1284p 4773632c html NOAA Cloudiness J anuary 2006 http lwf ncdc noaa gov oa climate online ccd cldy html 7 Insolation at Specified Location NASA January 2006 http aom giss nasa gov srlocat html Campbell Scientific Institute Dataloggers December 2005 www campbellsci com dataloggers Campbell Scientific Institute Dataloggers December 2005 http www campbellsci com cr1000 Campbell Scientific Institute Dataloggers December 2005 www campbellsci com dataloggers Campbell Scientific Institute Dataloggers December 2005 http www campbellsci com scwin P SaVi December 2005 http www geom uiuc edu worfolk SaVi images iridium_coverage gif P About Iridium November 2005 http www iridium com corp iri corp understand asp NALRESEARCH February
7. param clause values clause logging debug s sqlStmt Ss FUNCTION sqlStmt ro self connection query sqlStmt except pg ProgrammingError e logging error s s FUNCTION e raise return if name main reading Reading reading phone number phone number reading timestamp datetime datetime now reading values 1 2 3 4 print reading toString Datalogger usr bin env python import serial import time import datetime from time import gmtime strftime import struct import logging import sys import StringlO import os import string import sri ayoung DatetimeUtils import Reading which data is sent ETTING THIS VALUE dE db dE P at S PERIOD 60 60 seconds CRLF r n BEGIN HEADER Q4 END HEADER S END FILE CONNECT CONNECT NOCARRIER NO CARRIER ASCII READING SEP CRLF BINARY READING SEP pppp THANKYOU THANKYOU BAUDRATE 2400 OK OK def setupLogging logging_format message s logging format asctime s formatter logging Formatter log file log file setFormatter log file setLevel console logging StreamHandle console setFormatter formatte console setLevel rootLogger rootLogger rootLogger rootLogger logging getLogger setLevel logging D addHandler console addHandler
8. from stat import import sri Datalogger Reading import Database dbname datalogger user apache passwd apache def is_writable_dir p p is a string pointing to a putative writable dir return True p 111 dE db HE HE is such a string else False From usr lib python2 4 site packages matplotlib __init__ py try p test is string like except TypeError return False try t tempfile TemporaryFile dir p t write 1 t closet except OSError return False else return True Here is some weird voodoo http www scipy org wikis topical software UsingMatPlotLibInACGIScript matplotlib wants to run in a writable directory which when run through apache may be root unwritable by apache home HOME home dir os environ HOME if not is writable dir home dir home dir tmp os environ HOME import matplotlib dE dE dE de ode Here is some weird voodoo http www scipy org wikis topical software UsingMatPlotLibInACGIScript importing pylab w o calling matplotlib use Agg will have the import try to open the display which when run through apache which will fail matplotlib use Agg import pylab from matplotlib backends backend agg import FigureCanvasAgg as FigureCanvas import Mlab class Series def init self measurement None units None self measurement measurement self units units self range
9. Figure 25 Subsystems 5 1 Power System VECO Polar Resources gave the responsibility of designing and assembling the power system for the datalogging project to Tracy Dahl The preliminary design agreed on by the WPI team and Tracy Dahl is described in this section The block diagram can be seen in Figure 26 p memo odo S MORSINGSTAR D PULAVU CONTEOLITA aero 44 i 2 Deka 8G31 Batteries INDIA PV20 Solar Array MorningStar SunGuard Charge Controller Figure 26 Power System Block Diagram 29 5 1 2 Battery The battery chosen for the system is the Deka 8G31 100Ah Battery The battery is a gel cell sealed lead acid SLA battery which means it can be mounted in any position and still be operational The gel cell properties also provide recovery from deep freezes of below 40deg C that interrupt the chemical reaction inside the battery Once the temperature rises the battery will be able to hold charge again unlike ordinary SLA batteries which could be damaged by the freeze Two batteries are used in the system one battery provides power to the datalogger while the other powers the communications support The redundancy insures that any communications power malfunction does not sacrifice the information on the datalogger Each battery provides 100Ah which will be more than sufficient for both the datalogger and the communications for an entire year of operation without recharging Estimating from data sheet
10. Day of Year Western 3 Regional Extreme Max Ave Max Ave Min Extreme Min Climate Center Figure 4 Average and Extreme Temperatures for Kotzebue AK Due to the high latitude the daylight hours in Kotzebue vary greatly from summer to winter On the winter solstice December 21 the sun is only up for about 1 5 hours rising at 1 01PM and setting at 2 41PM During the summer solstice Kotzebue experiences 24 hours of sunlight In fact the sun rises on June 12 and sets July 2 providing a month of uninterrupted daylight Kotzebue receives 8 98in of precipitation annually on average significantly less than most areas in the United States Most of this precipitation comes from July to October as seen in Figure 5 KOTZEBUE 25 N ALASKA 505051 Period of Record 11 1 1982 to 5 31 1995 Precipitation Cin Mau Jul Sep Jun Rug Day of Year Figure 5 Average Monthly Precipitation in Kotzebue AK When harnessing energy from the sun it is also important to take cloudiness into account A cloudy day can greatly impact the output of a solar panel Figure 6 shows the monthly averages for clear partly cloudy and cloudy days in Kotzebue Cloudiness in Kotzebue AK m Clear m Partly Cloudy o Cloudy Jan Feb Mar Apr May X Jun Jul Aug Sep Oct Nov Dec Month Figure 6 Cloudiness in Kotzebue AK Solar energy also depends upon the strength of the sun Solar insolation is the
11. Figure 54 Hyper Terminal CR1000 Data Point The data in Figure 54 shows four characters which represent a 4 byte float Hyper terminal interprets everything as ASCII so a binary to ASCII conversion process was necessary Quickly converting ASCII to binary and then binary to IEEE float shows that this number is 10 1438 Volts which approximately matches the voltage shown by the power supply Conclusion Through the send data program it was successful in demonstrating how to use the serial port of the datalogger to send collected data This function simulated the end goal which was to ultimately collect and periodically send data over the Iridium satellite network 6 1 6 Control Port Time Purpose This program entitled Realtimesub utilizes the Realtime instruction to access the datalogger s real time clock as opposed to a counter used in previous programs Additionally the datalogger s control ports are activated to simulate power cycling the Iridium transceiver Setup Power on CR1000 load Realtimesub with PC200W Software Monitor the datalogger s control port 1 using a multi meter Solution 01 Declare array 66 Figure 55 Timing control ports program The second row of this program defines an array with 9 parameters called rTime 9 Lines 2 11 setup aliases corresponding to each of the 9 parameters in rTime The main program passes two values to the sub routine Realtimesub The first value passed
12. Table 3 Energy consumed per yOHE 2 uode tede tpon tod en tinta d oa ciun ioia ienie 22 Table 4 External Microcontroller vs CR1000 Internal Microcontroller 26 Table 5 Data Transfer Mobile Originated vs Mobile Terminated Weighted Chart 27 Table 6 Overnight Frequency amp Reliability Test dedos taut ie i abs eetixnc tt beihn 70 Table 7 Weekend data rate tes T 73 Table 8 Energy Consumption of System iiec sec ie eese aea bea bus ei ron egii une 78 vii Abstract The primary goal for this project was to create an autonomous remote data transmission system for environmental researchers collecting data in remote locations The system will be used in Kotzebue AK to monitor environmental conditions To design this system an interface between the Iridium Satellite Network and Campbell CR1000 datalogger was implemented and analyzed The solution includes a fully functional prototype which is able to provide near real time access to collected data viii Executive Summary The Center for GeoSpace Studies at SRI conducts research relating to the upper atmosphere and space environment This research often entails experiments using incoherent scatter radar satellite communications and optical instrumentation The Center is also one of the entities that make up VECO Polar Resources VPR the National Science Foundation s Arctic logistics contractor VPR supports research stations in Alaska Canada Gre
13. While the WPI team s system did not employ a webcam it did implement the Iridium satellite network to transmit data in real time Being able to review previous applications of Iridium s technology was useful in the design of this project Meteorological pressure _Ice instrument i Downward we tower Ax of looking P Snow ice floe sounder covered sea ice floe Figure 1 NOAA PMEL Webcam shot Another similar project is called the Summit Station It is located at the peak of the Greenland Ice Cap and like the WPI Sullivan project Summit is also sponsored by the NSF through VECO Polar Resources Summit Station is home to the Greenland Environmental Observatory also known as GEOSummit which provides real time monitoring of climate conditions This station is positioned on 3200m of ice which is almost 400km from the nearest point of land GEOSummit supports a diversity of scientific research including year round measurements of air snow interactions that provide crucial knowledge for interpreting data from deep ice cores drilled both at GEOSummit and elsewhere Zl http www summitcamp org Summit Camp Weather Microsoft Internet Explorer E3 File Edit View Favorites Tools Help J i D Oy 2B eh O O DAG Pwo dem 8 3 m LU Address httpe mww summitcemp orgftransport weatherstation i w Ee ws Ai weather station monitor 5 zl Current Conditions Outside Tenperature 8 207 aa Ht ore 8 Feb
14. a 5 F 56 Oo KR 8 9 S9 F A Day Figure 64 Iridium Battery State of Charge Estimated Energy Available in Datalogger Battery 100 90 80 70 o 5 60 o T 50 o amp 40 30 20 10 0 T oo 10 A Q O c Q TF st 0 i O0 CN DAM DK sb r CO T7 OW O O O N M O Fl O O QN cO i0 PF DON s 10 T rrr C Q QN OQ Q QN Q oe c co o Day Figure 65 Datalogger Battery State of Charge These graphs show that the system can operate independently throughout the entire year 82 7 Recommendations for Future Work The following section describes some potential extensions to the data transmission system While some may have greater impact than others researchers have shown interest in all of these possibilities 7 1 Universal Datalogger Communications One useful extension to this project would accommodate data transmission from any Campbell Scientific Inc CSI datalogger While creating an interface for any CSI datalogger model would certainly require external processing it would also make this project useful for deployments other than the Sullivan project Two of the more common CSI dataloggers are the CR10X and the CR23X Since these models are older than the CR1000 there are more of these deployed throughout the world Backward compatibility is a desirable feature for a communications system Creating this feature would make the project far more marketable To accomplish this the programmer could create several routines within the mi
15. d is d long FUNCTION line number len lines len line if line number 1 phone number lin logging debug s phone number s FUNCTION phone number elif line number 2 column names string split line logging debug s column names s FUNCTION repr column names else logging warning s line d d is unexpected s 109 line logging debug s d lines in data __FUNCTION__ len lines line_number 0 for line in lines line number 1 logging debug s line d d is d long FUNCTION line number len lines len line reading Reading fromString line if not reading logging error s can t parse Ss FUNCTION line else reading setTransmitTimestamp transmit timestamp reading setPhoneNumber phone number reading setColumnNames column names if message format version 0 1 readings insert 0 reading elif message format version 0 2 readings append reading else logging error s message format version Ss FUNCTION message format version logging debug s exiting d readings FUNCTION len readings return readings def store self readings This function stores the values into a reading table nm FUNCTION sys getframe f code co name Open a connection to the DB try logging debug s entering connecting to s s F
16. e logging error s exception opening serial port FUNCTION logging error s s _ FUNCTION e return None if self ser None or not self ser isOpen logging error s serial None or not open FUNCTION return None self setupModem self ser logging debug s waiting for a set of readings _ FUNCTION readings self receive self ser logging debug s closing ser FUNCTION self ser close logging debug s exiting FUNCTION return readings def simulate self readings 98 def for i in range 10 reading Reading Reading reading phone_number 800 555 1212 reading timestamp datetime datetime now reading values 1 2 3 4 readings append reading return readings setupModem self ser __FUNCTION__ sys getframe f code co name logging debug s ser s _ FUNCTION ser cmd ats0 1 s CRLF time sleep 0 5 logging debug s sending s _ FUNCTION repr cmd ser write cmd self readUntil ser cmd self readUntil ser SSOK s CRLF CRLF time sleep 0 5 receive self ser FUNCTION sys getframe f code co name logging debug s enter FUNCTION Gl if self readUntil ser CONNECT None return None if self readUntil ser s s BAUDRATE CRLF None return None phone_number col_names self receiveHeader ser
17. if phone_number None return None readings self retreiveData ser phone number col names if readings None return None Update the period logging debug s period d _ FUNCTION self period cmd PSTRSdPEND self period logging debug s writing Ss FUNCTION cmd ser write cmd time sleep 5 delay before hang up HEERE mnm Increment period for testing nm self period 5 60 seconds THE f Close connection logging debug s entering command mode FUNCTION ser write 99 if self readUntil ser OK None return None logging debug s issuing hangup ath FUNCTION ser write ath r n self readUntil ser OK ignore None return logging debug s exit __FUNCTION__ return readings def receiveHeader self ser This function reads in the header nm FUNCTION sys getframe f code co name logging debug s entering FUNCTION if self readUntil ser BEGIN HEADER None return None None header str self readUntil ser END HEADER if header str None return None None phone number col names str string split header str 1 col names string split col names str logging debug s exiting FUNCTION return phone number col names def retreiveData self ser phone number col names Read in the data from the serial port u
18. logging error e raise if ro None logging warning s sql Ss returned no result object name sqiStmt return None else results ro getresult log msg s sql s gt d rows FUNCTION sqlStmt len results readings for result in results reading Reading reading phone number result 0 reading transmit timestamp Sri ayoung DatetimeUtils parselso8601Datetime result 1 reading timestamp Sri ayoung DatetimeUtils parselso8601Datetime result 2 def reading values None TODO readings append reading if len readings O0 logging debug log msg return readings else logging info log msg return None insert self reading None FUNCTION sys getframe f code co name 95 logging debug s reading s __FUNCTION__ reading if reading None logging warning s reading is None __FUNCTION__ return a list for value in reading values NOTE It seems no whitespace is permitted in postgres Version 8 0 does allow whitespace value str g value a list append value str values s string join a list param clause phone number transmit timestamp timestamp values values clause s Ss s s amp N reading phone number reading transmit timestamp reading timestamp values try sqlStmt INSERT INTO readings s VALUES s
19. 00 12344 5 03 29 5 03 39 5 04 37 0 01 08 0 00 58 12210 6 03 45 6 03 54 6 04 54 0 01 09 0 01 00 12393 7 06 10 7 06 20 7 07 18 0 01 08 0 00 58 12402 Average Connection Time 0 01 12 0 01 00 12158 32 Table 7 Weekend data rate test This table shows an average data transmission rate of 201 01 Bps the Iridium specification sheet shows 300 Bps Now when calculating power and cost estimates for the amount of data that is going to be sent a more accurate number can be calculated Also to send 12 kilobytes of data across the Iridium Network there was an average total time of 1 minute and 12 seconds This was also useful in estimating how long a call would take and therefore how much the call would end up costing Iridium rounds the minutes used up so a 1 minute call would be charged for 2 minutes or 2 40 per call 6 3 5 Full System Reliability Purpose The purpose of conducting the final system weekend test February 24 27 2006 was to prove that the system could function and transmit data reliably An analysis of the number of successful data points received as compared to the number of points recorded gave an impression on the overall reliability of the data transmission system Set Up The entire system required assembly for this test to be successful On the remote end the datalogger was connected to the Iridium transceiver through a null modem cable The PV panel enclosure temperature sensor and Iridium battery were all connected to
20. 16 8262 21 2536 02 27 06 09 15 00 22 3641 12 0477 12 5291 16 777 21 4149 02 27 06 09 16 00 22 3347 12 0476 12 5292 16 5423 21 512 02 27 06 09 18 00 22 3347 12 0475 12 529 16 1907 21 1901 02 27 06 09 19 00 22 3076 12 0534 12 5288 16 2649 21 2867 02 27 06 09 20 00 22 3076 12 0477 12 5292 16 3565 21 1904 02 27 06 09 21 00 22 2782 12 0531 12 5292 16 4275 21 0944 02 27 06 09 22 00 o 22 2511 2 2511 12 0531 12 5309 16 8261 20 9664 9 o OO Figure 48 Website s Dai s Database View This view is meant for viewing a smaller amount of data Figure 48 shows the data from 9 00 AM to 9 22 AM on February 27 2006 This view shows the same values as the graphs except it shows the exact numerical value for each minute The database is useful for the scientist to begin analyzing the data 59 6 Testing and Results This section documents the testing and results for the project system The purpose set up steps and results are described in each test sub section Test sections include the CR1000 code Iridium network Python code and the full system 6 1 CR1000 Code Tests To develop the CR1000 code it was important to understand how to perform each function required of the end solution The following sections document code that was used to test the CR1000 functions related to bi directional serial communications data collection transmission and control port switching 6 1 2 Hello World Purpose The H
21. 2006 http www nalresearch com QuickRef_Gateway html P Data Transport Network December 2005 http transport sri com TransportDevel howitworks 16 Remote Data Retrieval with the Data Transport Network Valentic December 2005 http transport sri com files PolarTechSlides pdf 17 Personal Communication with Deka Batteries January 2005 18 Communications Network for a GPS Atmospheric Imaging System Candlish Daniels Hamman Lynch WPI MQP C05 Solar Electric Power Association November 2005 http www solarelectricpower org power pv_q amp a cfm 03 Alaska Average Insulation November 2005 http www absak com design sunhours html Charging the lead acid battery November 2005 http www batteryuniversity com partone 13 htm World Communication Center December 2005 http www weclp com Service Iridium SBD 23 Iridium Satellite Date Services White Paper Version 1 0 June 2003 Campbell Scientific Institute Dataloggers December 2005 http www campbellsci com cr1000 128 25 Deka 8G31 12V 100Ah gel cell battery February 2006 http us st11 yimg com store1 yimg com I yhst 705 15121304670_1886_70944078 129
22. Drawn when taking Measurements esesseeesesreesesreerreerrsrresrreresresss 77 Transmission Mode Current PrOBIG oot tetris etate te iare 78 Effect of temperature on Battery Capacity in Kotzebue AK 79 Expected Energy Output for Each day of the Year 80 Indium Dattety Stats Df Charge s sessssrsierestnsesiei ieiti iiaii 81 Datalogger Battery State of ChaT ge sccestcsecasehentessdecesueadh ctesenescauaiebeateaatentacss 81 Fall sensing SEUODEE d cese cuite pe esiti dai ena rie ita eias o iino pid ten iU d e gees 117 DDatalogser ene lise cocoa QN UII EE DU an bM Utd NU UD UA AREE RD TENE UteE 117 Datalogger with all connections necessary for communications 118 VDIV10 1 voltage divider for reading greater than 5V 119 M unbme panel GOBBOCHODS uus iioii i ADR e e ERI RF AE EERO 120 Mounting Panel Sebembsflesediacses ngehtld iaucibr de e idle upe doa dide nudas 121 Deka 8G31 12V 100Ah gel cell battery ec eciscs occ cedsaxeurcecsventeeaaneuelerssuactecabnerage 122 Header Information siens iseen aia 123 Scan Store Interyals EXalple see iioc dried Rar odierna udi 124 Read Sensors Surin scies ior iL ai ie EoE eos Ebo Sas p Ut pss a L Ege i nb dais 125 Data Table Declaration iiss oi eor bt acted e aisi todo Sis asa c i SM ANE 125 vi Table of Tables Table 1 Comparing the Amount Frequency of Data to Send ertet 21 Table 2 Cost per c c 2
23. Initialize Local Variables Y Send Header While Time Stamp of the Last Transmitted Record Time Stamp of Iterated Record Y Get the Time Stamp of the Current Iteration i Send the Time Stamp of the Current Iteration Y Send The Column Data of the Table y Increment Record Index Send End Data Tag Y Exit Sub Figure 34 Data Out Flowchart 5 3 6 Accept New Transmit Period Immediately following the completion of the send data routine a subroutine which accepts a new transmission period from the local user is executed This is essential because it allows the user to change the transmission frequency and gives confirmation that all the data was sent successfully This is particularly useful if energy availability is low The subroutine starts by executing a SerialIn instruction which waits for the termination characters PEND or a time out period of three minutes If a string PEND is received the program checks the received characters to find a PSTR tag The PSTR tag signifies the beginning of a new transmission period and the PEND signifies the end of this information Any characters between these two tags will be the new period adjustment information The transmission period information is then placed into a 32 bit integer container This binary data is a 4 byte integer to represent the number of seconds upon n
