Every NSW High School has had data loggers for about four years. Unfortunately like a lot of new technology in schools these are often underutilized. Most Science teachers have completed "inservice" courses on their use and most can use them to record temperature or voltage over time as a demonstration but the data loggers are very rarely used in an ongoing role within normal classes. A data logger is the perfect A/D interface for a seismograph to connect to a computer over weeks or months.
Each school chose whichever brand of data logger they felt was most suitable for their requirements there are about five common types of data logger (Pasco, Vernier, Data Harvest, Fourier and Pico) in schools and possibly a few less common types as well. Because of this I have designed the detector circuits to produce a voltage output; voltage input sensors and temperature sensors are supplied with nearly all marketed data logger packages. The detector circuit allows for high or low level outputs for different brands of datalogger. Multiple channel data loggers could even log two seismographs set as North/South and East/West or Short Period and Long Period or Horizontal and Vertical
I have used Pasco, Data Harvest, Pico and Fourier data loggers so far and have used the software to record and display seismograph records. The Pasco, Pico and Fourier software allow logging over many days when samples are taken every 5 or 10 seconds. These could be checked each week or when earthquakes are mentioned on the news or on web sites. The US Geological Service and Geoscience Australia have excellent earthquake sites updated hourly.
The Data Harvest Easysense Data Logger unfortunately can only log up to 4000 points; about 12 or 24 hours depending on the logging rate. The Australian suppliers have suggested that the software might be able to be rewritten to allow the data to be dumped to the computer hard drive to allow continuous logging. Use the "trigger" option to allow a threshold voltage level ( at the start of an earthquake) to trigger the logging of an event for the next hour or two.
All of the commercial data loggers come with "student friendly" display options and graphing. The Pasco Xplorer display above is typical. Scales can be easily moved, compressed or stretched, samples taken, labels changed and files can be stored. Several allow "export" of data as txt files to Excel programs. I will develop flow charts for each of the common data loggers to allow teachers and students to easily build the seismograph bar and detector and setup the data logger and computer programs to record, print and analyse earthquake wave patterns.
|August 2005 Update|
A "Picaxe" A/D converter circuit is now available for the seismograph so that people without a data logger can detect and record earthquakes. The original article is to be published in Silicon Chip magazine, probably in September or October 2005. The Silicon Chip people have designed a PCB that can be used for the detector circuit alone or the detector and Picaxe together. Their PCB fits inside a Jiffy Box with a slot cut so that the screen attached to the seismograph bar can block the light beam inside the box. This will reduce the "drift" caused by changing light levels within the room. You might be able to "borrow" this idea and have the LED and LDR inside the box.
The Picaxe is an 8 pin mini computer that can be programmed down through a serial cable from any early Pentium computer, detect up to four different types of analogue or digital inputs, process the data and output results through up to four channels as ASCII, digital signals, music or even pulses to drive robot motors; all for about $5. Information and the download program are available from http://www.rev-ed.co.uk/picaxe ;suppliers include http://www.microzed.com.au/ , http://www.oatleyelectronics.com/ and http://www.altronics.com.au/ in Australia and http://www.sicom.co.nz in NZ. Remember that you must order the 08M Picaxe chip, not the earlier 08 chip which is not as powerful.
Searching through the Picaxe forum files from Revolution-Education the Picaxe producers in England I discovered two letters from "Manuka" in NZ about using a Picaxe with a graphing program called "StampPlot" designed for using serial data from a "Stamp" chip. Stan Swan also refers to StampPlot Lite in Silicon Chip articles in August 2003 and October 2004. The free!!!! "StampPlot Lite" program can be downloaded from SelmaWare (http://www.selmaware.com/ ) in the US and the only changes needed are to specify the baud rate at the top left of the graphing screen to 4800 (the same as the 08M picaxe), and to specify the same COM port as the Picaxe is using. The StampPlot Lite can record over 200000 points, and dump them every few minutes to the computer hard-disk as a txt file, the scales of the graphs can be altered, the x-axis can be scrolled to show the last few thousand points (1 to 2 hours at one point per second), maximum and minimum recorded values can be shown and times (from the computer clock) can be saved with each data value.
