I have been slowly building the infrastructure needed for Otter Creek Research Station, my long term amateur science project. The basic concept is to have a variety of atmospheric condition sensors continuously recording data and made available on a web site. The sensor platform will be located some distance from the house so there is a need to communicate the data back to the home network. The platform will be a bit further away than the maximum length allowed for an Ethernet cable so some sort of radio link is needed.
I bought a wireless access point and set it up in my office and tested the range with my cell phone. It gives good signals out past where I will locate the sensor platform. If need be I can add some directional antennas.
SparkFun sells a little module by Roving Networks (RV-XN) that is a WiFi radio and processor. The module includes an 8 input A/D converter with 14 bits resolution. The software that comes with the module includes function to sync the internal real time clock to the NTP and report the sensor readings at a periodic rate to a web server.
I am currently working with a development setup using the RV-XN to get the bugs worked out and understand what I will need to use it.
The modules are just wired together
The development setup consists of a series of modules from SparkFun. In actual use the battery and regulator will be replaced by a system source of 5VDC power. The USB serial port is used to configure the module via a serial terminal program on a PC. In actual use the serial port on the module would be connected to a small control processor like an Arduino or not connected if the module can be configured to do everything needed. The Xbee explorer is an adapter board for the RN-XV and includes a power regulator to output the 3.3V needed by the RN-XV.
The module will be configured to set its internal real time clock but getting network time from a NTP server running on my PC. This will provide accurate time stamps for the sensor readings. It will report on a periodic basis the sensor readings to a web server also running on my PC. Software called by the web requests will log the sensor readings into a MySQL database. Other scripts can access the MySQL database of sensor readings and generate reports and graphs to show on a web page.
To try out the whole system there needs to be a sensor to monitor. For that I have chosen to use a pyranometer based on the pyranometer described on Dr. David Brooks web site. The RN-XV A/D converters have a full scale input voltage of 400 mV which is a reasonable match for the output of the pyranometer silicon diode and shunt resistor. This saves me from having to build a transimpedance amplifier for the detector. Although in the final version I may have an amplifier just to get the output level right where I want it to be.
Pyranometer Head
One could mount the sensor head on a little metal box, put the electronics inside, add a solar panel and have a stand alone sensor the could be left outside reporting the solar intensity continuously.
Now that the hardware is ready time to work on the software. When the software is working I can then work on testing and calibrating the sensor. I want to make sure the cosine filter works reasonably well.
References:
Dr. Brooks web site http://www.instesre.org/
SparkFun Electronics http://www.sparkfun.com/
Jim Hannon's Blog 