24. LO LO Lo 10 LO e o e o e e o e e o o e e e e e e e e e e e e e al N al N al al N al N N N al us i ng ino od M us s i ins ie X T x x d I N e D N o o o a T T T Day Figure 18 Estimated Solar Power Output 17 2 5 4 Charge Controller A charge controller controls the voltage or current delivered to the battery to prevent damage from overcharging and other irregular charging The best method to charge a lead acid battery is in three stages In the first stage a constant current is applied to the battery charging it to a certain threshold The second stage applies a constant voltage to saturate the battery Then in stage three a float voltage is applied to account for internal resistance losses in the battery See Figure 19 Voltage Cell Charge Current j j Stage 1 Stage 2 i sap 3 Constant current 1 Constant voltage Charge I I Current A Voltage iV Time hrs Figure 19 Three Stages of Charging L A Battery The simplest charge controllers only work in one or two stages These charge controllers are basically a switch which provides charging power to the battery until it reaches a certain voltage More modern charge controllers work with pulse width modulation PWM These charge controllers constantly monitor the voltage of a battery and change the duty cycle of the recharge voltage and current to most appropriately charge the battery This helps to maximize the am
25. None self data def append self datum self data append datum def getMeasurements self return self measurement def getUnits self return self units def setRange self range self range range def getRange self return self range def getData self return self data class Line def __init__ self x None y None 112 self x x self y y def getX self return self x def getY se f return self y class PlotGenerator def init self connection None FUNCTION sys getframe f code co name 429 logging debug s connection s X FUNCTION connection if connection self connection connection self local connection False else self connection self connect self local connection True def del self if self local connection self connection close def connection self FUNCTION sys getframe f code co name try logging info s connecting to s s FUNCTION dbname user connection pg connect dbname dbanme user user except pg InternalError inst logging error s db connection failed s _ FUNCTION connection None inst return connection def generate self date range pnd dir var tmp FUNCTION 2 sys getframe f code co name a logging info s s png dir s FUNCTION date range png dir readings self getReadings date range if readings None loggin
26. Open RS232 Port and set baud rate to 2400 Buffer Size is 2000 bytes SerialOpen ComRS232 2400 0 0 2000 90 Delay to let Iridium boot up Delay 1 20 sec SerialFlush ComRS232 Call Connect Dial Up connect success Call Mark Time last time sent If connect success Then Call SendData Call ReceiveNewPeriod Call Hangup EndIf Set port 2 to low to turn off transceiver PortSet 2 False Close the Serial Port SerialClose ComRS232 Exit Sub End Sub BeginProg MAX_CONNECT_TRIES 3 REMOTE_IR_PHONE_NUMBER 00881693151118 Xmit_period 60 60 24 seconds Dim now call Mark_Time last_time_sent Xmit period 1 transmission every day ts_of_last_sent_record 0 Scan Every n Seconds Scan 30 min 1 0 Scan 1 sec 1 0 Read all sensors Call Read Sensors Call Data Tables and Store Data CallTable Tablel Call Mark Time now Calculate the Change in Time from when the last transmission occured time Since sent now last time sent If it is Time to Transmit then call Transmit Data If time since sent Xmit period Then Call Transmit Data EndIf 91 NextScan EndProg 92 Appendix B Data Managment Code The following codes were run on the local computer and used to receive organize and make a database and graphs of the data Reading An object of type reading has a transmit timestamp version timestamp phone number and col
27. a variety of applications and can be embedded into other programming languages such as C 2 5 Power System To keep the overall system functioning a constant power source will be needed to provide the energy needed to collect compute and transmit data The power system will need to be self sustaining and independent from any power grid This requires energy storage and energy renewal VECO Polar Resources has assigned the task of designing the power system to Tracy Dahl an engineer from Colorado The WPI team therefore must collaborate with Tracy to make sure the power system is adequate for the project A good understanding of the power system components is necessary for the team 2 5 2 Battery The main source of energy for the system will be two 100 Ah batteries lead acid batteries In all batteries a chemical reaction inside the battery produces a voltage across the output terminals An important battery property to be considered is storage capacity or the amount of energy a battery can hold Batteries are rated to a certain voltage and Ampere hours Ampere hours Ah is the amount of current supplied at the battery s voltage multiplied by the hours it is being supplied So a battery that supplies 5A for 10 hours will have a rating of 50 Ah To find the Ah rating needed for this system the amount of input current needed to power the system needs to be known as well as the longest amount of time the battery may go without being recharg
28. can then answer the call and data can be sent Once the connection is established the intermediate connection with the Gateway is dropped the data is transferred directly from one ISU to the other through the satellite network Total set up time is estimated to be 25 seconds 15 seconds shorter than ISU to PSTN Both of these methods have a data rate of 2400 bits per second and cost 1 20 per airtime minute ISU to ISU requires two transceivers and ISU to PTSN only one However SRI has the resources for the system to have two transceivers and therefore this along with the shorter set up time has led to the choice to use ISU to ISU 25 4 3 Processing One of the fundamental challenges of this project was to create an interface between a Campbell Scientific Datalogger and an Iridium transceiver A major design decision that was faced with was deciding whether to use an external microcontroller to interface between the two devices or to perform all processing using the datalogger s internal CPU The internal processor on the CR1000 is a 16 bit Hitachi H8S 2322 Ideally the system would not use an external controller both to limit power consumption and to make a more simple system Energy consumption was one of the highest priorities because the power budget for this project will be limited particularly in winter months While an external microcontroller would draw additional energy it would not be significant enough to make this decision so
29. could be implemented Iridium Battery eur Cery Ure Figure 29 shows the schematic of the circuit used for this project 32 Iridium Battery 12V IRF9520 2N2222A Figure 29 Switching Circuit Schematic As seen in Figure 29 the only difference between the Iridium switching circuit used for this project and the sample circuit provided in Campbell s documentation is the power supply voltage and the transistor connected directly to the Iridium s power input The power supply used for this circuit was a 12V 110Ahr gel cell battery 5 2 3 Circuit Operation The functionality of this circuit is simple it is either off or on and supplying current to the Iridium modem When the datalogger s control port is off the BJT will not let current flow from the 12V battery Thus the voltage between R1 and R2 will be approximately equal to 12V When the voltage at this node is 12V the MOSFET will also be off Vgs 0V and not providing current to the Iridium modem As soon as the datalogger s control port is set high 5V the BJT will turn on allowing current to flow through from the Vcc This will drop the voltage between R1 and R2 to approximately 2V which means Vas will be greater than the MOSFET s threshold voltage of 4V and current will flow 5 2 4 Part Selection The 2N2907 BJT from CST s sample circuit needed to be replaced with a transistor that could handle a larger current through it For this purpose an IRF952
30. destination email server hosted by the Value Added Reseller for processing of the data message The architecture of SBD lends itself to integrate well with the Data Transport Network Data messages sent from the system will arrive in an email message as an attachment which can then be easily routed into the Transport Network A data dial up solution requires a program to collect the data being streamed across the network The problem with SBD is that the messages being sent are projected to be much larger than SBD is designed to handle The length of these SBD messages can range from 0 to 1960 bytes of data The message to send is expected to be 12 KB requiring multiple SBD messages for transfer 23 When sending multiple SBD messages the costs of communications skyrocket SBD charges per byte sent it costs 0 10 for the first 30 bytes and 0 003 for each additional byte A full 1960 byte message would therefore cost 5 89 Figure 22 compares the costs of sending larger SBD messages with the costs of using dial up airtime minutes Cost Analysis SBD v Dial Up 35 00 30 00 25 00 z 20 00 15 00 10 00 5 00 0 00 TTFTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTFTTT TT TTTTTT TT TTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTTI 100 2600 5100 7600 10100 of bytes sent SBD Dial Up Figure 22 Costs of SBD v Dial up Although SBD offers a simple solution to transferrin
31. end of a transmission indicating the remote system will go into this extremely low power mode and only transmit alive messages using SBD A similar extension to this project is to make the system in both SBD and dial up modes and be configured to either prior to deployment The mode selected would be based primarily off the volume of traffic expected For the Sullivan project dial up was found 83 to be the most cost efficient means of transfer but if the system were used for future deployments transmitting less data SBD capabilities might be preferable 7 3 Mobile Terminated Calling Another desirable feature that was discussed during the course of this project but never included in the final design is the ability to perform a mobile terminated call to modify program settings on the remote end The local end does send a tag at the end of each transmission indicating the period of transmission but expanding on this could add some very powerful options For example the ability to use Loggernet to either modify the datalogger s code or to load a completely different program on the datalogger from the local end could introduce some interesting prospects such as switching from ASCII to binary communications or disabling enabling particular sensors Additionally if problems occurred somewhere in the system being able to find and fix them over the satellite link could save the cost of a researcher returning to the site 7 4 Camera for Datalo
32. graphs of the collected data is to use the function getReadings to obtain the data from the database This function will get the data from when the data collection began until the present Figure 44 shows a basic overview of the plotGenerator program 53 getReadings Y Set Measurements Units amp Ranges NN For Column 0 to 4 getLine plotLine Next Colum Figure 44 Overview of plotGenerator Next the variables measurements units and ranges are declared as tuples A tuple is very similar to an array except instead of being associated with a position the elements have a unique key that are not dependent on location The three declarations can be seen below measurements 0 PTemp 1 Batt_volt 2 Batt Volt IR 3 PV_Voltage 4 ETemp units 0 Celsius 1 Volts 2 Volts 3 Volts 4 Celsius ranges 0 None 1 0 15 2 0 15 3 0 30 4 None The variable measurements is what the sensor is reading PTemp is the panel temperature of the datalogger Batt_volt is the voltage of the datalogger battery Batt_Volt_IR is the Communications systems battery voltage PV_Voltage is the voltage of the solar panel ETemp is the temperature inside the enclosure The numbers before the values in measurements are their individual keys Units and ranges use these keys to associate a 54 proper unit and ra
33. is where this information will go Replace Paddy0 Paddy1 PaddyN with the corresponding variable or sensor description names These must match with the order of the columns formatted by the datalogger The columns will go in the same order as declared in the DataTable instruction which will be described in the following sections 2 2 Sensor Readings When editing the final code it is necessary to change parameters in reading sensors and storing data The following describes steps to customize these settings inside of the final CR1000 code 2 2 1 Scan Data Storage Interval Changing the scan or sensor reading intervals requires editing two numbers in the CRBasic code The BeginProg structure of the CRBasic code shown in Appendix A shows an instruction called scan The scan instruction is used to control the looping inside of this function The instruction also controls how often the sensors are being read There are 4 parameters which are explained in the Campbell s CR1000 user manual or using the help menu in the CRBasic editor The first parameter of the scan instruction controls how often the scan occurs and the second parameter is used to control the scan units Editing these parameters will change the scan rate inside of the main program It is important to make the distinction between taking and storing a sensor reading The interval for reading a sensor is controlled with the scan function however storing a sensor reading i
34. spikes to about 22mA but for only about 2 seconds The other mode the datalogger operates in is transmitting mode During transmissions the datalogger must stream the data collected to the RS232 port requiring extra current To simulate the amount of data that will typically be sent by the system data was logged once per second to expedite the process The results are shown in Figure 61 78 Current Drawn By Datalogger during Transmissions Current mA Transmission Period 9 99 OPEV PHP My N N Time sec PPP usb O Ao C Qi X P should be disregarded since this is when the datalogger is measuring sensors once per Figure 61 Transmission Mode Current Profile The average current during transmissions is 28 9mA The spikes before the transmission second and will not apply to the final system only for the test 6 4 4 Battery State of Charge Expectation To prove that the system is able to operate on the two 100Ah batteries with solar charging the amount of energy consumed by the system as well as the amount of energy harnessed by the solar array must be computed Table 8 shows the estimated energy consumption from the communications and Datalogger According to the estimates the entire system will only consume about 10Ah for an entire year a very low power system System Total Table 8 Energy Consumption of System Dr
35. the same way except they use fiberglass to hold the electrolyte in place instead of gel When testing the gel cell versus the AGM battery the previous year s team found that the AGM battery s recharge characteristics are much more favorable than the gel cell The AGM is able to recharge over a greater range of voltages resulting in 96 minimum recharge efficiency while ideally the gel cell recharges at 90 efficiency Secondly the AGM is rated for twice the amount of lifetime discharges to 30 of capacity than the gel cell These results proved the AGM battery to be superior to the gel cell for the specifications of their system 2 5 3 Power Generation In order for the power system to be self sustaining some sort of power generation is needed The power generated recharges the battery and should be designed to keep the battery charged above 30 of its capacity Also there needs to be a charge controller between the power generation component and the battery to properly charge the battery Perhaps the simplest and least expensive form of renewable energy for Polar Regions is through photovoltaic cells PV cells These cells convert sunlight into DC power The 16 advantage of using solar power is that it requires no moving parts to produce energy A solar array simply lies on the ground or is mounted on a pole Also the efficiency of a solar panel actually goes up in cold temperatures Generally the efficiency of PV cells increase 0 596
36. timeunits signifies one of rTime s parameters and the second value passed time is an integer which will be compared to one of rTime s parameters specified by timeunits The subroutine receives the 2 parameters then proceeds to access the real time clock on line 15 which updates the array rTime every second Lines 16 21 implement an if statement which turns on control port 1 if the value in one of rTime s parameters specified by timeunits is equal to time The command Portconfig Mask Function on line 17 is used to configure control port 1 as an output port Mask specifies which port to configure amp B1 port 1 and Function configures the port as either input or output 1 output 67 The command Portset Port State on lines 18 and 20 activates a port either logic high or low Port denotes which port to effect and State indicates logic high or low False logic high 5V True logic low 0V Testing To test the program a digital multi meter was connected to control port 1 The program was executed and the real time clock was watched During the second half of a minute between 30 and 00 the port went high otherwise it was low Conclusion While this program is simple it does accomplish two functions that are critical to the project running off the real time clock and using control ports to power on the transceiver The previous programs ran off of a counter that simply incremented with each scan comm
37. to the sensor readings as shown in Figure 75 This example uses only one data table however it might be desirable in some cases to add more tables Please see the CR1000 user manual for further customizable options 2 3 Data Transmission Interval The data transmission interval is the period between transfers This period is initialized at the beginning of the BeginProg structure shown in Appendix A CR1000 Code The units are given as time in seconds Changing the period can also be done locally as described in the design documentation section Keep in mind that changing the period will also change the amount of records to be sent upon the next transmission To change the transmission period it will be necessary to edit the python code running the local end to send whatever the desired transmission period This period must be divisible by one hour and a maximum of one transmission per two weeks 3 Links and Resources Title Links 2 22 06 Description CR1000 Overview Attp www campbellsci com documents The CR1000 Overview manuals cr1000 ov pdf gives a brief description of the basic functions of the CR1000 datalogger Physical Specifications 126 are given PC200W ftp ftp campbellsci com pub outgoing f iles pc200w_3 1 exe Campbell s PC200W software to directly communicate with the CR1000 datalogger See the included help files for more information Campbell Downloads
38. week Bi directional communications will allow the local end to provide the system with confirmation that all the data was received In addition it would be desirable for the local end system to have the ability to change the transmission period should energy resources become scarce 3 2 Power Specifications The overall system must be powered perennially without maintenance The main power source is two 100 Ah 12 V gelled electrolyte batteries Deka Model 8G31 Whenever sunlight is present a 20W photovoltaic panel with a Morningstar Sunguard charge regulator will charge the batteries The two power specifications the system must meet are e Never allow the battery charge to fall below 30 e Operate for a complete year without power failure It is desirable to run the datalogger on one of the 12 V batteries and the communications system on the other By isolating the power to the datalogger and the communication 19 system the datalogger will still have power and be able to collect data if the communication system power fails for this design refer to Section 5 2 3 3 Physical Specifications The system is to be deployed in Kotzebue Alaska where the lowest recorded temperature is 46 C in February 1968 The physical specifications that must be met are e The system must operate at 40 C e The system must be able to withstand the harsh artic meteorological conditions As for the enclosure the electrical system including batteries
39. will be placed in a plywood box with 4 of insulation to provide a relatively stable environment The electrical components will be mounted on a panel and secured to the top of the box This box will open from the top keeping the small components out of harms way when the two batteries are moved in and out of the box The PV panel will be mounted vertically to shed ice and snow on a poll 2 5 meters high Figure 20 shows how the system will be set up North East 45m _ gt ee 54m Mela MET at 2m 25m Guy wires at 1 5 m 4 Guy wires at 15 m 4 25m D 7m 10 cm re bar guy anchors Figure 20 Physical setup of the System 20 4 Design Choices The following sections outline the major design decisions for the overall system These choices included the quantity and frequency of data transmission using short burst data vs dial up communications processing capabilities and data transfer methods All decisions were made based on several factors each critical to the success of the project 4 1 Quantity of Data amp Frequency of Data Transmission For the final deployment it was important to choose how often data should be sent from the remote end to the local user The specifications state that the data must be sent at a minimum of one sample per week but ideally the system would send the complete set of data daily po Cost Energy Convenience Toia Weight 02 o2 06 1 Once per Week One Sample 100 10
40. will continue as previously operated The transmit data subroutine is then ended by the execution of the datalogger s hang up routine 37 Establish Connection Send Data From Table Accept Period String Tags Accept New Tx Period If Any Mark The Time of Last Sent Transmission Exit SubRoutine Figure 32 Transmit Data Flowchart 5 3 4 Establish Connection Establishing a connection with the Iridium transceiver requires the use of an RS232 connection and bi directional communication The transceiver can be operated through the use of AT commands which can be executed by sending and receiving strings over the serial port The following section explains the subroutine used to establish communication with the Iridium transceiver The first step in establishing a connection is to initialize the AT command strings for establishing a connection and to open the RS232 port Following initialization the subroutine enters a connection retry loop Inside of the connection retry loop the program sends the ATDT command to establish a connection If the connection attempt fails then the program delays and retries establishing a connection with a limit of three attempts When the transceiver has successfully established a connection or exhausted the connection attempts the subroutine is exited Figure 33 shows this process 39 Initialize AT Command Strings i Open RS232 Com Port ot For Reconnect 1 To3 i Flu