The naming/numbering of the leg on the Picaxe chips is odd. Please be very careful. On the circuit diagram I have numbered the legs as 1 - 8 like the op-amp and this is how they will be connected on the board HOWEVER the programming language use a different number sequence. "readadc 4" actually means read the adc value at leg 3. I have not worked out the full reason but it could be buried somewhere in the development history of the chip.
The circuit uses the same pin arrangement as the "Schools Experimenter" in the May 2005 Silicon Chip. My first attempts used a LED output from the seismograph detector circuit shining on a LDR and resistor as a voltage divider as the input signal for the A-D converter. The circuit described here is simpler, with the op-amp connected to the Picaxe chip through a trimpot voltage divider.
The detector is identical to the previous design except that a 5k trimpot replaces the 3.3k resistor in the output voltage divider. This allows you to set the maximum input voltage to the Picaxe to 4 volts. Building and set up are unchanged; even with the extra components the Dick Smith H 5605 prototype board is still not overcrowded.
The 7805 only supplies a few mA and the output is filtered with a 0.1 capacitor. A heat sink will not be needed. The socket for the Picaxe 08M is soldered so that the actual "physical pins" of the Picaxe are the same way around as the op-amp; leg 1, the positive supply is a the lower left end of the chip. It is connected to the output of the 7805. The "serial in" (leg 2) is connected to a 22k and then to the download lead from the serial port. Leg 8 is connected to the zero volt line. Leg 3 is the analogue input from the detector circuit.
|A long serial cable could be purchased for the seismograph but I have tried several old "serial mouse" cables and they have all had the correct leads. They also had one extra lead that I just clipped off. You will need to cut the cable close to the mouse strip the wires back and use a multimeter to identify which lead goes to which serial socket hole.|
Before you fit the circuit board into the jiffy box and before you fit the op-amp or Picaxe chips you can make several voltage checks. Plug in the power pack and turn on. The LED should come on. Bend it and the LDR so that the light shines on the LDR. Use a digital multimeter to measure the voltages between the zero volt line and the pin holes in the empty sockets. Check your soldering if values are not close.
||Picaxe Socket; (socket pins NOT Picaxe function pins) |
Unplug the plug pack, fit the op-amp chip only and plug the plug pack back in. Check the voltage at pin 3 of the Picaxe socket with the variable resistor set half way through its range of motion. The voltage on pin 3 should move up and down as you cover the LDR with your fingers and then remove them. Check that the voltage at pin 3 never exceeds 4 volts. You can change the value of the 5k trimpot resistor in the output voltage divider to bring this close to 4 volts. Unplug the plugpack and fit the Picaxe 08M so that the notch is facing to the left.
The program for the Picaxe can be downloaded from the Rev Ed site mentioned above.The download program for the Picaxe is VERY, VERY simple.
|You can actually "Copy" the whole program and "Paste" it into the Picaxe Programming Editor. The comments on the right hand side are ignored. An option allows you to "Check syntax" of the program. To download the program to the Picaxe chip you must have the Picaxe circuit turned off when you press the download arrow, and then quickly turn the picaxe circuit on. A series of blocks, filling across the screen show that the program is loading onto the picaxe.|
Once the program is loaded and running on the 08-M (check by looking at the "Debug" screen) you must close down the Picaxe Programming Editor program to free the Com Port so that the StampPlot Lite program can use it.