41. 0 00 58 0 00 58 0 00 56 0 01 04 0 00 58 0 01 01 0 01 01 0 00 59 0 01 00 0 01 02 0 01 01 0 01 02 0 01 00 0 01 00 0 01 11 0 01 03 0 00 57 0 01 00 12116 12291 12288 12298 12362 12302 12399 12312 12391 12302 11609 12273 12012 11901 12198 11015 12293 12297 12258 11671 12221 11353 11709 12159 12250 12028 12397 12105 12326 12140 12378 12368 12288 12150 11682 12355 11590 12381 12377 12268 12279 12338 12366 12298 12339 12281 11922 12251 11577 12299 73 201 93 195 10 211 86 208 44 213 14 212 10 206 65 192 38 210 02 192 22 170 72 201 20 182 00 205 19 210 31 211 83 201 52 201 59 204 30 197 81 210 71 195 74 209 09 206 08 204 17 194 00 199 95 201 75 205 43 209 31 199 65 199 48 211 86 209 48 208 61 193 05 199 83 202 97 202 90 207 93 204 65 199 00 202 72 198 35 205 65 204 68 167 92 194 46 203 11 204 98 15 00 30 15 00 40 15 01 36 0 01 06 0 00 56 11657 15 59 12 15 59 21 16 00 19 0 01 07 0 00 58 11985 16 59 26 16 59 35 17 00 52 0 01 26 0 01 17 12377 17 59 53 18 00 02 18 01 02 0 01 09 0 01 00 12306 20 00 27 20 00 37 20 01 34 0 01 07 0 00 57 11643 21 00 45 21 00 55 21 01 54 0 01 09 0 00 59 12350 22 01 06 22 01 15 22 02 15 0 01 09 0 01 00 12355 0 01 47 0 01 57 0 02 54 0 01 07 0 00 57 11624 1 02 15 1 02 25 1 03 33 0 01 18 0 01 08 12347 2 02 43 2 02 52 2 03 53 0 01 10 0 01 01 12289 3 02 46 3 02 56 3 03 55 0 01 09 0 00 59 12342 4 03 06 4 03 16 4 04 16 0 01 10 0 01
42. 0 P Channel MOSFET was chosen with a drain source current of 6 8A This power FET was chosen for its current rating its Vos of between 2V and AV and its low static on resistance Rps 0 60Q The 2N2222A NPN BJT was not changed for this circuit It was designed 33 for high speed switching applications with collector current under 500mA which will not be exceeded in this circuit The resistors chosen for this circuit needed to be relatively high valued to limit the current draw The ratio of R1 to R2 was also significant The 10kQ and 2kQ resistive divider was used to provide 12 volts to the gate of the IRF9520 MOSFET when the 2N2222A is turned off and low voltage about 2V to the MOSFET gate when the circuit is turned on After analyzing the operation of the circuit it was concluded that the resistor values used in CSI s sample would work for this application 5 2 5 Control Port For the remote system to operate autonomously the Iridium transceiver would need to be powered on periodically as controlled by the datalogger Each of the datalogger s eight digital I O ports can be configured as an output port and set to either high SV or low OV using the portset instruction These digital output ports are often used to control switching circuits but not to provide significant power because the port itself has limited drive capabilities 2 0mA at 3 5V To implement a control port the datalogger program uses two commands Portconfig Mask Func
43. 1 index2 initialize counter Dim J Dim buffer size buffer size 100 Dim end str AS STRING 4 Dim start str AS STRING 4 Start str PSTR end str PEND Initialize New Period New Period 0 Wait for PEND String to Verify that the wait PERIOD for next transmission will occur 88 SerialIn InString ComRs232 18000 end str 100 Check for end str string For J 1 TO 90 Step 1 If mid InString J len end str end str Then index2 J EndIf If mid InString J len end str start str Then indexl J EndIf Next J If index2 index1 4 0 Then New Period mid Instring indexl 4 index2 index1 4 EndIf If The Time Between Transfers is greater than 2 weeks or less than 1 hour change to 1 transmission per day If New PERIOD mod 60 60 0 and New PERIOD gt 60 60 and New PERIOD 14 24 60 60 Then Xmit period New PERIOD EndIf Exit Sub End Sub SendData SubRoutine Sends Data for a specified number of records Sub SendData Dim MyPhoneNumber AS STRING 40 Dim ASCII Record AS STRING 1000 Dim k Dim MAX REC MyPhoneNumber 00881693151117 MAX REC 250 Assuming 24 records per day for 7 days Send The Header With Column Names SerialOut ComRS232 S 0 500 SerialOut ComRS232 MyPhoneNumber 0 500 SerialOut ComRS232 CHR 13 CHR 10 0 500 Health Status columns SerialOut ComRS232 PTe
44. 15 51 21 16 52 44 17 51 56 19 52 35 20 53 03 21 53 18 22 53 44 23 54 04 0 54 15 1 54 35 2 54 58 3 55 22 5 56 01 6 56 21 7 56 35 8 57 00 9 57 15 11 58 06 12 58 17 20 53 19 21 53 25 22 53 53 23 54 03 0 54 30 1 54 36 2 55 09 3 55 21 4 55 59 5 56 03 7 56 54 8 57 06 11 59 55 12 58 16 13 58 53 22 46 51 23 46 58 0 47 26 1 47 38 2 48 05 3 48 16 5 48 55 6 49 24 7 49 46 8 51 42 9 50 18 10 51 59 11 51 01 12 51 20 13 51 39 14 52 01 15 52 23 16 53 42 17 52 54 19 53 31 20 54 07 21 54 16 22 54 45 23 55 05 0 55 14 1 55 35 2 56 00 3 56 23 5 57 03 6 57 21 7 57 35 8 58 11 9 58 18 11 59 03 12 59 17 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 14 17 10 0 01 0 01 0 01 10 13 08 07 08 08 08 13 11 0 02 58 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 10 10 09 0 01 12 08 08 06 13 08 0 01 10 09 10 12 11 11 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 0 01 09 21 13 06 10 0 01 0 01 0 01 0 01 0 01 07 07 02 11 13 09 09 08 07 06 08 08 it2 11 11 0 01 00 0 01 03 0 00 58 0 00 59 0 00 58 0 00 58 0 01 00 0 01 04 0 00 59 0 01 04 0 01 08 0 01 01 0 01 06 0 00 58 0 00 58 0 00 52 0 01 01 0 01 01 0 01 00 0 00 59 0 00 58 0 00 58 0 00 56 0 00 59 0 01 00 0 01 02 0 01 02 0 01 00 0 01 00 0 00 58 0 01 02 0 01 02
45. 22 2006 01 25 UTC Degrees C Temperature 45 0C 49 0F Wind chill 578C 720F Wind speed 6kts 6 MPH Wind direction 215deg SW naa eae jUNSEUE IESE DEURSE RETESET 02 04 06 o8 10 12 14 16 18 20 22 00 Wednesday February 22 2006 Plot Duration pw View Wind Speed t Webcam Knots 04 06 08 10 12 44 16 18 20 22 00 Wednesday February 22 2006 Vind Direction Degrees Fullsize Archives Tug 7U4 us us au 42 14 46 48 20 41 wednesday February 22 2006 This page will automatically refresh every 10 minutes Copyright C 2006 SRI International All rights reserved For weather conditions in Kangerlussuaq see tha Gandractram du ig Figure 2 GEOSummit webpage www geosummit org Figure 2 shows a screen shot of the website relating data from GEOSummit in real time The table in the upper left corner presents current conditions at the Summit station In the center of the page there are graphs continually updated displaying outside temperature wind speed and wind direction Another interesting feature is the live webcam view which is shown on the left of the page For Paddy Sullivan s research project the WPI team will create a website similar to this one continually updating the page with meteorological data from Paddy s remote station in Alaska 2 1 5 A New Approach to Remote Data Trans
46. 4 0 j _ 10 cm re bar guy anchors 117 Figure 66 Full sensing station Datalogger Connections The datalogger will be mounted on the tower in its own Campbell Scientific secure enclosure Figure 67 shows the datalogger mounted in the enclosure with space at the bottom for a multiplexer gt P Figure 67 Datalogger enclosure Several wires will need to be fed from the datalogger s enclosure to the insulated box This is necessary to power the datalogger to communicate with the satellite modem and to collect data from the insulated enclosure Figure 68 shows all of these connections on the proper terminals 118 Differential voltage sensor Communications battery with VDIV10 1 Differential voltage sensor PV Panel with Thermocouple T107 Collects temperature data inside enclosure Datalogger Power and Ground Null modem cable Inserted into R3232 Port Figure 68 Datalogger with all connections necessary for communications The connections made from the datalogger s enclosure to the insulated box are explained in detail below Power and Ground wires are connected to the datalogger s power terminals from the datalogger s 12V battery in the insulated box These two lines are the Brown power and paired White ground lines in the eight conductor cable One control line is run from the datalogger s digital I
47. 46 Once per Day One Sample 14 70 47 6 Once per Day Complete Data Set 9 100 Table 1 Comparing the Amount Frequency of Data to Send Table 1 shows a value analysis chart of the each of the three choices The weights were chosen due to its importance to the project On this scale zero represents no importance with ascending importance until one Each decision is given a value from zero to one hundred for cost power and convenience This number was chosen by how well each choice meets the ideal situation with zero being the worst case and 100 being the best case The reasoning behind the selection of these numbers will be discussed in the pertinent sections below After weighing the options it was decided that a daily transmission of the complete set of data was the best option One specification of the system is that the communication cost cannot exceed 2400 per year There will be approximately 100 sensors in the field The data from these sensors are each four bytes Each sensor will take one reading every hour or 400 bytes per hour Therefore one sample is one reading from each of the 100 sensors in the field combined with some metrological data such as temperature and also some system health checks such as battery voltage This makes one sample approximately 500 bytes A full day s set of data would have twenty four samples one for each hour of the day and would be approximately 12KB Cost for Transmission per year Once per Week One
48. 52 13 25 16 0 00 24 1944 4 0 15 07 13 39 59 13 40 28 0 00 29 2923 5 0 20 23 14 00 22 14 00 53 0 00 31 0 6 0 02 02 14 02 24 14 02 56 0 00 32 3986 6 0 24 33 14 24 55 14 25 31 0 00 36 4481 7 0 31 05 14 56 00 14 56 44 0 00 44 6001 8 0 34 01 15 30 01 15 30 49 0 00 48 6868 9 0 35 21 16 05 22 16 06 08 0 00 46 6464 10 0 43 45 16 49 07 16 50 01 0 00 54 0 10 0 02 32 16 51 39 16 52 42 0 01 03 9170 11 0 48 23 17 40 02 17 41 08 0 01 06 9806 12 0 54 59 18 35 01 18 36 10 0 01 09 11284 13 1 00 01 19 35 02 19 36 15 0 01 13 12327 14 1 04 58 20 40 00 20 41 21 0 01 21 13311 15 1 10 02 21 50 02 21 51 29 0 01 27 14459 16 1 15 10 23 05 12 23 06 42 0 01 30 15523 17 1 20 12 0 25 24 0 27 00 0 01 36 16496 18 1 24 57 1 50 21 1 52 03 0 01 42 17420 19 1 29 49 3 20 10 3 21 58 0 01 48 18568 20 1 35 05 4 55 15 4 57 10 0 01 55 19573 21 1 40 06 6 35 21 6 37 19 0 01 58 20631 22 1 44 46 8 20 07 8 22 09 0 02 02 21647 23 1 50 12 10 10 19 10 12 29 0 02 10 22783 Table 6 Overnight Frequency amp Reliability Test The first column contains the record number which shows that 24 records were sent In this column record 6 and 10 appear twice This is because they did not successfully send on the first trial It took a retry before these two records were sent The next column is where the period information can be seen All of the periods were successfully transferred and upd
49. CRBasic s special instruction set to periodically measure and store data into tables Additionally CRBasic offers many hardware interface commands to integrate external devices Using the RS232 serial commands with CRBasic to program the CR1000 was required to integrate with an Iridium transceiver 2 4 Sending Data SRI funds projects in many remote locations around the globe Communication in some of these areas is often difficult This can make transferring data to and from these remote locations quite a challenge The Iridium Satellite Network paired with SRI s Data Transport Network will allow information to be sent to researchers from anywhere on earth 11 2 4 2 Iridium Satellite Data Network The Iridium Satellite Network uses three main components in its operation the satellites an Iridium Subscriber Unit ISU and Iridium Gateways There are sixty six low earth orbiting LEO satellites It is the only satellite network whose coverage spans the entire globe including Polar Regions oceans and airways At any time there is at least one satellite covering every region of the globe The Iridium Data Network transmits data to and from areas where no other form of communication is available Figure 13 Iridium Satellites cover the Earth The satellite network consists of 66 operating satellites as well as 14 orbiting spares The satellites are arranged into 6 polar orbiting planes with 11 satellites in each plane The orbiting altit
50. Net software would require no remote computation to send data The LoggerNet software would simply initiate a connection pull and manage data from the CR1000 This is a very simple design strategy and would also in turn lower the complexity of the final system Timing complexity was another important factor in the decision Establishing a connection between remote and local ends would require an Iridium transceiver to be awake and active at an expected interval for a mobile terminated approach This approach could be potentially disrupted from a system clock drift The CR1000 clock is rated to within 5 min per year To compensate for this clock drift longer Iridium active windows could be implemented however this increased complexity is undesirable The energy consumption could be greater using the mobile terminated approach As mentioned earlier the increased timing complexity would require an increased duty cycle for the Iridium transceiver of at least 10 more active minutes This increased duty cycle is unwanted and would increase the energy consumption of the system According to the testing and results section the average current draw of the Iridium transceiver during a transmission session is about 245mA lasting approximately 2 minutes Multiplying these values together gives an energy consumption value of approximately 8 1mAH During an idle period the average current draw is 60mA Adding an idle period of 10 minutes to compensate for a
51. O port 2 C2 to the circuit box and is used to activate the switching circuitry inside the insulated box This line is Blue wire in the eight conductor cable Two wires running from the datalogger s differential measurement terminal 2 which is connected through a CSI voltage divider are used to sense the Iridium modem s battery voltage Be sure to connect the positive battery terminal to the high or H terminal on the datalogger s differential measurement panel These two lines are the Green and paired White wires in the eight conductor cable 119 Two wires are used to sense the PV voltage they also run from the datalogger s differential measurement terminal 3 which is connected through a CSI voltage divider to the terminal strip shown in Figure 70 These two are the Orange and paired White wires in the seven conductor cable A null modem cable is linked between the datalogger and Iridium transceiver s RS232 port inside the insulated box A thermocouple T107 is used to measure the temperature inside the insulated box Four wires must be connected to the datalogger to operate this temperature sensor The red wire should be attached to the SE 1 terminal the black wire to the EX1 terminal and the Purple and Clear wires to a ground terminal CSI Voltage Divider Two CSI VDIV 10 1 voltage dividers are used in this system The datalogger can only measure voltages up to 5 volts so to measure the PV panel vo
52. REMOTE DATA TRANSMISION SYSTEM A Major Qualifying Project submitted to the Faculty of the WORCESTER POLYTECHNIC INSTITUTE and performed at SRI INTERNATIONAL in partial fulfillment of the requirements for the Degree of Bachelor of Science by Eric Hall Peter Kaineg Amanda Quigley Eric Young Date March 4 2006 Approved Advisors Roy Stehle Todd Valentic Andrew Young Professor John Orr Major Advisor Table of Contents WP SU MPH cp DEREN viii Pxecativye UIT socii oon oven iu M CssE pn M C M MEN SENE NM MM DM MD NM DM ix Introd cense ER 1 MES o c o oorr EE 2 2 1 SCLIN CEA AC Sera 100 RENS RN RA Rm 2 21 2 Why Transmit Data in Real Time see eciam dieta enda 2 2 1 3 Why Deploy Arctic Sensing Systems acce seioecis tree eterno src bet ptezetesan 2 2 1 4 Existing Systems for Real Time Data Transmission e 3 2 1 5 A New Approach to Remote Data TransmissioN csccecsseeeeseeeeeeeeeees 4 2 4 Alaskan cg M 5 2 3 Coll cting Data Mc V 8 2 32 System DOSDEIDEIODI iesenii ume on up utu DIS aE Sat t pA Ur USUS 9 225 lag ri ME C ERE 9 2 3 4 vir M 10 4 3 5 Short Cut Redi T 10 2 3 6 dono A n 11 2 4 euni gb rp P 11 2 4 2 Idiom Satellite Data NetWORE ose eoaoc discs icta tn Db ior ade Ep Duae ta uineis 12 2 4 3 Data Transport NeBWOEE ai qu stetit a id tiu
53. RING 40 System vars Public InString AS STRING 100 Public last time sent Public Xmit period Public ts of last sent record Public time since sent Health Status data Public Batt Volt Public Batt Volt IR Public PV Voltage Public ETemp C Public PTemp C Paddy s data Units Batt Volt Volts Units PTemp C Deg C Units PV Voltage Volts Units Batt Volt IR Volts Units ETemp C Deg C Define Data Tables DataTable Tablel True 1 DataInterval 0 1 min 0 Sample 1 PTemp_C IEEE4 Sample 1 Batt_Volt IEEE Gl Gl EJ A Sample 1 Batt Volt IR I Sample 1 PV Voltage IEE Sample 1 ETemp_C IEEE4 Paddy s columns to follow EndTable Lu B 86 Sub Mark_Time temp_var Dim rTime 9 Dim temp_var Alias rTime 1 Year Alias rTime 2 Month Alias rTime 3 Day Alias rTime 4 Hour Alias rTime 5 Minute Alias rTime 6 Second Alias rTime 7 uSecond Alias rTime 8 WeekDay Alias rTime 9 Day_of_Year Read The System Clock RealTime rTime The number 366 is used to acount for the number of days in a leap year if it is not a leap year the 366th day will be skipped temp var Year 2006 366 24 60 60 Day of Year 1 24 60 60 Hour 60 60 Minute 60 Second Exit Sub End Sub Sub Read_Sensors Default Datalogger Battery Voltage measurement Batt_Volt Battery Batt Volt Wi
54. Rate Format TXDelay BufferSize on line 09 is used to open a serial port for communications The ComPort parameter specifies which communications port to use RS232 CS I O or Digital I O The Baud Rate is the speed of transmission and Format is the data type that will be transmitted through the port TXDelay is used to introduce a delay and BufferSize limits the packet size that can be transmitted The command Serialflush ComPort on line 12 is used to clear the communication ports input buffer 62 The command SerialIn Dest ComPort TimeOut TerminationChar MaxNumChars on line 13 is used to read data into a destination array Dest through a specified communications port ComPort This read is terminated if the TimeOut parameter is exceeded if the terminating character TerminationChar is received or if the maximum number of characters is exceeded MaxNumChars Testing To test the Confirmation program the datalogger was directly connected to a PC running hyper terminal The user would receive a start prompt saying Enter confirmation The user would then have 20 seconds to type into the terminal If the confirmation string was typed then the program would output Confirmation_Received otherwise the output would be Try Again Conclusions This program accomplished another key feature of the final program The end solution required bi directional communications with the datalogger which this example program ac
55. Sample Once per Day One Sample 551 15 Once per Day Complete Data Set 876 00 Table 2 Cost per year 21 The cost to send one sample of data is 1 51 As Table 2 shows the daily transmission cost at this rate is 551 15 per year and the weekly transmission at this rate is 78 52 per year The cost to send a complete day s set of data is 2 40 per day or 876 00 per year Even though sending one sample of the data per week is cheaper than the other two methods it is important to notice that all three choices would be well within the yearly budget To choose a weighting from zero to one hundred for the cost parameter Table 2 was used To do this one hundred was chosen for sending one sample per week because it was the most cost efficient Then dividing one sample per day by one sample per week showed a factor of seven Dividing one hundred by seven gave a rating of fourteen The same was done with sending a full day s data everyday and the result was nine The system must run all year without draining its battery source Therefore the energy used by the system must be minimal The energy is calculated by the length of time the transceiver is on and the amount of time the datalogger is collecting and sending data Sending the complete set of data will keep the devices on the longest with once a day next and once a week consuming the least amount of power The energy consumption of the three choices can be seen in Table 3 Ener
56. UNCTION dbname user connection pg connect dbname dbname user user except pg InternalError inst logging error s db connection failed s name inst return reading table Reading ReadingTable connection for reading in readings try reading table insert reading except pg ProgrammingError e if re search timestamp key str e logging warning s duplicate reading _ FUNCTION else raise Ss exiting closing connection to s Q0 s FUNCTION dbname user connection close logging debug 110 return def setupLogging logging format asctime s name s filename s levelname s message s formatter logging Formatter logging format log file logging FileHandler Store log w log file setFormatter formatter log file setLevel logging DEBUG console logging StreamHandler sys stdout console setFormatter formatter console setLevel logging DEBUG rootLogger logging getLogger rootLogger setLevel logging DEBUG rootLogger addHandler console rootLogger addHandler log file if name main setupLogging logging debug s starting name Store sys argv run sys exit 0 plotGenerator usr bin env python 3 port StringIO port logging port datetime port tempfile port time port sys port os port os path port pg 3 3538 H H H H H H H H H 3 385 3 3
57. a datalogger In the past the data was retrieved only when the scientist visited the site which could be as infrequent as once a year Several scientists have explained that their system was working great while they were there but upon their return a year later they found that it failed shortly after they left With real time access to the data they not only have constant monitoring capabilities but can also determine if the system is functioning properly 2 1 3 Why Deploy Arctic Sensing Systems Scientists and researchers are placing increasing importance on understanding environmental effects of changes in temperature moisture and other shifting climate conditions This project will be used specifically to support a research station which will gather information about soil moisture and temperature and the effect of the growth and regression of the latitudinal tree line over time Dr Patrick Sullivan is a researcher supported by a NSF grant and being assisted by VPR Dr Sullivan has hypothesized that the careful study of changes in soil moistures temperature and tree line over time can lead to conclusions about the widespread effects of global warming Evidence implies that temperature has significant control over the latitudinal tree line position Traditionally it has been viewed that rising temperatures are associated with increases in growth of tree line trees and the invasion of forests into tundra land However numerous recent studi
58. account Figure 63 takes into account these variables to provide an estimated energy output of the 20W PV panel 80 Energy Output of Solar Panel by Day Dv M 69 e ol o Watt Hours o 10 5 0 Lo Lo Lo Lo re LO Lo ire Lo LO 1o Lo e e ce e ce e e e e e ce e e e e e e e e e e e e e y e e e e e y e y e e e E Bac iet b c bo ies b nu Es e ceo t ike co N co o e N a Day Figure 63 Expected Energy Output for Each day of the Year According to the estimates the solar panel should provide much more than enough extra energy in the summer months while having no output in the winter This means the only time the batteries should start to become drained is in the winter Integrating the capacity of the battery charging from the solar panel and current drain from the components a graph of the state of charge of the battery was determined Figure 64 and Figure 65 show the energy available in each battery throughout the year Both graphs look similar since the batteries are in the same weather conditions being charged by the same solar panel Also both systems draw close to the same amount of current each day 81 Estimated Energy Available in Iridium Battery 100 00 90 00 80 00 70 00 n 5 60 00 o T 50 00 3 E 40 00 30 00 20 00 10 00 0 00 wo wo wo wo wo 1 wo wo wo wo wo wo e Q e e ce ce Q Q e e e e e e e e e e e e e e e e e Yo A N e e e e e y e e
59. ain A Duration Day hr Energy day Ah Energy year Ah Iridium 0 156253 0 0844 Quiescent 0 0005 24 0 0120 4 3800 Measurements 0 022 0 0133 0 0003 0 1071 Transmitting 0 0289 0 0844 0 0024 0 8908 Datalogger Total 79 Using two 100 Ah batteries with this low power system seems to be excessive but in the extreme cold the lead acid batteries lose much of their capacity getting as low as 40 during the coldest months in Kotzebue In order to determine the battery capacity of the Deka 8G31 at any point during the year the low temperature for each day is used as a temperature reference The results are seen in Figure 62 The capacity dips into the low 4096 range throughout the winter months The figure shows the worst case scenario since the temperature reference used is the low for each day The actual batteries will be in an insulated enclosure Effect of Temperature on Battery Capacity in Kotzebue AK 100 90 80 70 60 50 40 30 20 10 0 o d d od d od ELE EE O S O S S SU XS uS Qu qu uS WS uS KF KF qd v V WAV AV S WAV AV EAN AEN XU AW XV Soq S Wy qq QU A Q2 QoS WN oD Capacity Day Figure 62 Effect of temperature on Battery Capacity in Kotzebue AK Extensive solar array testing has not been performed around Kotzebue AK where the system will be deployed In order to accurately predict how a solar array will perform the solar insolation cloudiness and daylight hours must be taken into
60. alizes variables to be used in various subroutines These variables could be counters or anything which affects a decision in the program The program then continues taking readings and storing these readings into a data table at a predefined interval The researcher will provide the necessary code for recording data To test the system readings of the power supply voltage and the panel temperature were taken to provide real data Marking the time is an important block in the program because this is used to decide when it is time to transmit data Unfortunately the time reading functions of CRBasic are somewhat limited to simply reading the present time However by using the following algorithm it was possible to calculate the time in seconds since the beginning of 2006 Seconds _ Since _ 2006 Year 2006 366 24 60 60 Day _ of _ Year 1 24 60 60 1 Hour 60 60 Minutes 60 Seconds 35 Using equation 1 a calculation was made of the time difference from the last transmission session to the present time The last transmission session is initialized to the present time so if this is the first check then the difference will be 0 However if there is a calculated difference in the last sent time to the present time that is greater than the transmission period the program executes the transmit data function of the system This function is explained in more detail in the next section Declare Initiali
61. amount of incoming solar radiation that reaches the planet measured in Watts per m Figure 7 shows the average insolation values throughout the year measured at the top of the atmosphere for the latitude and longitude of Kotzebue Solar Insolation at Kotzebue AK 600 500 I o eo Watts per M wo e e to e eo 100 re re re I re D D D D 9 D re 3 3 3 3 s 3 e 3 s 3 o o o o o o o o o o o o a a a a a a a a a qi SI a A N A N N N A A N A N A ee T ae mu ae t Pond hund me a 3 D N o Ss a Day Figure 7 Insolation in Kotzebue AK The insolation values drop extremely low in the low angle sun of the winter but peak close to 500 W m in the summer roughly half of what is received at the equator Another important weather condition for energy generation is wind speed Figure 8 shows average wind speeds to be around 1 1mph with gusts ranging from 30mph to 48mph In the winter the prevailing wind direction is from the East while in the summer the wind mostly comes from the West Average 56 Calm Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Figure 8 Wind Speeds in Kotzebue AK 2 3 Collecting Data SRI has considerable experience in remote sensing applications Through prior efforts they have concluded that using a single datalogger with multiple inputs as well as integrated memory and processor is the most efficient way to implement data collec
62. and running off the real time clock provided a more versatile solution 6 1 7 Iridium Hello World Purpose This program established a connection through the Iridium satellite network to a computer running hyper terminal the string Hello World was streamed The purpose was to demonstrate communication from the CR1000 datalogger to the Iridium transceiver Set Up Power on CR1000 load program with PC200W software connect datalogger RS232 to Iridium transceiver using a Null Modem Cable Set up a computer to another Iridium Transceiver and run using hyper terminal Solution 68 Figure 56 Iridium Hello World Program To communicate with the Iridium transceiver the datalogger must send out AT Commands To establish a dial up connection to another Iridium transceiver the AT command ATDT number is used and initialized in line 10 of Figure 56 After opening the RS232 port this command is output to the transceiver in line 16 Lines 19 to 22 flush the input buffer and wait for the string CONNECT to verify a connection If this string is not received the SerialIn command will timeout after two minutes Directly following the connection the hello world program is executed and infinitely streams Hello World in lines 24 26 Testing To test the Iridium hello world program the datalogger was connected to an active Iridium transceiver which established a connection with a computer also linked to an Iridium transceiv
63. as possible to see how often a successful transmission was received The purpose of this test was to verify that sending the period worked properly and to observe the reliability of the system Set Up This test required the full system to be set up On the remote end the datalogger was powered by one battery and the Iridium transceiver by a different battery The switching circuit was setup to turn the Iridium on and off as indicated by the datalogger On the local end the data was received by the python program put into the Data Transport Network and then made into a database Testing The datalogger took sensor readings every second and stored those values until they needed to be sent The period to send the data was originally set to approximately two and a half minutes After the first transmission the local side sent a new period to the remote end starting with five minutes and then incrementing after each successful transmission After the datalogger sent the data it waited for three minutes to receive the new period To test the frequency and reliability the test was run for twenty three hours and a log file was kept of all the data to review in the morning Conclusion A table of the data collected and calculated can be seen in Table 6 70 Record connect hang up total bytes period time time time sent 1 13 09 47 13 10 11 0 00 24 1954 2 0 05 01 13 14 48 13 15 08 0 00 20 840 3 0 10 04 13 24
64. as shown that the reliability of a system which transmits all data points daily will easily meet the minimum requirement of at least one data reading per week This was of course the primary goal as defined by Patty Sullivan Testing of the final solution also shows that the system is well within the power budget described in the specifications section of this report Although trivial using a design which toggles power to the Iridium transceiver was an important milestone to meet this goal The Iridium transceiver consumes the most energy in the system even when idle and therefore it was desirable to create a solution which disconnects power to it In addition to meeting the minimum requirements set forth by Patty accomplishments were made in providing more end user control The final solution allows the end user to control the data transmission period of the remote system Allowing this communication gives insurance that if the daily transmission period unexpectedly drains energy resources the user can reset this period Providing the researcher with a user web interface allows the data to be viewed and analyzed in real time Important data such as battery voltages system temperatures and photovoltaic panel readings will give vital information on the system status This information can provide clues for potential causes of problems should the system fail Designing this remote data transmission system in only seven weeks proved to be a chall
65. as temperature wind speed moisture and a host of other environmental indicators Many of the other CSI dataloggers have similar functionality with variations simply being in the size number of inputs or layout Figure 11 A CR1000 used in a weather station 2 3 2 System Description The Wiring Panel uses screw terminals to connect input sensors and controlled devices to the output On the CR1000 there are 8 differential analog inputs 16 single ended analog inputs 8 digital I O ports as well as 5 and 12 volt terminals Additionally there is a 9 pin RS232 port for serial communications A 9 pin CS I O port is also included for connection of other peripherals such as the CS keypad The measurement and control system can sample input sensor voltages at a maximum rate of 100Hz The CR1000 implements both battery backed SRAM and non volatile flash memory to store data and programs Standard memory on the CR1000 is 2MB SRAM with expansions available The CR1000 uses the CR OS 8 operating system which was designed by CSI specifically for data acquisition This operating system can be used to run the datalogger s complete set of process arithmetic and program management instructions used to operate the system 2 3 3 Peripherals To add even more flexibility to this device CSI multiplexers and synchronous devices for measurement SDM can be implemented to expand measurement and control capabilities Multiplexers increase the number of
66. ata length d not mod d s FUNCTION len data SIZEOF FLOAT repr data reading values Distorted Data return None readings ii z EA thus we better have mults of 4 bytes numbers nValues len data SIZEOF_FLOAT for i in range nValues value data i SIZEOF FLOAT i SIZEOF FLOAT SIZEOF FLOAT number struct unpack f value numbers append number logging debug s ts s nNumbers d FUNCTION reading ts len numbers reading values numbers return reading def readUntil self ser x maximum O This function reads characters until it reads the string x mnm FUNCTION sys getframe f code co name logging debug s waiting for s maximum d FUNCTION repr x maximum 102 def def message buffer StringlO StringIO i 0 while True value ser read message_buffer write value text message_buffer getvalue if len text gt 25 xx text 24 else xx text logging debug s recv d s FUNCTION len text repr xx if self isNoCarrier value return None if value x i logging debug s received s of s FUNCTION repr x 0 i 1 J repr x i 1 else i 0 if i gt len x logging debug s received s completely FUNCTION repr x break if maximum 0 and len text maximum logging de
67. ated except at record 8 the period should have increased from 35 to 40 minutes However this did not happen because either the period was not received or a byte got distorted during transmission leading to an invalid period value The positive side to this is that the datalogger realized this problem and kept the period at the same value it previously had In addition all of the records were sent successfully from the remote end to the local end with only 2 records not sending on the first try The period was successfully sent 22 out of 23 times and when it was not sent successfully the datalogger handled the situation by maintaining the current period 6 3 3 NO CARRIER Purpose If the remote side Iridium tries to establish a connection with the local end Iridium transceiver and receives a NO CARRIER it will try to reconnect three times 71 If all three of those reconnection tries are unsuccessful the datalogger will take the data it was trying to send and add it to the next data transmission The purpose of this test was to make sure this feature was working properly Set Up This test was run with the complete system set up The remote end was connected with the switching circuit being controlled by the datalogger and calling the Iridium transceiver The remote end was set up waiting for incoming data from the remote end To start this experiment the remote side s antenna was unplugged to ensure a NO CARRIER for the first data tra
68. bug s received maximum chars _ FUNCTION message message buffer getvalue return message 0 len x isNoCarrier self value This function checks for a NO CARRIER mnm FUNCTION sys getframe f code co name self read buffer write value result self read buffer getvalue find NOCARRIER if result O0 logging debug s found NOCARRIER self read buffer StringlO StringIO return True else return False output self readings This function write the formatted data to a file used during testing mnm 103 __FUNCTION__ sys getframe f code co name filename 3d txt self counter logging debug s opening s for writing FUNCTION filename myfile open filename w myfile write HEADER n myfile write s readings 0 header myfile write n myfile write DATA n for reading in readings what do do if it gets a corrupt piece of data if reading None myfile write Distorted Data else myfile write TS s reading ts myfile write VALUES for value in reading values myfile write e value myfile write n myfile close self counter 1 rf name main setupLogging logging debug s starting name datalogger Datalogger serial port 0 readings datalogger getReadings readings datalogger run if readings l
69. ch goes through the subject and returns the version and the transmit timestamp The transmit timestamp is the time at which the data was pulled from the Data Transport Network Next storeReadings calls splitPayload The function splitPayload breaks the payload back into a header and data and returns those values to storeReadings 52 Once the header is separated from the data it also has to be parsed with the function parseHeader This function separates the header into its components the remote Iridium phone number that is sending the data and the column names for the values stored in the database Both the phone number and the column names are returned to storeReadings The final parsing is of the data All the values that were returned to storeReadings in the previous three functions are the parameters that are passed into the function parseData Here the parameters that were passed into the function are combined with the collected data into a variable named reading This complete reading is then returned and sent into the function store to be added to the database The function store is the last step in creating the database First it opens a new connection to the database and then attaches the reading that was obtained from parseData onto the database Store then calls PlotGenerator which will produce the graphs of the data 5 4 3 2 plotGenerator The first step in generating
70. clock drift would add an energy consumption value of approximately 10mAH this would more than double the energy consumption of the Iridium transceiver Design flexibility was another deciding factor in choosing a transfer method In the future more dataloggers could be implemented easily with a mobile originated approach Using 27 a mobile terminated approach would require the LoggerNet software which is designed to access a single datalogger Although this approach offers a simple solution as described earlier this would give the solution little flexibility For the reasons mentioned above the decision was to use a mobile originated approach Although mobile termination offered a simple solution the mobile originated benefits were too important to ignore as shown in Table 5 28 5 Design Documentation Figure 25 shows a top level block diagram of the full system The following sections outline the design of each subsystem and address how each component is linked together to form a fully functional remote data transmission system REMOTE END PV Panel Charge Charge Controller Controller v Datalogger Communications Iridium Satellite battery Battery gt Network LOCAL END Sensors E Datalogger gt Iridium Transceiver Ly Switching Circuit PC E Website
71. complished 6 1 4 Data Collection Purpose The test program used the internal sensors of the CR1000 to collect and store data into a data table every minute This was created using the ShortCut software Setup Power on CR1000 and load program with PC200W software Solution 63 Figure 52 Data Collection Program The first step in creating a data collection program was to declare any public variables which hold sensor readings to be sent to a data table this is accomplished in lines 5 7 The units for these sensor readings are defined in lines 9 12 In lines 14 18 the data table declaration for Tablel defines which readings to store how often to do this and what data type to use Inside of the main program is a scan instruction which tells the program to execute each instruction under scan until line 31 or NextScan The NextScan instruction then tells the program to sleep for a period of time until which the program will loop back to the scan instruction Lines 24 28 are instructions which call the sensors to store values to the variables These values are then transferred to the data table in line 30 Testing To test the functionality of the data collection program the CR1000 was powered on and allowed to run for 5 minutes The CR1000 keypad was then used to view the contents of tablel to verify that reasonable data was being tabulated Conclusion Although minor this program was at least is similar to how the data wa
72. crocontroller to accommodate the various existing dataloggers This way if a researcher desired to transmit data from a CR1000 the datalogger would be connected to a microcontroller the user would then set a switch either physical or in software to execute a routine for CR1000 data transmissions Building off this extension even further could lead to a third party microcontroller accommodating data transmission from non datalogger sources capable of RS232 communications Some data sources might have the resources to format the data while others may not but the ability to interface a wide variety of devices with an Iridium modem could prove to be advantageous Two examples of possible data sources are a GPS receiver and a webcam real time access to both of these are desirable to researchers 7 2 Dial up and SBD communications The system created for this project uses the Iridium Network s dial up mode of operation to transmit data A potentially useful extension to this project is the ability to transmit using either dial up or short burst data If for some reason the remote system needed to go into an extremely low power mode and only transmit a message indicating that the system is functioning this could be accomplished while minimizing the communications budget by using short burst data As explained earlier for shorter messages SBD is more economical than a dial up connection To activate this feature the local end could send a tag at the
73. develop the final code to be deployed in the field Please refer to the help file s included with the following development tools for further assistance 1 1 1 CRBasic Editor The final CR1000 code was compiled and debugged using the CRBasic editor which is available through Campbell Scientific s Loggernet package The software can be purchased through the Campbell download link page as provided in the links and resources section of this manual 1 1 2 PC200W PC200W is a free software package used to upload programs or change the system clock of the CR1000 This software was used to upload new versions of the code but can also be used to directly retrieve collected data See the links and resources section for download information 2 Customizing Code This section will explain the code alterations which will need to be made before deploying the final system Additionally this section will also explain some features in the code which can be changed if desired 2 1 Header Information It is necessary for the header information to be edited inside of the CRBasic code The header is transmitted before the data and provides the backend system with column information for each data reading Figure 73 shows the code to transmit the header information 123 Figure 73 Header Information To customize the code so that the header columns contain the correct information simply edit the SerialOut command shown in line 11 The second parameter
74. ducis i RE ete irae ni pras aUbi rud 9 Iosue 12 Screen shotOf SC WEB od eodieieieici Sero Riu E Gan bIM dnd ica UI aN qu Sin Eq Midas 11 Figure 13 Indium Satellites cover the Earth sq ssscsc censsedenoueccesessedeciestesdeccesssnsetpuesdviasesgoceds 12 Figure 14 Ericus CO VERA eussi essiens inie ainai dab mM alti eE iiaa 13 Figure 15 Message Queue at Local Iridium Connection ses soussscorcerasuaiascaaveeden menteevancercenss 14 Figure 16 Publishing and Subscribing to Data from Central Newsgroup Server 14 Figure 17 Temperature Effect on Battery Capacitance esesseeseseeesesreesrreresreserreresreese 16 Figure 18 Estimated Solar Power Output essseessesrsesessessrssresresserererreesetsrireresressesseesresee 17 Figure 19 Three Stages of Charging L A Battery assiettes a tette 18 Figure 20 Physical setup o the S ySlelll ieu ieiunii rd itn levlaecnseeegnectecatadeeaiwicd 20 Figure 21 Basic architecture of the Iridium SBD iau iei eret inet e bus pne utar speeds 23 Figure 22 Costs of SBD y Dial peus Spass tena limen ad ad binmaciqi iin ies 24 Fig re 23 ISU to PSTN OVBEUIBQ ee cna iced inten telecine cuneis rei deduc ai uL Utt Eua 25 Figure NIE URBE cun E S 25 Figure 25 SUbsyStEMS setaroi kiaina ia e n ieaie 29 Figure 26 Power System Block Diagram uis occisione pnbterd etu eet dpt Der putent 29 Figure 27 Charging Two Batteries with Single PV 12 Glisten nied 31 Ligure 28 CSI San le swi
75. e I O ports and gather more data Since the goal of this project is to interface the datalogger with the Iridium network the data collection program will be created by another VPR partner who will implement the final system However Short Cut will still be an integral component of the remote sensing system 10 4d SCWIN CR10X C ASCWinYoulback SCW Scan Interval 10 0000 seconds Sensors Pile E3 Fle Edt SeRrgt Hep Available Sensors Selected Sensors Borometnic Pressure z E Precipitation J censore o mem Relative Humidity amp Temperature B 10162 Enclosure Relative Humidity 207 Temperature amp Relative Humidi C5205 Fuel Temperature Sensor CSS00 Temperature amp Relative Hun CS505 Fuel Moisture Sensor B o144 Wind Speed Sensor 9244 Wind Direction Sensor A 03001 Wind Speed amp Direction Sen B 03101 Wind Speed Sensor 03301 Wind Direction Sensor Vaisala HMP4SC Temperature and Relative Humidity Sensor Units for Aie Temperature Deg C Deg FK Sensors Z Weing Diagram Ag Winng Text l 1 Figure 12 Screen shot of SCWin 2 3 6 CRBasic The SCWin software allows users to create programs into CR1 files written with CRBasic with simple drop down menus as described above The CR1000 uses this programming language which is similar structured language to Basic Using this relatively high level language a programmer can easily create programs with the