Maybe an "Alarm" loop could be included later to tell you that b1 has exceeded a certain value (and warn you to run to higher ground). A Magnitude 7.2 quake 160 km off California on June 14 this year led to a tsunami warning. We all know about the Boxing Day Tsunami but few Aussies realize that quakes near NZ and Macquarie Island possibly pose a threat to SE Australia.
| Open the StampPlot Lite program
|September 2005 Update|
|AmaSeis; the best seismograph display program yet|
|The Jan 27, 2006, magnitude 7.7 quake from the Banda Sea, Indonesia, recorded with a light detector. Note how the sensitivity and hence the background noise has changed over the 13 hours shown, as the resting position of the seismograph bar has moved slightly. This is probably a result of the heating and cooling of my home. In a school lab with a large thermal mass this will be less of a problem but it will still need to be rezeroed every few days. Maximum sensitivity occurs when the edge of the shadow is near the centre of the LDR. The big advantage of the magnetic detector is that the sensitivity stays constant.|
In December, I discovered that the Dataq DI 194 RS, can feed directly into a seismograph recording program written for US schools called AmaSeis. It is available at http://www.geol.binghamton.edu/faculty/jones/AmaSeis.html. The program can even be downloaded onto two floppy disks so that you can load the program on an older computer without internet access. The AmaSeis program was written to accept data from a US$500 vertical seismograph that came with its own amplifier and A/D converter. It could also work with up-market Dataq A/D converters. When I had trouble ordering the up-market devices, I tried to use the cheaper DI 194RS converter and found that the program has recently been modified to work with it.
This program can run continuously and shows a whole day's recording as an old fashioned helical recording. The advantage of this is that students can see the last 24 hours as 24 traces each 1 hour long. The current hour is always the bottom line and the traces all move up each hour. The time on the display is Universal Time not local time (when I first ran the program I thought that there was a mistake because it would not show "Sydney" time). This makes it easier to compare your data to quake information from Geoscience Australia and USGS which use Universal Time.The IRIS site, giving loads of information about AmaSeis and its original seismograph is at http://jclahr.com/science/psn/as1/ . It even has suggestions for putting the display live onto the internt. I will try this from Turramurra High School in 2006. The official setup manual is at http://www.scieds.com/spinet/pdf/AS1AmaSeis.pdf.
|4. You will have to "Select Device" as "DATAQ 194".|
|5. Restart AmaSeis and hopefully you will see the "Helicorder" display|
The Jan 27 2006, Magnitude 7.7 quake from the Banda Sea in Indonesia, detected using the low noise magnetic amplifier describe in the September update.
6. In SETTINGS , THIS STATION give your latitude and longitude so the program can correct for universal time. I could not get it to accept north and south so I put Sydney in as 35 and 151. You must restart the program again |
7. SHOW DATA VALUES will let you see the "number" being recorded. You should to adjust the "Zero Level" until the "Data Value" is close to 0.
8. SET ZERO LEVEL tells the program to add or subtract a value (about 2000 for a Magnetic detector and about 2800 for a Light detector) so that the data value is normally close to zero. The Zero level will probably drift slightly over a few days.
9. The "Helicorder" option in "Settings" allows you to adjust some parameters of the display. You will need to adjust the gain of the detector and the gain in the program so that the background noise is small but so that the "blow test" shows like the real quakes above.
|10. The screen will show a series of horizontal lines (the last 24 hours) and will begin to record on the bottom line (now). Check that the time on the display stays accurate. If it is fast or slow you will have to adjust the computer clock in Window's "Control Panel".|
|The program saves the data files. You can also capture the image on the screen by clicking BOTH Alt and Print screen at the same time. This copies a bitmap of the screen to the clipboard. You can then "paste" it into "Paint" for printing.|
|DataQ DI-194-RS data logger; the best and cheapest data logger I have found.|
Information about the DataQ DI-194-RS is available from the DataQ website at http://www.dataq.com/ or from the Australian suppliers Total Turnkey Solutions at http://www.turnkey-solutions.com.au
|The DataQ DI-194-RS data logger is one of the easiest to use with the seismograph as well as being extremely cheap to buy (about $60 + GST and delivery). It has no internal memory or display so most of the real data logging functions, are carried out by the computer running Win 95 to XP. It is a 10 bit, 4channel analogue to digital converter that is powered by the serial port of the computer via the supplied cable. Because it works so well and is so cheap, I would recommend that schools (particularly those schools with Data Harvest data loggers) consider purchasing one, to free up their data logger for other uses.|
The inputs of the DI-194-RS are screwdown clamps that take the wires from the detector. Input voltage must be between +/- 10 volts. The display can be altered to read 0-10 volts. If you find that the detector output rises to over 10 volts on strong signals you can use a lower voltage plug pack; a "6volt" unregulated, plug pack will probably give about 9 volts maximum output. A second solution is to add a 3.3k ohm resistor between the output of the op-amp and the high level output of the detector. A "12volt" unregulated plug pack gave up to16 volts output. This dropped to 9.5 when the 3.3k ohm resistor was added to the circuit. A third solution is to use the low level output of the detector and alter the scale in the program.