76. e e aE EEES aia 33 5 2 6 Assemb Yosser een are a aE Ran a S QM 34 5 0 CRIO00 Codeseira E ees EE ETA E EA A a TEREG 34 5 3 2 Top Level System m 34 5 3 3 Transmit br mr m 36 5 3 4 Establish Connection ossia d eile Doe dede ou ind aeniea enis 37 5 3 5 Data QUt 39 5 3 6 Accept New Transmit PMO aj saciiasaudcasaesaavnintaaantsdeaadeaavendeanisiaustiveearsaatins 42 54 Data turis nd OG c 43 5 4 2 Becermg tbe Data eto tho iun euet e Ia a EEKE eai ME 44 5 4 3 Prepare Data for Viewing sedo scrisat sa tio te apad Mind ancoldiu di PI ina aee 49 5 4 4 Dio 54 6 Les ngand Besults emot hanenn Ee E umet studui MO dE ESE 59 6 1 CR1000 Code MI T kE aae iai 59 6 1 2 Hello World m 59 6 1 3 Seral I acca dan eiii p a 60 6 1 4 TD eile BU i oui etait iri e EEEa 62 6 1 5 Send Vala sje a iei 63 6 1 6 Control Port TIME uide ee deed isa bie baie dun cis ndegayadetaacasiienadecess 65 6 1 7 Itidiam Hello World Pm 67 6 2 Python Cage Vests esisiini ris cece ata ESEG 68 6 2 2 C omp ter to Comp tet eee ee mesi eiiie ame ene etn Mcd eer Esenas 68 6 3 Pull System TESIS ee E E E PU n HE ERU as 69 6 3 2 Reliability amp Fregen y edebant edi actas pre abiens edunt 69 6 3 3 NO CARRIER c P 70 6 3 4 Data Tra
77. e string x again Once M 49 becomes equal to the number of characters in x the string has been successfully read into the modem and it returns the data buffer Once the call is connected retrieveData is called This function reads and formats the data in and returns it to receive A basic overview of the function can be seen in Figure 41 readUntil Collect all the data Split data on an end of row tag parseReading Seperates each row into a Timestamp and values Figure 41 Overview of the retrieveData function The program first calls the readUntil function until it gets an end data tag Each new row of the data is separated by a parsing tag so the data is split at each of these values and stored as separate elements in an array The last step in sorting is parseAsciiReading Each data reading contains a timestamp followed by data values so the first step in parseAsciiReading is to separate the timestamp from the values and save them accordingly Each value is then casted from a string to a floating point number and appended together as an array of values The timestamp and array of values are then returned as the final data to be posted to the data transport network 5 4 3 Preparing Data for Viewing The second task of the local end is to take the data that is stored in the Data Transport Network and put it into a user friendly format This is done in two ways pu
78. ed It should be designed to not drop lower than 30 of its capacity to increase the lifespan of the battery The temperature the battery is operating at greatly affects the capacity of a battery A chart showing the relationship between temperature and capacitance can be seen in Figure 17 As temperature decreases so does the capacity The Ah rating of a battery is given for 80 F When the battery is operated at 40 F the actual Ah of the battery is 75 of the rating and at 0 F the actual Ah is at 50 of its rating 15 Temperature Effect on Capacitance 100 90 80 70 60 50 40 30 20 Capacitance 10 0 Qc Q PF sb c wownonrrsrrhre OM OHO DN YO WO NN NN Eoo O iycs urs o NN AN Temp Figure 17 Temperature Effect on Battery Capacitance The SRI sponsored MQP team from 2005 Communications Network for a GPS Atmospheric Imaging System performed a great deal of research and experimentation on batteries in Arctic climates Their work showed two specialized deep cycle lead acid batteries to be most applicable the Gel Cell battery and the Absorbed Glass Mat AGM battery The gel cell battery uses a thickening gel usually fumed silica to immobilize the electrolyte making the battery able to perform even if the walls are cracked or damaged Also the battery functions under any orientation unlike flooded lead acid batteries which need to be sitting flat on the ground AGM batteries work much
79. ello World Program was the first test in programming the datalogger Using the serial I O instructions the goal was to create a program which could send the string Hello World out a serial port Setup Power on CR1000 and load program with PC200W software Connect datalogger over an RS232 serial cable to a computer running hyper terminal Solution Figure 49 Hello World Program The Hello World program defines a string in line 5 called HelloWorld which has a maximum length of 11 characters In lines 8 9 the main program begins by initializing the Hello World string The program then enters an infinite while loop in lines 10 13 which opens the RS232 port and continuously streams out the HelloWorld String 60 Testing To test the Hello World program the datalogger was directly connected to a PC running hyper terminal Upon completion the program was able to continuously stream Hello World to hyper terminal as shown below d HyperTerminal File Edit Yiew Call Transfer Help Dua 3 SERE Hello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHel lo MorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello Wor IdHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldH ello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHell o MorldHello WorldHello Wo
80. en no satellite signal could be found Unplugging the antenna simulated the loss of signal The second attempt was able to connect briefly but the call was dropped then on the third attempt a successful transmission was made This current profile illustrates one of the worst case scenarios as far as erngy consumption is concerned The average current drain is about 156 mA for 304 seconds These figures are used to estimate the overall energy consumption of the Iridium 6 4 3 Datalogger Energy Consumption The power consumption of the CR1000 Datalogger was determined using the same MultiTec 330 Digital Multi meter to measuring the current drawn from the datalogger battery during operation There are two modes of operation for which the current was measured the measurement mode and transmitting mode Measurement mode is the time when the datalogger takes measurements of sensors For this project the datalogger measures once every hour and is in sleep mode the rest of the time To test the current drawn during measurements a test program was written to read 80 sensors every minute The results are shown in Figure 60 TT Datalogger Current Drawn During Measurements Measurement i taken ee Current mA S TQ HOSP PHP GY ECD HPS OVWeCDP ADAG S Time sec Figure 60 Current Drawn when taking Measurements The average quiescent current between measurements is around 0 53mA while the current during measurements
81. enge however lessons were learned in handling such a task One of the most important lessons learned through the completion of this MQP is the importance of thoroughly testing the final solution Even when things appear to be completed unexpected results can occur if one does not take the time to meticulously test and professionally document a prototype Another lesson gained from this experience is the importance of breaking a large system into smaller subsystems When developing a large scale system such as this it is much easier to accomplish if it is approached one small step at a time This also proved to be true in debugging It is far simpler to solve an issue when the problem is isolated to a smaller component of a much larger system then to search through an overwhelmingly large and complicated structure The lessons learned from this experience will no doubt be applied throughout the professional careers of this project team 85 Appendix A CR1000 Code The following appendix gives the full version of the CR1000 code written in CRBasic developed by the WPI team This code will be edited by Dr Patrick Sullivan and used in the final deployment of the data acquisition system Remote Data Transmission System cr1 WPI Team 2 24 06 Eric Hall Peter Kaineg Amanda Quigley Eric Young Declare Variables and Units Default data type Float System consts Public MAX CONNECT TRIES Public REMOTE IR PHONE NUMBER AS ST
82. enland Iceland Norway Russia and the Arctic Ocean Throughout these regions over 100 grants and 500 scientists are supported year round for a total of 55 different field locations SRI s role in VPR is to provide communications services for VPR s field research projects including the deployment of data collection and field communications systems SRI and VPR have received increasing requests from scientists for real time access to data from their research stations in remote locations These are often small stations collecting data from a few sensors and storing samples into a datalogger In the past the data was retrieved only when the scientist visited the site which could be as infrequent as once a year Several scientists explained that their system was working great while they were there but upon their return a year later they found that it failed shortly after leaving With real time access to the data they not only have constant monitoring capabilities but they can also determine if the system is functioning properly This report provides design documentation as well as user and maintenance manuals for future use and upkeep of the system The major specifications for this project are the system must be able to sustain itself year round without maintenance in 40 C with ice and snow the system must collect and send meteorological data at least once per week over the Iridium satellite network and the system communications costs mus
83. er This is a simple function that calls readUntil and records all the data up until a specified end header tag The header is then returned to receive Figure 40 shows the architecture of the readUntil function 48 Wait for an incoming lt character Y Store character in a buffer Is NO CARRIER found Return None Ye No Is the character x M Is M gt the length of x Yes Y Return the character buffer Figure 40 Overview of the isNoCarrier function The program waits for an input character to be received and when it receives the character it adds it to a buffer that will contain the data Next it calls a small program isNoCarrier which checks the buffer to see if a NO CARRIER was read A NO CARRIER signifies the call is over it has either been hung up or the connection was dropped If a NO CARRIER is found the function immediately returns None If it is not found it proceeds through the function Next the program checks if the character is equal to the Mth element in x where M is a counter that starts at zero and increments if it finds the first character of the string it is looking for and x is the string that is being searched for If it has found the character the index M will be increased to look for the next character if not it will set M back to zero and look for th
84. er The computer also had an active window of hyper terminal to view the data being streamed After receiving the CONNECT 9600 string on the local computer hyper terminal showed a similar output as shown in Figure 50 Conclusion Using the hello world program through the Iridium transceiver was a major milestone for the team This program demonstrated the datalogger s capabilities in interfacing with an Iridium transceiver Making a direct connection between two Iridium transceivers effectively built a wireless RS232 cable which was the method of streaming real data 6 2 Python Code Tests This section of the testing results briefly document the methods and results from testing the python code which is used on the backend system 6 2 2 Computer to Computer Purpose Before introducing the datalogger and Iridium transceiver into the testing The Datalogger py code see Appendix B was tested to make sure the code worked as it was intended to Set Up Two computers were connected with a null modem cable One computer ran the python program and the other had HyperTerminal open Testing The computer with HyperTerminal simulated different situations that the datalogger would produce First a successful transmission was simulated by entering 69 values into HyperTerminal A successful transmission consists of CONNECT lt BAUDRATE gt a start header tag some header text an end header tag data readings and
85. es have observed negative growth trends in the late 20 century among arctic and alpine tree lines studied The most significant tree line regressions have been noted in particularly dry areas Clearly there are changes occurring in the earth s climate studying areas affected by these changes can provide insight into the broader implications of phenomena such as global warming 2 1 4 Existing Systems for Real Time Data Transmission As previously stated recent years have seen an increase in researchers need for real time status reports from their remote systems As a result several organizations have begun using satellite network technology to transmit status reports instrument data and even digital images This section will briefly overview a few such deployments One notable deployment took place in September of 2005 Since April of 2002 The National Oceanic amp Atmospheric Administration in conjunction with the Pacific Marine Environmental Laboratory NOAA PMEL has been deploying web cams to view the North Pole in the summer warmth and daylight They are set up from April to October and redeployed each spring The images from the cameras track North Pole snow cover weather conditions as well as the status of PMELS North Pole instrumentation This includes meteorological and ice sensors seen in Figure 1 Among the sensors are downward looking sounders ice thickness poles and camera images which are relayed via the Iridium satellite system
86. essor 100 100 70 100 Table 4 External Microcontroller vs CR1000 Internal Microcontroller Table 4 clearly shows that for this application the CR1000 internal processor without a third party microcontroller is the best choice After working with the CR1000 to test its 26 capabilities this decision was reaffirmed and the team proceeded without a third party controller to interface the datalogger with the Iridium transceiver 4 4 Data Transfer To transfer data from the datalogger to the local user it was important to choose a method of initiating a connection The options were to use either a mobile originated approach where the remote system dials to the local user and data is pushed across the network or a mobile terminated system where the local user initiates the call and pulls the data S Complexity Eeg Desin Flexibility Weight Mobile Originated Mobile Terminated Table 5 Data Transfer Mobile Originated vs Mobile Terminated Weighted Chart Table 5 shows a value analysis of the pros and cons of each approach The system was designed under the mobile originated approach Mobile Originated implies that the system initiates a dial up connection from the remote end The local host accepts communications and manages the data being pushed from the remote end The computation required with each approach is significant to the design decision Using a mobile terminated approach with Campbell Scientific s Logger
87. ext transmission If a valid number is received then the transmission period is changed to this value A valid value is defined as divisible by one hour and less than once every 2 weeks If a number is received that is outside of this range then the program will reject this change and the previous transmission period is continued being used Following these instructions the subroutine then exits back to the call origin Figure 35 shows a flowchart of this sub system 43 Initialize Local Variables Wait For PEND Put Period Information Into New Period Container New Period Valid Period New No Figure 35 New Transmit Period Flowchart 5 4 Data Management Code The data management code is located at SRI It is responsible for receiving the data and then posting to a website for the scientist to view There are two main components in the data management code The first is the Python program used to receive the incoming data from the satellite network organize it and post it to the Data Transport Network 44 The second component is the Python program that takes the data from the Data Transport Network and archives it and provides graphs Data gt Receive Transport Store Database Network Returns Data Datalogger plotGenerator Graphs Figure 36 Basic Overview of the End System Code Figure 36 shows a basic overview of the entire end system code T
88. for every decrease of 1 C P Other aspects of the climate reduce the effectiveness of the solar power however The effectiveness of a solar cell is optimized in full sun hours or times of the day when the sun s intensity is equal to 1000 watts per square meter Most of the time full sun hours are about a quarter of the total sunlight hours in a day In Kotzebue AK total sunlight hours in a day can get as low as 1 5 hours This means there is virtually no sunlight close to the winter solstice Another obstacle in solar power generation is cloud cover Figure 6 in Section 2 2 shows that Kotzebue experiences more cloudy days than sunny days As a rule of thumb the output of a solar panel during cloudiness is only 20 of the output in full sun Precipitation can severely limit the output of a solar panel as well Snow can accumulate on top of the solar panels during the winter months greatly decreasing the efficiency of the panel Without any human interaction with the system snow and other debris can stay on the panels for long periods of time inhibiting any power generation By taking into account the amount of solar insolation sunlight hours and cloudiness the estimated output of a 20W Solar Panel in Kotzebue AK for each day of the year can be seen in Figure 18 This estimation does not take into account any obstruction blocking the sun from the PV panel such as ice or snow Estimated Power Output of PV Panel lo Lo LO lo LO LO Lo
89. g StreamHandle console setFormatter formatter console setLevel logging DEBUG rootLogger logging getLogger rootLogger setLevel logging DEBUG rootLogger addHandler console rootLogger addHandler log file 115 System Requirements These programs were written with Python 2 4 2 A downloadable version and overview of this version can be found at http www python org 2 4 2 Also to utilize the serial commands pyserial is needed to properly run these programs Pyserial 2 2 can be downloaded through http pyserial sourceforge net This site also contains information about Pyserial and its functions Lastly if this is being run on a windows PC win32 is required This can be downloaded with pyserial also at http pyserial sourceforge net For this download the file named pyserial 2 2 win32 exe 116 Appendix C User s Manual This manual is intended to aid the researcher using this system in properly setting up the data collection station Additionally it documents the programming tools used to create the interface between a Campbell Scientific datalogger and an Iridium modem Physical Setup The following sections will cover recommended steps to ensure proper set up of the physical system on site in Kotzebue Alaska Figure 66 is a physical diagram of the research station PV Panel 54m L MET at 2m m MET at2m Guy wires at 15 m 4 Guy wires at 1 5 m Insulated enclosure
90. g data the costs associated with it allowed us to quickly discount it as a viable communications option From here attention was to dial up service 4 2 2 2 Dial Up Service There are two main forms of Iridium dial up service Iridium Subscriber Unit ISU to the Public Service Telephone Network PSTN and ISU to ISU The ISU to PSTN provides connectivity from the onsite modem to an offsite computer LAN or Internet Service Provider ISP A basic overview of the dial up data method can be seen in Figure 23 24 tte eT HTTMIIDLLLLLLZ Host Application Application using the LBT Figure 23 ISU to PSTN overview A call request from a mobile ISU is routed over the satellite network to an Iridium Gateway for user authentication and call set up The switch then makes the connection to the number dialed into the ISU The analog modem in the gateway and the analog modem in the host application then synchronize and then the end to end connection is established Overall set up time for an ISU to PS TN is estimated to be 40 seconds The second data service is ISU to ISU which provides a connection between two Iridium units A basic overview of this method can be seen in Figure 24 ISU 2 LLL INter Working Deco Equipment Figure 24 ISU to ISU overview ISU 1 dials the ISU 2 The call is set up and connected to inter working equipment at the gateway A ring alert and call set up is then issued to ISU 2 by the gateway ISU 2
91. g warning s no data available FUNCTION return None measurements 0 PTemp 1 Batt volt 2 Batt Volt IR 3 PV Voltage 4 ETemp units 0 Celsius 1 Volts 2 Volts 3 Volts 4 Celsius ranges 0 None 1 0 15 2 0 15 3 0 30 4 None colum numbers 0 1 2 3 4 for column_number in column_numbers measurement measurement column number unit units column number 113 49 9 png filename s png measurement png path os path join png dir png filename line self getLine readings measurmenr unit range column number self plotLine date range line png path def getReadings self dr Returns a list of readings within the date range 44 FI IT FUNCTION 2 sys getframe f code co name Here is the structure of the series xs Series time time ys Series measurement units ys setRange range for reading in readings x matplotlib pylab date2num reading getTimestamp y reading getValues column number xs append x ys append y l Line xs ys return 1 def plotline self date_range line png_path __FUNCTION__ sys getframe f code co name logging debug s s png path s FUNCTION date range png path figure matplotlib pylab Figure figsize 8 2 25 f Canvas for figure canvas FigureCanvas figure
92. gger This deployment could certainly further benefit from having a digital image of the site transmitted periodically with the collected data VECO Polar Resources has deployed webcams for other research stations They have even used the Iridium satellite network to transmit the images so implementing a camera is a possibility in the future However adding a webcam to the system may or may not require an external microcontroller to perform buffering or protocol adjustments between the camera and the satellite modem 7 5 ISU to PSTN Communications In the early stages of the design choices a solution using ISU to ISU communications was chosen As described in the design choices section some of the most important reasons for using ISU to ISU were the simplicity and modem training advantages Since SRI International had the resources to provide an additional transceiver without the added cost it made the most sense to take advantage of these benefits for Dr Sullivan s project However the solution of ISU to ISU might not be the most cost efficient solution to future projects If an additional Iridium transceiver is required it might make more sense to use ISU to PSTN communications because of the fact that the communications price is the same without the added transceiver 84 8 Conclusion After successfully implementing a remote data transmission system the preliminary goals and specifications set forth for this project were met Testing h
93. gy used for Transmission Wh Once per Week One Sample 13 00 Once per Day One Sample 91 24 Once per Day Complete Data Set 113 91 Table 3 Energy consumed per year Table 3 shows the energy in Watt hours Wh that each method consumes per year The communication system will be powered by a 12V 110 Ah battery which when fully charged will provide 1320 Wh per year The battery should only be drained to approximately 30 of its original charge capacity which leaves 924 Wh In addition the battery self discharges 2 per month when not used At worst case there would be twelve months that it is not used which leaves 725 Wh to run the system for a year As with the price budget all three methods stay within the energy specifications Table 3 was used to determine the weightings for the energy parameter As before with the cost weightings a 100 was given to the best option Then dividing one sample per day by one sample per week showed a factor of seven Dividing one hundred by seven gave a rating of fourteen The same was done with sending a full day s data everyday and the result was eleven The last category considered among the three choices was the convenience for the scientist At a minimum the system must send one data sample once per week Ideally the system would send the full set of data each day If all the data is sent daily the scientist can begin to analyze it and would not have to wait a year to retrieve the da
94. he codes used to receive and sort the data from the Iridium Satellite Network and post it to the Data Transport Network are the Python codes Receive and Datalogger Store and plotGenerator take the data that was posted to the Data Transport Network and create a database as well as graphs of the data 5 4 2 Receiving the Data Two python codes were written to receive the data coming across the Iridium Satellite Network Receive and Datalogger The full text of these codes can be seen in Appendix B Receive is an infinite loop that calls Datalogger Datalogger receives and sorts the incoming data and returns it to Receive Receive then posts this data to the Data Transport Network 5 4 2 1 Receive Receive calls Datalogger Section 5 4 2 2 which will collect and organize the data It then sends this information to the Data Transport Network where other Python programs will take that information and produce a database and graphs of the data Figure 37 shows the overview of the Receive program As shown the program is an infinite loop So the program will call Datalogger to read and return the data post the data to the Data Transport Network and then start again and wait for the next set of data to be sent across the Iridium Satellite Network 45 5 4 2 2 Datalogger Datalogger waits for incoming data stores the data sorts it and returns it to Receive A basic overview of Datalogger can be seen in Figure 38 Call Datalogge r py
95. he design can be seen below in Figure 27 Using this diode isolation configuration the datalogger and communications system can run on separate batteries The 6A2 6A 200V diode was selected for the job for its high current rating and voltage rating as well as low forward voltage drop 30 PV Array m Indium Battery Solar Battery DL Battery Figure 27 Charging Two Batteries with Single PV 5 2 Switching Circuit The design of a switching circuit to turn the Iridium modem on and off was critical to the system s functionality A relatively simple circuit was implemented to perform this task One of the digital control I O ports from the datalogger was used to activate and de activate the circuit and a 12V battery was used for power The following sections detail the selection of parts and reasoning behind the design decisions 5 2 2 High level Design The CR1000 datalogger manual provided a section describing a sample switching circuit which is shown in Figure 28 This figure shows a simple circuit often used for switching external power to a device without the use of a relay which typically would draw more power than using a transistor 31 SUPPLY 40 VDC MAXIMUM 10K MAXIMUM CURRENT TO PERIPHERAL IS 75 mA AT 25 C 40 mA AT 50 C 2N2907A 2K PERIPHERAL TO BE POWERED Figure 28 CSI sample switching circuit For the purpose of powering on the Iridium modem a similar circuit
96. ies 53 Figure 46 Figure 47 Figure 48 Figure 49 Figure 50 Figure 51 Figure 52 Figure 53 Figure 54 Figure 55 Figure 56 Figure 57 Figure 58 Figure 59 Figure 60 Figure 61 Figure 62 Figure 63 Figure 64 Figure 65 Figure 66 Figure 67 Figure 68 Figure 69 Figure 70 Figure 71 Figure 72 Figure 73 Figure 74 Figure 75 Figure 76 Website s System Heal Check uoo dioe ccactasoenndescwsetunusuecsseeadeacsebssebivneeesy 56 Website s Query page eee soie dena etd nitas a adusta cr A aianei 57 Websites Database VIO osetia datas aici cdad uude uasa uus ela iiae seee enis 58 Hello World Pine team uii inui re Rob psu utto epi eR a aE ENS 59 Hyper Temuinal Bello WO ieehidecscieeu ducti ph eta uieacaientennedtiae 60 Timing control ports PEO SEAN assy eec re sincrone niii ainiin 61 Data Collection PPORPID nope gi niei i a Eom Eaa 63 send Out Data DEOS ESO Guo ces rdi issos isini i lien Digesta ndisse cmd 64 Hyper Terminal CRIODO Data Poihi ee ooco id rer bib rhe tema dab bb ir tios mph deu 65 Timing control ports program aua uedeacs ed nios odi edutop i dieti tdi tasi tinis ddind 66 Indium Hello World Program uacua eu b terc ka artis eost petere 68 Curent Profile Connect First Attempt 555 2 pai Rr ipae Ese ARE b ERE Se va lego DUbS 75 Current Profile Connect Second Attempl nasi tree torn eene 75 Current Profile Connect Third Attempt ssessssesssessssssesssessesnsssesssesresseeess 76 Current
97. ill include ix another L Band Iridium modem and a dedicated PC running a program to receive and format the transmitted data The local end program which was created using the Python programming language will send updates to the Data Transport Network which will load the most current data onto the website The datalogger and the Iridium modem were both specified by and provided by SRI The datalogger was chosen for its high level interface to the processing and data controls The Iridium modem was chosen for its versatility of communications modes including standard dial up and short burst data as well as its ability to communicate in Polar Regions After thorough testing and the completion of the project the system was shipped to Kozebue Alaska where it was deployed A user manual and a maintenance manual were also provided to the researcher leading the project The system successfully demonstrated the functionality of this design through extensive testing The communications system has been thoroughly tested also the ability to adjust the transmission period has been implemented and the data has been displayed on a website for easy access This report documents background information related to the project major design decisions tests performed and recommendations for future work 1 Introduction The National Science Foundation NSF was created in 1950 by Congress to promote the progress of science and national prosperity Today NSF is c
98. ine number 1 logging debug s line d d is d long FUNCTION line number len lines len line mo re match 4 line if mo section name mo group 1 logging debug s now reading section Ss FUNCTION section name if not sections has key section name 108 logging debug s sections else if section name sections if sections has key HEADER header sections HEADER data sections DATA logging debug s exiting new section section name section name append and sections has key line header d lines data FUNCTION DATA od de lines FUNCTION len header len data return header data else logging warning s exiting header None data None FUNCTION return None None def parseHeader self This function parses the header into a phone number amp column names www lines return phone_number def parseData self __FUNCTION__ column_names message_format_version lines phone_number transmit_timestamp column_names __FUNCTION__ sys getframe f code readings line number CO name len lines FUNCTION sys getframe f code co name logging debug s d lines in header _ FUNCTION len lines line number 0 for line in lines line number 1 logging debug s line d
99. lely on energy consumption By not designing an overly complex system potential sources of error can be limited and the system can be made easier to implement Using an external microcontroller would increase external circuitry and create a dual processor system which for this application seems to be excessive Using an external processor would potentially mean more issues for the end user to deal with and possibly absorb time which could have been spent performing tests to ensure system robustness The ability to make future alterations to this system was also a considerable factor While this project does have a specific end user in mind it is worth while to design the system such that it can easily be adapted to achieve other goals This is where using an external processor would be desirable An external processor would allow more complete access to its low level controls The Hitachi processor on the CR1000 can only be controlled through CRBASIC which is converted to machine code so it is a less versatile solution The cost of using an intermediate processor had to be considered but since small processing units can be purchased for less than ten dollars and the total hardware budget is on the order of thousands of dollars cost was not a major deciding factor Energy Consumption Complexity Flexibility Cost Total Weight 0 35 0 3 0 25 0 1 1 0 External Microprocessor 90 60 100 90 83 5 CR1000 internal proc
100. log fil class Datalogger def init self serial por self serial port An Iridium manual sugges f computer and Ir modem TOO LOW MAY CAUSE S asctime s logging FileHandler ERIOD is tranmitted back to the datalogger to set the rate I EXCESSIVE BATTERY DRAIN S name s filename s message s logging format Datalogger log w formatter logging DEBUG r ii sys stdout logging DEBUG Cut EBUG e t202 serial port ts a 19 200 baud rate between However the modem modem link is at levelname s 97 best 2400 baud When cpu modem baud rate is set above 2400 the modem must buffer self baud_rate BAUDRATE self period PERIOD self counter 1 self read buffer StringIO StringI0 self ser Non def _del__ self if self ser and self ser isOpen self ser close def run self __FUNCTION__ sys getframe f code co name logging debug s entering FUNCTION while True logging debug s waiting for a set of readings FUNCTION readings self getReadings if readings None and len readings gt 0 self output readings break makes it a one shot def getReadings self __FUNCTION__ sys getframe f code co name logging debug s entering FUNCTION try self ser serial Serial self serial port self baud rate except serial serialutil SerialException
101. ltage which could reach up to 25V and the Iridium battery voltage 12V the 10 to 1 voltage dividers shown in Figure 69 are used Figure 69 shows how the voltage dividers are inserted into the datalogger s terminals The resistor values are 90k 2 and 10kQ to form a 10 to one voltage divider Figure 69 VDIV10 1 voltage divider for reading greater than 5V Mounting Panel Connections Several hardware components of this system are mounted on a sheet of plywood and cemented onto the top lid of the insulated enclosure Items attached to the mounting panel include the Iridium modem the battery charge controller the switching circuit and diode box a fuse block and several junction terminals The mounting panel is affixed to the inside of the top lid Figure 70 shows the mounting panel with all connections made 120 Charge Controller EE Circuit Box Fuse Block Bre E a 5 CHARGE CONTROLLER 1 pe M 3iDIUM TP SCEIVER S DC 5C CONVERTER M Iridium Modem 4 DC to DC converter Junction Terminals Figure 70 Mounting panel connections Make sure all connections on the mounting panel are secure Gently pull on Iridium DC to DC converter to verify it is fastened securely to the modem as shown Ensure that the circuit box connectors are secure Assure that all terminal connections are screwed down tightly Insure that circuit box lid is screwed down tightly Make sure RS232 cable from the Iridium m
102. m the previously mentioned function getLine The first step of this function is setting the figure size This was set to be 8 wide and 2 25 tall The title is set to be the measurements value for the specified key and y axis label as the units variable The data array is now used to plot the lines onto the graph The y axis limits are set as they were declared in the declaration for the variable ranges The x axis limits are set from the first data point to the current data point The last step is to write out the figure for the website to use 5 4 4 Website A website to display the scientist s data was developed This website contains the database and graphs as described above The website address is http polar sri com datalogger The side bar can be used to navigate the website This can be seen in Figure 45 55 Information Home m Contacts Status m Health Data m Query Updates m Release Notes Figure 45 Website Sidebar The two main pages on this site are the Health page and Query page The health page contains the graphs of the datalogger s panel temperature datalogger s battery voltage communications system s battery voltage solar panel voltage and the temperature in the enclosure The health page can be seen in Figure 46 56 nformati information __ HEALTH TUE FEB 28 14 44 47 2006 m Home es m Contacts Last transrnission was received on Mon Feb 27 16 53 00 2006 Stat
103. mission As seen above VECO Polar Resources has had a fair amount of experience with remote data transmission systems However the Sullivan project will be their first experience interfacing a Campbell Scientific datalogger with an Iridium modem for remote transmission purposes While the collection of soil temperature and moisture data will be of great interest to the scientific community this project will have broader implications as well It will serve as a prototype system for future VPR deployments of CSI dataloggers using the Iridium network to transmit data in real time 2 2 Alaskan Weather The implementation of the data logging system will be in Kotzebue Alaska for treeline research Kotzebue is located on the northwest coast of Alaska just above the Arctic Circle 67 28 N 162 1 W Kotzebue AK Ea at Figure 3 Location of Kotzebue Harsh weather conditions can be expected in Kotzebue The maximum recorded temperature is 85 F in June of 1991 and the minimum recorded temperature is 52 F in February 1968 The average and extreme daily temperatures can be seen in Figure 4 KOTZEBUE 25 N ALASKA 505051 Period of Record 11 1 1982 to 5 31 1995 M ARES Eau Rp E M redi i VI RR vw LISTE Aene UA AED 977 ul ER RR aa Ucar LIV T TE Ean EEN Ea wen pi NOREED ET ENERGES RT ES RS Le o c 2 m t o a z o LT Jan 1 Mar 1 May 1 Jul 1 Sep 1 Nov 1 Dec 31 Feb 1 Apr 1 Jun 1 Rus 1 Oct 1 Dec 1
104. mp C Batt Volt Batt Volt IR PV Voltage ETemp C 0 500 MAX RI Paddy to add his columns here SerialOut ComRS232 PaddyO0 Paddyl PaddyN 0 500 Send The Header With End Tag SerialOut ComRS232 S 0 500 k 1 Do While ts_of_last_sent_record lt Tablel Timestamp 1 k and k lt EC 89 GetRecord ASCII Record Tablel k SerialOut ComRS232 ASCII Record 0 1000 k k 1 loop End of Data Tag SerialOut ComRS232 0 500 ts of last sent record Tablel timestamp 1 1 Exit Sub End Sub HangUp Sub Routine tells the transceiver to disconnect and close the serial port Sub Hangup Initialize AT Command Strings Dim AT COMMANDP AS STRING 40 Dim AT COMMANDhu AS STRING 40 AT COMMANDP AT COMMANDhu ATH chr 13 chr 10 SendOut the AT Command to exit data mode SerialOut ComRS232 AT COMMANDP 0 100 Flush the Buffer SerialFlush ComRS232 Wait for the OK confirmation maximum of 10 seconds SerialIn InString ComRS232 1000 OK 100 Send Out AT command to hang up SerialOut ComRS232 AT COMMANDhu 0 100 Flush the Buffer SerialFlush ComRS232 Wait for the OK confirmation TimeOut after 10 seconds SerialIn InString ComRS232 1000 OK 100 Exit Sub End Sub Sub Transmit Data Dim connect success Configure and turn On Transceiver From Control Port 2 PortsConfig amp B10 1 PortSet 2 True
105. mport Datalogger dbname datalogger user transport class Store ProcessClient NewsPollMixin 106 def __init__ self argv ProcessClient __init__ self argv NewsPollMixin __init__ self callback self process def process self message FUNCTION sys getframe f code co name logging debug s entering FUNCTION subject message get Subject payload message get payload logging info s message subject s FUNCTION subject self storeReadings subject payload to datetime datetime now fm sri ayoung DatetimeUtils parse iso8601 2006 02 24 18 00 00 date range sri ayoung DatetimeUtils DateRange fm to png dir var www polar polar static images datalogger plot generator sri Datalogger PlotGenerator PlotGenerator plot generator generate date range png dir logging debug s exiting FUNCTION def storeReadings self subject payload FUNCTION sys getframe f code co name logging debug s entering FUNCTION message format version transmit timestamp self parseSubject subject if message format version 0 1 and message format version 0 2 logging error s unknown message format version NTESNTT FUNCTION message format version return header data self splitPayload payload if header Non logging error s header is None FUNCTION return phone numbe
106. n 13 2 4 4 Python Programming Language Lao ias iese ebd tun tpe ii adeps Deasesidisan 14 23 POWE SyS Mien Gebot ee eee ner Beene ere fendi nd dan pP ED Den FO Dd I UEM 15 2 5 2 fil MT T 15 2 5 3 Power Generatio Mses n eE nte tps vatnF ag ku pn UM SEU EEE MA 16 2 5 4 Charge 4 Onttollgr dde ipie qupd t RN em dpud uus 18 Dv GSBEBDEIOCOBDBS c osten aa tart qup mM pO Cu Up ea I GIRO UEE 19 341 Communication SpeelTICabloBs ooo secet so raro iesise innii beide annees 19 C MEM V der celi ii Pm 19 3 3 Physical Specifications assostaire eanais 20 X4 oon CHOICES RP 21 4 1 Quantity of Data amp Frequency of Data Transmission eese 21 4 2 Sbort B rst Data vs Dial Up cedimus nonono uan ide Etiese 23 A221 Short Burst Data ServiCe ioi rran r RRS 23 LAC MEE ICI EA iis Pm 24 4 3 gis jj C M 26 4d Data Transfer n 27 De Design DOCubeld ID ss cde uiaennteui td xanh se ip tas car ende naui vid D d Eaei 29 5 1 POWSE yy SU ossi crind ro Unde pb ose es ms derbRDctisdisttan EP eee MEME MUS 20 3 1 2 zio MT Q 30 5 1 3 Solar Pane lerria e E EEN EE E E E 30 5 1 4 Charge Controlle seisin A a EE ERS 30 5 2 UC TN Cirit CE 31 5 2 2 High l vel Design CL ei eri a a 31 ii 5 23 Circuit eT MOM MEE C 33 5 2 4 Part Selection MEME ccc 33 5 23 Control DOfEc aute ere een eer ennei EEEa s r
107. nge with each measurement If the range is None the plotter will automatically choose a range based on the data points For the voltages the range was selected so that the OV reference point would always be visible Therefore the first number in the range represents the lower boundary and the second number the upper Each key corresponds to a column of the data table The program uses the key to go through all five sensor readings For each key the functions getLine and plotLine are called The function getLine uses that key s measurement unit range and data readings A new series is created for the x and y axis The series includes a measurement unit range and data When the new series x and y are created the measurement and unit for x are both set to time the range is set to None and the data variable is initialized as an empty array The y axis is next initialized with its measurement being set to the measurement value matching its key in the variable declaration shown above The unit and range are set the same way Its data variable is also initialized as an empty array Once those variables have been set the data must be set For each data point an x coordinate and y coordinate are drawn from the database and then appended onto the empty data array described above This obtains each x and y coordinate and returns these values to be graphed The function plotLine graphs the data that is returned fro
108. nsfer Bate utem Hebe ui Datus Meet S 71 6 3 5 Pull System INoldbl by asiento editione iiaia 73 6 4 up o C M 74 6 4 2 Communications Energy ConsumiptLiot 2 poni rrt tmr Rete etires 74 6 4 3 Datalogger Energy ColSuIpblOfi au uscaa saepe cipit uberes ente dina Re tpe epica 76 6 4 4 Battery State of Charge Expectation iioi perte tass Due SU Urb ubsot tau sonis 78 7 Recommendations for Future WOEE aues nod do aa tar eeu dei rnt nude aedis 82 7 1 Universal Datalogger Communications sse iad ope proderit in Unna eo ceo dida 82 17 2 Dial up and SBD comtibicatlofs ai eei p teo ates be teo eb tesas ties 82 da Wiebke Terminated Cal IE serere eti ao a b desino adii rion qute da abemus 83 JS Camera for Datalo getse ienesis aires gu s scd eii eih 83 1 5 ISU to PSTN COIUIDICaIODS iscissi ee riot dat ole laetus iiai ie 83 ME SOT E E eer 84 Appendix AO OOD Gode T 85 Appendix B Data Managment Code i coiscssia sea cehizseceunsaueccniasadeuadssdaduniysedeentaetabee nde dues 92 UCR MOT 92 Datalogger Mer HM 95 pop d 103 jr 105 iii pies cuj eR i 110 System PRG CUNT BTS sade cause ane ard edes TU edad eia EE ri R Aaria 115 Appendix C Users Manual L2 tine cossis eiie iei aiii ai 116 Physical SetUp mD
109. nsmission attempts Testing To test this system the datalogger collected data every 20 seconds for a 10 minute period After 10 minutes the remote side attempted to send 30 records across the Iridium Network Since the antenna was unplugged the data was not able to transmit successfully The program then collected data every 20 seconds for 10 more minutes While the datalogger was collecting data the antenna was reconnected After the ten minutes the remote side attempted to send the data again Conclusion This system collected 30 records for the first 10 minutes and was unable to send it due to the fact that the antenna was not connected to the Iridium transceiver The datalogger then collected more records for the next 10 minutes this time the antenna was connected and the data was transmitted successfully On the local end it was confirmed that all 60 samples were sent This proved that when the first transmission was unsuccessful the data was attached to the next record and sent along with the next transmission 6 3 4 Data Transfer Rate Purpose The power and cost analysis was originally calculated using the data rate on the Iridium Specification sheet of 2400bps However to accurately calculate the power and cost a test was run to see how long it took to send a known amount of data across the Iridium Network This test was run by sending approximately 12 kilobytes of data for every transfer This is the amount of data that the scienti