|The "WinDaq" programs supplied on CD with the data logger include WinDaq Lite recording program, WinDaq Waveform Browser, software and hardware manuals and sample files. The WinDaq Lite program can "only" sample at up to 240 samples per second and so is not suitable for the "CRO" uses that the full featured program allows but is still much better than is needed for the seismograph. Because the computer is recording the data the storage file size can be made very large, allowing more than one week of recording even at 10 samples per second.|
Like all data logger programs some of the useful features are "buried" under unrecognisable names or icons:
|The "WinDaq Waveform Browser" allows you to "browse" back through stored files. It can even look at the file currently being recorded.|
"F7" again allows compression. 100 times compression allows you to look through a file quickly OR 10 times allows detail to be seen if you know when the quake occurred.
In the "Options" dropdown click on "Cursor Time" so that the Date and Time will be shown at the bottom of the screen, allowing you to move the cursor to the time of a quake.
"Print" in the "File" dropdown allows printing of parts of the file.
Below is a printout of a 7.7 Magnitude quake from PNG on Sept 9. Note that the "Cursor Time" shows 09-09-05 (the date) and 17:35:28 the time shown by the black line. Each division of the display represents 80 seconds. The time is derived from the computer clock and should be checked every few weeks as these do drift significantly and should not be relied on for accurate arrival times. The relative arrival times however will be accurate. If the computer is a not connected to a network or internet then the clock can be changed to UTC time.
The top channel is the N-S recording and the second channel is the E-W recording. The vertical motion sensor on the third channel was not sensitive enough to show any recognisable pattern.
If you want you seismograph display to look like a "real" seismograph drum chart like those produced by pre-computer seismographs then there are two options.
A free program called HeliPlot is one of several seismic programs in a file called seismic_dataq.zip written by Bob McClure and available at http://jclahr.com/science/psn/mcclure/wdq_utilities/. HeliPlot can directly read data from the DI-194-RS and other DataQ data loggers and display it "live" or it can read stored files of quakes. The advantage of running the program "live" is that it can show the last day on one screen. Windows XP needs to have a Visual Basic update from Microsoft to run the HeliPlot program. This worked well when I tried it. The only problem with a drum display is that it shows every "blip", "artefact" and false reading over the last day. There are lots more of these than there are real earthquakes.
The HeliPlot above shows the N-S recording of the September 9 earthquake in Papua New Guinea.
The same event recorded by a N-S long period seismograph run by Geoscience Australia at Riverview in Sydney. Seismograph records of all quakes in Australia and quakes larger than 6 worldwide are available at http://www.ga.gov.au/bin/listQuakes
Yes their's is bigger and better than mine but I am getting closer.
It is interesting that the CD with the DI-194_RS had a sample file called seismograph that was just a single line graph that could have been a recording of data from a real earthquake but with no additional useful information. The CD should have mentioned the programs written by Bob McClure.
Some of the more advanced DataQ data loggers including DI-154-RS can feed data to a seismograph recording program called AmaSeis available free from http://www.geol.binghamton.edu/faculty/jones/AmaSeis.html in the United States. A number of US schools use this program with a simple vertical seismograph.
The serial data output of other data loggers might be able to be read by these programs if you can change the data stream or program. If you do manage to get them to work please let me know. If you know how to post the seismograph display live to the internet please also let me know.