110. ntil the end of file marker parse it and return a list of readings nm FUNCTION sys getframe f code co name logging debug s entering FUNCTION Read from the serial port data self readUntil ser END FILE if data None return None reading elements string split data ASCII READING SEP 1 logging debug s d reading elements FUNCTION len reading elements readings i 0 for reading_element in reading_elements i 1 reading self parseAsciiReading reading_element reading setPhoneNumber phone_number reading setColumnNames col_names if reading None logging warning s None reading on d of d element 100 FUNCTION i len reading elements else The readings are stored in a stack on the datalogger and are transmitted in reverse time order We want to post them in time order readings insert 0 reading ol p logging debug exit receiveData read d elements return readings def parseAsciiReading self reading element Parse an ASCII line of data and return a reading mnm FUNCTION sys getframe f code co name logging debug s parsing d bytes FUNCTION len reading element if len reading element logging warning s zero length reading element FUNCTION return None ts str data str string split reading element 1 ts str string strip t
111. odem is screwed securely to the DB9 RS232 cable coming from the datalogger box 121 DATA LOGGER ENCLOSURE JHR BATT S MRBATT z DL BATT DL BATT PV PV j JADE PWR D PWR Circuit Box T107 Temp Sensor Iridium Transceiver Mounting Panel Figure 71 Mounting Panel Schematic PV panel and Antenna The PV panel should be mounted vertically at the top of a 2 5 meter pole facing south Refer to Figure 66 The voltage leads from the panel will be fed into the insulated enclosure The antenna should be mounted at 1 5 meters up the pole and the cable should be fed into the insulated enclosure and connected to the Iridium modem Batteries At the heart of the power system are two 12V 100Ah gel cell batteries which are housed at the base of the tower in an insulated box 122 Figure 72 Deka 8G31 12V 100Ah gel cell battery CR1000 Code User Manual 1 Devlopment Tools This section will document the tools used to
112. ogging debug s got d readings name len else logging debug s None readings name sys exit 0 Receive usr bin env python import import import import import import logging sys string datetime StringlO Pg from Transport import ProcessClient from Transport import NewsPostMixin readings 104 import Reading import Datalogger Initial MESSAGE FORMAT VERSION 0 1 mnm This version receives the list of readings from Datalogger in a reverse order The initial version received them in reverse time order decending Now Datalogger s getReadings returns them in forward time order assending nmm MESSAGE FORMAT VERSION 0 2 def setupLogging logging format asctime s name s filename s levelname s message s formatter logging Formatter logging format log file logging FileHandler Receive log w log file setFormatter formatter log file setLevel logging DEBUG console logging StreamHandler sys stdout console setFormatter formatter console setLevel logging DEBUG rootLogger logging getLogger rootLogger setLevel logging DEBUG rootLogger addHandler console rootLogger addHandler log file class Receive ProcessClient NewsPostMixin def init self argv ProcessClient init self argv NewsPostMixin init self def run self FUNCTION
113. ontinuing to keep the United States at the leading edge of discoveries from astronomy to geology to zoology With an annual budget of 5 5 million NSF is responsible for 20 of federally supported research at America s colleges and universities Research in the Polar Regions of the Earth is of particular importance in these times of climate change and global warming In order to support scientists in the Polar Regions NSF contracts VECO Polar Resources VPR for all Arctic Logistics VPR supports 500 scientists working in 55 different field locations around the Poles The GeoSpace Center at SRI International supplies field communications services to VPR funded projects The WPI team worked at SRI to provide communications support to Dr Patrick Sullivan s study of the constraints on the physiology and growth of trees at the latitude tree line As communications technology pushes forward so does the demand for immediate access to data from sensors in remote locations around the world The resources expended to send a researcher to one of these remote locations to manually retrieve data is both uneconomical and impractical given today s ability to communicate autonomously Real time data coming from these remote systems would allow researchers to monitor operating status and keep up to date records while saving time and money The primary goal of this Major Qualifying Project was to design a general system that can provide real time access to
114. ount of power delivered to the battery in the shortest time without compromising the life of the battery 18 3 Specifications The system must provide autonomous and robust communications support for a Campbell Scientific CR1000 Datalogger using the Iridium Satellite Network The system must meet the transfer needs of Dr Patrick Sullivan for deployment in Kotzebue Alaska The system specifications can be broken into three main categories communication power and physical specifications 3 1 Communication Specifications The communication specifications include the amount of money that can be spent to send the data the amount of data to be sent and the type of data to be sent The system must meet the following communication specifications e Communication budget of 2400 per year e Minimum of one transmission per week e Each transmission must include o Battery Voltage o Enclosure Temperature o Atleast one sample of data from each sensor e Allow bi directional communication As the specifications state the system must send at a minimum one transmission a week that includes the battery voltage enclosure temperature and at least one sample of the sensor readings The datalogger will collect a set of sensor readings every hour so one sample signifies one of these data sets Ideally the system will transmit the battery voltage enclosure temperature and the full set of collected data every day but must transmit at a minimum of once per
115. r column names self parseHeader header readings self parseData message format version data phone number transmit timestamp column names self store readings 107 def def parseSubject self subject This function parses the subject into a transmit timestamp amp version www __FUNCTION__ sys getframe f code co name logging debug s entering FUNCTION pattern version a z0 9 mo re search pattern subject if mo version mo group 1 else logging warning s using NA for version __FUNCTION__ version NA pattern timestamp s sri ayoung DatetimeUtils iso8601_pattern mo re search pattern subject if mo ts mo group 1 transmit_timestamp sri ayoung DatetimeUtils parse_iso8601 ts else logging warning s using now for timestamp FUNCTION transmit timestamp datetime datetime now logging debug s version s transmit timestamp s FUNCTION version transmit timestamp return version transmit timestamp splitPayload self payload This function parses the payload into a header amp data nm FUNCTION sys getframe f code co name logging debug s entering FUNCTION sections lines string split payload n logging debug s d lines in payload _ FUNCTION len lines section name None line number 0 for line in lines l
116. ring Panel Temperature measurement PTemp C PanelTemp PTemp C 60Hz Voltage measurement of IR battery VoltDiff Batt Volt IR 1 mV2500 2 True 0 60Hz 0 01 0 0 Voltage measurement of PV panel VoltDiff PV Voltage 1 mV2500 3 True 0 60Hz 0 01 0 0 temp measurement of Enclosure Therm107 ETemp C 1 1 1 0 60Hz 1 0 0 0 Paddy s data to follow Exit Sub End Sub Connect Dial Up SubRoutine sends AT Commands to the Iridium Transceiver to connect via dial up mode Sub Connect Dial Up tmp connect success DIM tmp connect success DIM try reconnect tmp connect success 1 Initialize counter Dim D 87 Initialize AT Command to Connect DialUp Dim AT_COMMAND AS STRING 40 AT COMMAND ATDT REMOTE IR PHONE NUMBER Chr 13 Chr 10 For try reconnect 1 to MAX CONNECT TRIES step 1 Flush the Buffer SerialFlush ComRS232 Send Out the AT Command to DialUp SerialOut ComRS232 AT COMMAND 0 100 Gl Wait for CONNECT string to verify connection TIME OUT UNITS IN CENTI SECONDS SerialIn InString ComRS232 6000 CONNECT 100 For D 1 to 90 step 1 if mid instring D 7 CONNECT Then Exit Sub EndIf Next D Next try_reconnect if try_reconnect gt MAX_CONNECT_TRIES Then tmp_connect_success 0 endif Exit Sub End Sub Sub ReceiveNewPeriod Dim new_period initialize index Dim indexl index2 0 0 index
117. ringIO buf write s self timestamp a list for value in self values str Sg value a_list append str Str string join a list yt buf write str return buf getvalue The next six definitions set or get the transmit timestamp phone number or column names setTransmitTimestamp self timestamp self transmit timestamp timestamp def getTransmitTimestamp self return self transmit timestamp def setPhoneNumber self phone number self phone number phone number def getPhoneNumber self return self phone number def setColumnNames self column names self column names column names def getColumnNames self return self column names class ReadingTable def def init self connection self connection connection get self id None where None if id None return self getOne id id elif where None return self getWhere wher else return self getWhere 94 def getOne self id None where id d id readings self getWhere where if readings and len readings 1 return readings 0 else return None def getWhere self where None __FUNCTION__ sys getframe f code co name sqlStmt SELECT id meter id settz UTC timestamp legs w hrs max demand FROM readings if where sqlStmt WHERE s where try ro self connection query sqlStmt except pg ProgrammingError e
118. rldHello WorldHello WorldHello WorldHello WorldHello W orldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello Worl dHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHe llo WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello Wo rldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello World Hello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHel lo WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello MorldHello WorldHello WorldHello WorldHello WorldHello WorldHello Wor ldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldH ello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHell o WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello W orldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello Worl dHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHe llo WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello Wo rldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello World Hello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHello WorldHel lo WorldHello WorldHello WorldHello WorldHello Wo_
119. rogram can send an e mail alert to the administrator News group Data Data Data Data Data Data Data File File File File File File File 101010 HB ov 0o co 0o 3o ogu6 oc l E mail alerts 101030 record dat Figure 16 Publishing and Subscribing to Data from Central Newsgroup Server 2 4 4 Python Programming Language To configure the Data Transport Network Python will be used to manipulate data Python is a high level interpreted interactive object oriented programming language that is used in many applications and by many companies including Google Yahoo and Industrial Light amp Magic Python operates using automatic memory management and dynamic data typing Similar languages to Python that use dynamic data typing include Scheme Lisp Perl PHP and Ruby The portability of Python is convenient as it can be implemented using most operating systems including Windows Unix Linux and Mac 14 Python is also free and available for download from their website www python org The python website provides free tutorials and helpful links for programmers The language is intended to be fun to use as reflected in the name origin after Monty Python s Flying Circus and the humor implemented in most of Python s online tutorials One of Python s biggest goals is to make their programming language easy to use understand and implement By providing Python with built in modules and extensibility it can be used for
120. roughout an entire year without depleting the batteries energy consumption testing was performed Since the datalogger and the communications run off separate batteries energy consumption was measured separately Upon completion of these measurements it was possible to estimate the state of charge for each battery throughout a year 6 4 2 Communications Energy Consumption In order to determine the energy consumption of the power system the MultiTec 330 Digital Multi meter with RS232 connection was used logging the current drawn from the Iridium battery every 500ms The multi meter began logging the current when the datalogger switches on the transceiver before a transmission took place The meter logged the current throughout the dialing and data call until the call was over and the Iridium was switched off The current was logged for when the transceiver connects on the first attempt second attempt and third attempt The data size of the transmission is approximately equal to that in which the system expects to send once per day Connect on First Attempt Sending Data Hangs up 400 350 300 250 200 Current mA k oa o 100 50 0 14 28 43 57 71 84 98 112 Time sec 75 Figure 57 Current Profile Connect First Attempt Figure 57 illustrates the current drawn by the Iridium when it is able to connect on the first dialing attempt The most current is drawn during actual data
121. s stored and collected in the final design This program also allowed a method of collecting data which proved to be useful in testing programs which require data to be transferred from the datalogger 6 1 5 Send Data Purpose Following the successful implementation of a data collection program the next step was to take the data stored internally on the CR1000 and output this data through a serial port This furthered the understanding of the CR 1000 data retrieval functions Set Up Power on CR1000 load program with PC200W software and connect datalogger RS232 to Computer running hyper terminal Solution 64 Figure 53 Send Out Data Program The send data program builds off of the previous workings of collect data program This program collects the battery voltage into tablel every minute and then sends this data point In line 17 a counter is initialized and incremented with every scan in line 20 When the counter hits 12 this means that it is time to send the data point The data point is sent with the SerialOutBlock command which sends the data point as a binary float representation Testing To test the functionality of the data sending program hyper terminal was used to view the data stream from the CR1000 After a minute the following characters were shown on hyper terminal 65 f HyperTerminal File Edit Yiew Call Transfer Help Dae 538 28 Connected 0 00 47 Auto detect Auto detect SCROL APS apture Prinvecho
122. s accomplished by editing the data table declaration Line 12 in Figure 74 shows a DatalInterval instruction Parameters 2 and 3 of this instruction can be edited to change the storage rate from sensor readings Similar to the scan instruction shown in Figure 74 parameter 2 controls the storage rate while parameter 3 is used to control the units corresponding to the storage rate Figure 74 also shows an example of a program which scans and stores at different intervals Notice that on line 13 the data is recorded every sixty minutes but in line 20 it shows that the sensors are scanned every five seconds 124 Figure 74 Scan Store Intervals Example 2 2 2 Adding Sensors Inside of the developed CR1000 code is a subroutine which is built to hold the sensor readings This subroutine can be seen in Figure 75 Figure 75 Read_Sensors Subroutine Editing this section of the code is trivial Simply add or remove any measurement instruction calls into this section Please refer to Campbell s CR1000 user manual for specific details on measurement instructions In addition to editing the subroutine the data table declaration of the code will also need to be edited Figure 76 shows the data table declaration which was used in testing the data transmission code Figure 76 Data Table Declaration To ensure data storage simply add any processing instructions into the data table structure The variables in this data table structure correspond
123. s str 4 remove quotes timestamp sri ayoung DatetimeUtils parse iso8601 ts str numbers numbers_str string split data_str for number_str in numbers_str try number float number_str except ValueError e logging error s s is not a float FUNCTION number str return None numbers append number reading Reading Reading reading timestamp timestamp reading values numbers logging debug s timestamp s nNumbers d FUNCTION reading timestamp len reading values return reading def parseBinaryReading self reading element header This function parses binary input from the datalogger DEPRECATED mnm 101 __FUNCTION__ sys getframe f code co name logging debug s parsing d bytes FUNCTION len reading element if len reading element 0 logging warning s zero length reading element FUNCTION return None if len reading element 19 logging warning s length not even that of a ts s FUNCTION repr data return None reading Reading Reading reading header header reading phone number 800 555 1212 4 TODO iso8601 str reading element 0 19 ts sri ayoung DatetimeUtils parse iso8601 iso80601 str reading timestamp ts data reading element 19 I f The data is Il SIZEOF FLOAT 4 if len data SIZEOF FLOAT 0 logging debug s d
124. s the overall energy consumption of the entire system is less than 20Ah for a year of operation Further tests on the battery capacity confirming this estimation will be shown in Section 6 4 4 5 1 3 Solar Panel The India PV20 20 Watt Solar Model will provide renewable energy to the system The panel is monocrystalline and carries a 10 year warranty The panel will be mounted on a 10 foot pole at a vertical angle so that the panel faces the horizon The vertical alignment is designed to optimize the panel performance for the winter months when the stays low to the horizon throughout the day in Alaska This is desirable since the stronger summer sun provides more than enough power to keep the batteries fully charged 5 1 4 Charge Controller The Morningstar SunGuard will be used to control the charging of the two batteries The SunGuard provides a simple and economical charge control while having proven field reliability V ECO has used them in previous projects with success The efficient PWM charging of the SunGuard can provide up to 4 5A to the battery while monitoring the ambient temperature to adjust the charging cutoff voltage for the battery to avoid overcharging in the cold Due to this temperature monitoring the charge controller is to be mounted inside the box with the batteries In order to charge both of the batteries two charge controllers are used Diodes are inserted in between the PV panel and each one of the charge controllers T
125. scientific instruments located in remote regions of the world To implement the system the WPI team provided a comprehensive interface between a Campbell Scientific Datalogger and an Iridium Satellite Transceiver Additionally the team employed the Data Transfer Network created by SRI to make this information available to researchers To interface with the Data Transport Network the team used Python programs to process and distribute the data transmitted across the Iridium Network 2 Background This section provides background information on all relevant aspects of the project data collection data transmission system power and weather conditions 2 1 Scientific Background In order to gain a greater appreciation for the impact this project will have on SRI VECO Polar Resources VPR and the scientific community at large several topics must be understood In the following section the significance of polar research and real time data transmission will be explained Furthermore a short overview of previous arctic remote sensing deployments will be discussed This section concludes by outlining theories behind this project and how they will improve future remote sensing missions 2 1 2 Why Transmit Data in Real Time SRI has received increasing requests from researchers for near real time access to data from their instruments in remote locations These are often small stations collecting data from a few sensors and storing samples into
126. sensors that can be read by the CR1000 and its predecessors The two main sensors that will be implemented by Dr Sullivan are the CS615 soil moisture sensor and the CS107 thermocouple SDMs are peripherals that expand digital I O ports and analog output ports CSI uses NEMA 4X enclosures to protect the datalogger even in extreme weather conditions 2 3 4 Software CSI has developed proprietary software for use with its entire line of dataloggers This software package supports programming the device communicating with a PC and displaying data on the software s graphical interface There are several software packages available SCWin Program Builder allows the user to create programs using only sensor measurement and data output PC200W Starter Software allows the user to transfer and retrieve data from the CR1000 LoggerNet 2 X is CSP s comprehensive software package there are several other software packages that can be used to increase the capabilities of the LoggerNet 2 X software Real Time Data Monitor RTDM displays real time or stored data in a multitude of graphical formats 2 3 5 Short Cut SCWin Short Cut for Windows is a software package that is designed to make datalogger programming easy Short Cut implements a four step process to create simple programs with a user friendly graphical interface This package is compatible with a wide array of sensors Furthermore it permits the use of multiplexers with the datalogger to expand th
127. sh Serial Buffer i Send ATDT Command i Wait for Connect String onnect String Found Yes Exit Sub Next Reconnect Try Exit Sub Figure 33 Establish Connection Flowchart 5 3 5 Data Out As described in the Transmit Data subroutine after establishing a connection the next step is to send the appropriate data To do this the sub system as shown in Figure 34 was used 40 The first step in sending the records in a data table is to initialize local variables These variables will be used in the data transmission loop The next step is to send the header with header information containing an id number and appropriate column headings Header begin and end tags are also used to be interpreted by the local end This transmission is accomplished with a simple SerialOut command The subroutine then enters a loop to send out a variable number of data points To calculate exactly how many data points to send the timestamp associated with each row in the data table was used The loop then iterates through each record in the stack and continues sending until the last sent time stamp is greater then the current iteration The time stamp of the last sent record is then tagged to be used upon the next execution of the send data subroutine An end data tag is also transmitted and finally the subroutine exits and enters the period change routine 4
128. st will be sending so having the connection time is useful for having a complete time including any connection or hang up times Set Up For this experiment the complete system was used including the complete remote and local ends Testing The test was run over the weekend from Friday February 17 at 7 52PM until Tuesday February 21 at 7 06AM The local end collected all the data and stored it into a log file to be reviewed and analyzed after full completion of the test Conclusion The data that was collected during that test can be seen in Table 7 Begin Begin data Total Transfer bytes Connect transfer End Call Time Time sent Bps 19 52 27 19 52 37 19 53 46 0 01 19 0 01 09 12762 184 96 72 20 52 09 21 52 12 22 52 45 23 52 56 0 53 22 1 53 28 2 54 01 3 54 08 4 54 48 5 54 49 7 55 37 8 55 56 11 56 57 12 57 09 13 57 46 22 45 49 23 45 47 0 46 13 1 46 29 2 46 56 3 47 08 5 47 48 6 48 18 7 48 38 8 50 34 9 49 06 10 50 48 11 49 51 12 50 10 13 50 30 14 50 50 15 51 11 16 52 34 17 51 46 19 52 25 20 52 54 21 53 08 22 53 34 23 53 55 0 54 05 1 54 25 2 54 48 3 55 12 5 55 52 6 56 10 7 56 26 8 56 50 9 57 05 11 57 57 12 58 07 20 52 19 21 52 22 22 52 55 23 53 04 0 53 32 1 53 38 2 54 09 3 54 17 4 55 00 5 54 59 7 55 46 8 56 05 11 58 49 12 57 18 13 57 55 22 45 59 23 45 57 0 46 25 1 46 38 2 47 06 3 47 18 5 47 57 6 48 28 7 48 47 8 50 42 9 49 16 10 50 57 11 50 01 12 50 20 13 50 41 14 50 59
129. stent transfer properties from unreliable limited bandwidth network connections The Iridium network fits this description with calls often being dropped in handshakes between satellites and a maximum bandwidth of only 2400 bps Figure 15 shows the workings of the DTN as it applies to this system Data files being collected on the local end of the Iridium connection are saved to a networked file system The data is stored here until a posting program notices the new data file and posts it to a central newsgroup The information can then be accessed by anyone by logging onto the newsgroup where it was posted 13 Networked filesystem Data ata EE rc uum Pia nt x News group Local Iri udum Figure 15 Message Queue at Local Iridium Connection The DTN provides for a convenient system to distribute the data that was transmitted from the datalogger From the central newsgroup server where the raw data files are posted a Python based programming architecture can archive plot and monitor the data as seen in Figure 16 The Python Programming language will be described in the next section Typically a processing program will have one component which watches the server for new data from the remote site That program can trigger a program which archives the raw data and plots the processed data to a website for review by researchers The system can also be implemented to monitor the health of the overall system When problems are reported the p
130. sys getframe f code co name logging debug s entering FUNCTION datalogger Datalogger Datalogger while True logging debug s top of while __FUNCTION__ readings datalogger simulate readings datalogger getReadings if readings None or len readings 0 logging error s no readings to post _ FUNCTION else buf StringIO StringIO buf write HEADER n 105 o buf write s n readings 0 getPhoneNumber buf write timestamp col names string join readings 0 getColumnNames buf write s n col names buf write DATA n for reading in readings buf write Ss n reading toString timestamp datetime datetime now subject version s timestamp s MESSAGE FORMAT VERSION timestamp logging debug s posting message subject s FUNCTION subject self newsPoster setSubject subject self newsPoster postText buf getvalue f break f one shot logging debug s exiting FUNCTION LE name _ main_ setupLogging logging debug s starting name Receive sys argv run sys exit 0 Store usr bin env python import logging import sys import StringlO import string import datetime import pg import re import sri ayoung DatetimeUtils from Transport import ProcessClient from Transport import NewsPollMixin import Reading i
131. t be under 2400 per year The project is divided into two systems which will communicate with one another a remote end and a local end The remote end located in Kotzebue Alaska will collect data periodically and send it to the local end at SRI in Menlo Park CA The local end will receive the data format it and send it to the data transport network which will update a webpage displaying meteorological conditions at the remote site The remote end can be broken down into three systems data collection communication and power The data collection system consists of a multitude of sensors attached to a Campbell Scientific Institute CSI datalogger through CSI multiplexers The datalogger was programmed to collect data from the array of sensors and transmit it at user adjustable intervals The datalogger will also send a system health update with each data transmission The communications system includes an L Band 390MHz 1 55GHz Iridium modem and transmitting antenna The power system is comprised of three major components a photovoltaic panel a charge controller and two 100Ah gel cell batteries The solar panel converts the available sunlight into power which is stored in the battery via the charge controller increasing the efficiency of the charging system This testing station will collect soil moisture and soil temperature readings and transmit the data over satellite communications to be received by the local end The local end w
132. ta from the datalogger Since sending the full set of data daily is within the power and cost constraints it was decided that this was the best option for the scientist 22 4 2 Short Burst Data vs Dial Up After deciding the quantity of data to be transferred and the frequency it will be sent an Iridium communications service was chosen There are two ways to send data over the Iridium network short burst data SBD and data dial up calls 4 2 2 1 Short Burst Data Service SBD is designed to send and receive shorter data messages at a more cost efficient rate than dial up When using SBD the user does not need to establish a connection with another modem eliminating the dialing and connection time A basic overview of how the SBD connection can be seen in Figure 21 Internet Email Sensor or other inputs outputs Value Added Reseller Host System eae End User Applications lrid um L Band Transceiver Microprocessor Based System RS232 Senal Mobile Application Figure 21 Basic architecture of the Iridium SBD The mobile application loads the data message into the transceiver and then instructs the transceiver to send the SBD message to the Iridium Gateway The Iridium Gateway SBD equipment receives the message and sends an acknowledgement back to the mobile application An email message is then created with the SBD data message as an attachment to that email The email is then sent to the
133. tching Ci OU a aao tit haie dare Rus quM rena Dodo s et Ea 34 Figure 29 Switching Circuit S CHOMIALIC oio ie ioo srodne eri bd es SES Spass dU du 32 Fig re 30 Assembled Circuit BOX ois nodi ibus eai iaoeia inen i iiaa 34 Figure 31 System Flowchart Top Level ssssssseeeseeseesresresresseseresreesrrsrrreresressrseresresse 35 Figure 32 Transmit Data Flowehalt ie encontinen oo e ini ea 37 Figure 33 Establish Connection Flowchart esseeseseeseesreeseesersrrssesrresresrireresresseseeesreeee 39 Fig re 34 Data Q t PUTNAM MR 41 Figure 35 New Transmit Period Flowchart need Debe ice r Uttar oto neas 43 Figure 36 Basic Overview of the End System Code detti e ade rtoniindida 44 Figure 27 Overview of Roce esisiini gecs ieroi a i a UE DU MER E QURE 45 Figure 38 Overview of DatalOBSBE eei taie i penates eati basa decns idus 45 Figure 39 Overview of the receive function a ooasidediede inui hd rui ecpe it qud ae Dresd 47 Figure 40 Overview of the isNoCarrier function aei iit dt rhone rbd rin Een ass docen 48 Figure 41 Overview of the retrieve Data DUBCGUODL sai ieu icc haie bb ea dcinde ndotirindde 49 Fig r 42 Overview Of Uo C 50 Figure 43 Overview of store eadIngs uiuos sputo ete Ud enaska 51 Figure 44 Overview of plot Generate eite aai rtm need ibl rid Set ro Gas ia is Rtc tede 53 Figure 45 Website SIdeDAt cceccacccnsesdscescedsttnietaecdansduieteueidedassedasienietducdeutdasetpuesdeiasunact
134. the datalogger s A D ports to record health status and the internal panel temperature and battery voltage sensors were programmed to record at each storage interval Separate batteries were used to run the datalogger and the Iridium transceiver The switching circuit to toggle power with the Iridium transceiver was used with the appropriate CRBasic code to turn this port on and off Testing The test program was set to store health system data once per minute and transmit the data once per hour Each transmission session included sixty data points If the Iridium transceiver was unable to establish a connection with the remote system the system was programmed to append this data to be transmitted on the next transmission period 74 208 16 206 64 160 74 205 10 204 26 209 32 205 92 203 93 181 57 201 46 209 19 205 73 210 52 206 55 213 83 201 01 Conclusions The results from this test were very positive Out of the 3960 data rows collected over this weekend test 3720 of these table rows were successfully collected This implies that 4 hours worth of data out of 66 total hours of data were unsuccessfully transferred Looking at the log file shows that 4 transfer calls were dropped correlating to the 4 hours worth of data which was unsuccessfully transferred Looking at the overall reliability percentage for this test shows that 93 of the data was successfully transferred 6 4 Energy Testing To ensure that the system can operate th
135. the end data tag 96 Next a variety of unsuccessful transmissions were simulated An unsuccessful transmission means that the end data tag was not received and a NO CARRIER appeared meaning the connection had been lost The unsuccessful transmissions were simulated by typing NO CARRIER into HyperTerminal at different stages of the transmission For example halfway through the data reading a NO CARRIER would be typed Conclusion This test proved that the Python code worked as expected When a successful transmission was simulated a file containing the header and data was produced When an unsuccessful transmission was simulated the program quit and no file was created 6 3 Full System Tests This section of the testing results documents the results from testing the final prototype system System reliability and characteristics are stressed in this section and documented through various procedures described below 6 3 2 Reliability amp Frequency Purpose After each successful transmission the back end program sends the remote end a new transmission period in intervals of five minutes with a starting period of five minutes The benefit of sending the transmit period is that 1f the communication system is draining too much of the battery the local user can set a new frequency to send less often and therefore save power The second factor this tested was the reliability of the system This system ran overnight so it w
136. tion Given this knowledge researchers at SRI have elected to employ a Campbell Scientific Institute CSI datalogger to collect environmental information Continuous Analog Outputs Pulse Counting Switched Excitation Channels ce NH Channels ipia Digital a A ey Inputs Outputs rrr dart e e Single ended Differential Analog Inputs ER Char menos p n HEC Sr Power and Ground Qe ge c Connections CS I O Port I RS 232 Serial vo Keyboard and Display Figure 9 Standard components of a CSI Datalogger CSI dataloggers have been used in a variety of applications including meteorology agriculture air quality soil moisture HVAC systems water resources and geological research These dataloggers are used all over the world to provide accurate and reliable measurement and to control system performance CSI has intended these systems to execute one time data collection as well as ongoing data monitoring Additionally some CSI dataloggers can be programmed to respond to input conditions by executing operations such as actuating a motor or toggling a switch feo lt f f ANIPEr 1 sc L 5 ING TRI E Figure 10 CR1000 Datalogger This project will be designed specifically for the Campbell Scientific CR1000 Datalogger but the desire for a generalized solution will be kept in mind as well All CSI Dataloggers accept input from multiple peripheral devices These sensors may indicate things such
137. tion and Portset Port State The Portconfig instruction is used to configure a control port as either output or input Mask specifies which port to configure amp B1 port 1 and Function configures the port as either input or output 1 output The Portset instruction activates a port either logic high or low Port denotes which port to effect and State indicates logic high or low True logic high 5V False logic low 0V 5 2 6 Assembly After testing this circuit with the full system the circuit was soldered onto a piece of proto board and placed in the circuit box shown in Figure 30 The box has a sheet of insulting material on the bottom and ground wires are tied to the standoffs to create a chassis ground The circuit box was then screwed onto the mounting panel which would be later cemented into the final insulated storage box for deployment 34 Figure 30 Assembled Circuit Box 7 5 3 CR1000 Code Coding the CR1000 was done in CRBasic and required a divide and conquer approach To complete the final system it was important to start from a top level view and then divide the large system into smaller subsystems The following sections document the CR1000 code starting from the top level approach and working down into each sub level of the program The full CRBasic code can be read in Appendix A 5 3 2 Top Level System Figure 31 shows a top level flowchart of the solution The first step in the CRBasic code initi
138. to answer after just one ring with the command atsO 1 Here the serial port and the baud rate were set The period in which the data will be sent was initialized along with a buffer to store data After Logging and Initializing the program is ready to be run Receive will call getReadings which retrieves the data and formats it for integration with the Data Transport Network Retrieve and Format the Data The data that is received from the remote site is a continuous string of data that needs to be formatted so it can be easily read and interpreted The function getReadings calls the function receive which handles this data The function receive calls the functions needed to retrieve and sort through all the data and return this data back to getReadings Figure 39 shows a basic overview of the receive function 47 receive Calls functions to obtain data and is returned readUntil Receive CONNECT Return the Header amp Baud Rate Return the Data receiveHeader Reads the file header retreiveData Returns the data in the proper format End Connection Figure 39 Overview of the receive function First a function readUntil is called twice The first time it looks for the CONNECT string and the second time it looks for the baud rate The next function called is receiveHead
139. transmission but spikes are also seen when the Iridium is starting up and when commands are sent to the transceiver The average current drawn during actual data transmission is around 325mA The average current drawn during the entire time the Iridium is switched on is 245mA The following plots illustrate the current draw when the Iridium transceiver is not able to connect on the first attempt Connect on Second Attempt Second First Connection 400 irst_ E Connection ful 350 Unsuccessfu 300 z 250 E 200 9 5 Waiting to 150 Diat Again 100 50 0 0 28 56 83 111 139 167 194 222 Time sec Figure 58 Current Profile Connect Second Attempt Figure 58 shows when the transceiver is unable to connect on the first try During this test transmission the log file on the local program shows that there was a brief connection but the call was quickly dropped The modem then waited two minutes and then attempted to call again and was able to connect and transfer the data successfully 76 Connect on Third Attempt Third Connection Second Successful 400 onnection Unsuccessful First Dialing Attempt Unsuccessful Current mA Time sec Figure 59 Current Profile Connect Third Attempt Testing showed that unsuccessful transmissions have a variety of different profiles The first dialing attempt in Figure 59 shows the profile of dialing wh
140. tting the data into a database and into graphs The python program Store puts the data into a database and plotGenerator graphs some of the data Both of these codes can be seen in Appendix B 50 5 4 3 1 Store The Store program obtains data from the Data Transport Network and puts these values into a table that will be posted onto a website so the scientist can access the data A basic overview of the Store program can be seen in Figure 42 process Obtain the subject amp data from the Data Transport Network storeReadings Collects the information that needs to be stored store Opens a connection to the database amp adds any new readings Figure 42 Overview of Store Store begins with the function process which obtains the subject and payload from the Data Transport Network Payload here is referring to all the data including the header and the values These are both passed into the function storeReadings An overview of this function can be seen in Figure 43 51 storeReadings Return version amp transmit timestamp Return header parseSubject amp data Return phone number amp column names Return readings splitPayload parseHeader parseData store plotGenerator Figure 43 Overview of storeReadings First storeReadings calls parseSubject whi
141. ude is 485 miles at 16 832 miles per hour This configuration ensures that any part of the earth is covered by at least one satellite at all times 12 Figure 14 Iridium coverage Figure 14 shows the Iridium Networks coverage The darkest areas indicate the best coverage The Iridium Network is ideal for systems in the Polar Regions because of the concentration of satellite coverage in this area The second component of the Iridium Network is the ISU which places and receives calls When a call is placed from one ISU to another the call is directly routed by passing the call from one satellite to another until it has reached a satellite above the intended receiver For a call to a remote local area network LAN or to establish connection through the public switched telephone network PSTN the Iridium Gateway must be used to establish connectivity The third component of the Iridium Network is the Iridium Gateways There are currently two commercial Iridium Gateways one in Arizona and the other in Fucino Italy Each user is registered to one of these Gateways The Gateway is responsible for keeping information about its users It also routes calls from an ISU to the PSTN or other land based networks 2 4 3 Data Transport Network The Data Transport Network DTN was developed by SRI as a way to manage the collection of data from an instrument and deliver the information to interested parties It was made in response to the inconsi
142. umn names and is used in the four codes that were discussed in Section 5 4 usr bin env python 3 port datetime port struct port logging port sys port StringIO port os port string port pg 338 3 3g H p b H H H H H E 3 import sri ayoung DatetimeUtils def fromString str This function separated the timestamp from the data then converts the data from a string to a float mnm FUNCTION sys getframe f code co name try ts str numbers string split str 1 except ValueError e logging error s unable to split Ss FUNCTION str return None reading Reading reading timestamp sri ayoung DatetimeUtils parse iso8601 ts str value strs string split numbers for value str in value strs try value float value str reading values append value except ValueError e logging error s unable to convert s to float FUNCTION value str return None logging debug s parsed ts s d values FUNCTION reading timestamp len reading values return reading 93 class Reading def def ww www def __init__ self self phone_number Non self transmit_timestamp None self timestamp None self column_names None self values toString self This function creates a string with the timestamp and values ww __FUNCTION__ sys getframe f code co name buf StringIO St
143. us PTemp 25 m Health 24 s 23 22 Data ca 21 20 19 Query asl Updates z 02 24 02 25 02 26 02 26 02 27 21 00 12 00 03 00 18 00 09 00 m Release Notes Batt_Volt 12 10 28 6 4 2 SUY H H 9 02 24 02 25 02 26 02 26 02 27 21 00 12 00 03 00 18 00 09 00 Batt_Volt_IR 12 10 28 6 4 2 OY H H 9 02 24 02 25 02 26 02 26 02 27 21 00 12 00 03 00 18 00 09 00 PV Voltage 02 24 02 2 02 2 02 26 02 27 21 00 12 00 03 00 18 00 09 00 ETemp 02 24 02 25 02 26 02 26 02 27 21 00 12 00 03 00 18 00 09 00 Figure 46 Website s System Health Check Figure 46 shows data that was collected hourly from February 24 2006 until February 27 2006 Store will check the Data Transport Network for new values these values will be added to the website If there are no new values it will check again The last recorded data transmission can be seen just below the HEALTH heading 57 The graph is useful for the scientist as a quick confirmation that the system is still functioning properly However the complete database is available on the website also This can be found by clicking on Query on the sidebar The screen in Figure 47 will appear The date range can be selected so only the data of interest is displayed There are two options on how the data is displayed It can either be saved as a comma separated value csv file or viewed on the current web browser This method is suggested for viewing a large amount of data query
144. ze Variables Y Take Sensor gt Readings Call Table Y Mark The Current Time Y No Calculate The Difference In Previous Time Sent to Current Time Time Difference Tx Period Yes Transmit Data Figure 31 System Flowchart Top Level 5 3 3 Transmit Data Figure 32 shows a flowchart of the transmit data subroutine This larger subroutine is basically a collection of smaller subroutines which will be explained in more detail in later sections 36 The first step in transmitting data is to turn on the transceiver As explained in section 5 2 this is accomplished by turning on a 5V control port After turning on the 5V control port there is a delay of 20 seconds to allow the transceiver to power on The program then continues on by establishing a dial up connection After a connection is established the program then starts streaming data out of the RS232 port Immediately following the transmission of the end data tag the program executes the period adjustment subroutine The period adjustment subroutine allows the user on the local end to bi directionally communicate with the datalogger and gives allowance to remotely change the transmission frequency should there be a need for it as well as give the datalogger confirmation that the data was successfully transmitted If the user does not change the frequency or gives an invalid frequency then the program

Download Pdf Manuals

Related Search

Related Contents

Samsung RSH1STPE 用户手册  Yealink SIP IP Phones V51 Firmware Change Log  GB/US.........GH1/GH1.Q.Ceiling.Hoist  Gel documentation systems UVI UVP chemiluminescence gel  MTX-D, Oil Temperature/Pressure Gauge  取扱説明書 - 三菱電機  User`s Manual  OPERATING AND SERVICE MANUAL  Emulex Drivers for VMware ESXi 5.0 User Manual  Manuale d`lstruzioni Benchman - BioPak 240 Revolution 4  

Copyright © All rights reserved. Failed to retrieve file