Data Harvest Easysense Advanced
You will need
Plug in the serial cable to the computer and data loggerStep 2 Graph program options
It would be best if the program could record data every second for five or ten days like some data loggers but the Data Harvest program only records a few thousand points. This means that you have two possible ways of recording data from the seismograph; continuous or triggered
Continuous RecordingStep 3 Graph Program
Click on the "Red Arrow" at the bottom left of the screen to start the recording run and on the "Red Square" to stop the run if required.
The "File" drop down menu has "Print Setup" and "Print" options as well as "Save As". I suggest that you save files with file names using the date and 24 hour time that you began the recording run (eg Mar 20 1433). You could also use this as the "Title" of the graph.
To reset the program or change recording times stop the recording by clicking on the "Red Square" and then click on the "White Sheet of Paper" which opens the "Recording Wizard"
Below is the March 29 magnitude 8.7 quake recorded with the Data Harvest Easysense Advanced set on trigger.
|Using the "Meter", "Stop", "Enter" and arrow buttons it is also possible to run the Easysense independently of a computer and download data later.|
|November 2005 Update|
The DrDAQ data logger from Pico ( www.picotech.com )in England is a good and cheap school level entry device. For about $200 it allows continuous logging using several built in sensors (Light, Temperature, Sound, Voltage or Resistance) or using external sensors (such as pH). The voltage input is via screw terminals.
Power and data feed for DrDAQ is through the printer port cable. It comes as a naked circuit board and so must be kept away from prying fingers. The PicoLog program gives DrDAQ its real power, enabling up to 1 million points to be plotted over days or weeks at one sample per second. With the inbuilt microphone it can be used like a CRO to display voice wave patterns at up to 15000 samples per second.
Updates of the PicoLog program are available from the website. Like all data logging programs this one also has a few tricks.
In the "File" dropdown you must use New Data to give each file a name. When you first open the program the suggested name is NONAME 1. The best naming is as below.:
At the right hand side of the screen are three small icons click on the middle "View Graph" icon to display the graph. The horizontal row of icons allows you to alter the compression and position of the horizontal scale. When you start a new file the graph seems to remain at the end time of the previous recording. You have to use the minus icon to compress the scale and then repeatedly click the double left arrow to race back to the start of the graph. Use the plus icon to expand the scale back to normal.
The vertical icons do the same. The "Tick" icon allows you to change the label on the horizontal axis to "Date/Time" and "Scroll" keeps moving the graph to the left as the line reaches the right hand end of the screen. The "Scroll" instruction should be the last instruction, otherwise it sometimes drops back to the fixed scale.
The MultiLog Pro is one of the better and easier data loggers to use for seismograph. The graph option allows you to alter the axes and the small hand icon at the bottom allows you to make the X-axis show Time and Date. Even after 6 month's use I never worked out how to get rid of the table on the right hand side of the graph. For display in a lab it would be better if the graph occupied more of the screen. I used several channels at the same time for N-S, E-W and vertical detectors but using one channel only allows a longer timespan before needing to be reset.
The display show was from a Magnitude 6.5 quake from PNG
The Pasco Xplorer was used as the datalogger and worked well if the data was fed through to the computer. The newer Xplorer GLX has a bigger memory and could log for weeks to its own memory. Data, could be downloaded if an earthquake was reported on the USGS or Geoscience Australia websites.
Nearly all schools will have voltage sensors with their data loggers. If not low or medium level light sensors could be used to monitor the light output of a LED connected across the high and low level outputs of the detector. A better solution would be to purchase a DATAQ DI-194RS.
|Picaxe Seismograph January 2006 Update|
|A quake detected from PNG on December 12, 2005 on a "Picaxe Seismograph". This is the design in the Silicon Chip magazine. A light detector feeds to a $4.70 Picaxe chip programmed to operate as an analogue to digital converter. The serial data stream is then graphed by the "StampPlot" program.|
|A much larger 7.7 quake from the Banda Sea in Indonesia on Jan 27. The data file can be exported into Excel if needed.|
Page Maintained By: BioSciCH@mail.usyd.edu.au
Last Update: Monday, 30-Apr-2012 14:59:09 EST