I think at least one of the people wh visit this forum have been developing a weather station.
I have been thinking the same thing. I want to develop a wireless, solor-powered, battery-backed weather station with Temp, humidity, rainfall, solar irradiation. Maybe wind speed and direction too.
I am thinking about having the uC sleep for a while (longer when running on battery at night, shorter sleeps during the day, even shorter sleeps when measured varables are changing fast).
It will awaken and do some readings. Maybe use a tipping bucket for the rainfall gauge which triggers an external counter. The counter is read when the ZX wakes up. Maybe have the rainfall (and or) windspeed trigger an interrupt or a counter on the ZX itself.
If you have already answered these issues, I'd like to pick your brains, including the selection of sensors.
-Tony
Weather Station?
Mine is a commercial LaCrosse WS3600, wireless sensors. Its serial port connects to a PC. A program on that PC, written in C by the user community (open source) and named open3600, grabs the current weather (or internal log data).
My software runs every 15 minutes, runs open3600, and takes the result and puts it into a MS access database. The PC is on 24/7 as it also runs a web server and home automation and other stuff.
I have asp code that sends out web pages of current and historical data to web browser.
I thought about having a micro read the WS3600 serial port and bridge to ethernet or some such.
But the weather stations with temperature/wind/pressure/humidity sensors, indoor/outdoor with rain and wireless are well under $100 now, so I just can't rationalize it.
The 1-wire sensor set from http://www.aagelectronica.com/aag/ are a good start for a bare metal system.
My software runs every 15 minutes, runs open3600, and takes the result and puts it into a MS access database. The PC is on 24/7 as it also runs a web server and home automation and other stuff.
I have asp code that sends out web pages of current and historical data to web browser.
I thought about having a micro read the WS3600 serial port and bridge to ethernet or some such.
But the weather stations with temperature/wind/pressure/humidity sensors, indoor/outdoor with rain and wireless are well under $100 now, so I just can't rationalize it.
The 1-wire sensor set from http://www.aagelectronica.com/aag/ are a good start for a bare metal system.
Those are quite good prices, though when you all all the pieces to the system, the costs mount up.
This does look like a good system. I just wonder if having the ZX be the 1-wire master would make interfacing easier (other than to an ethernet lan).
It looks as if many of the units are powered by coin-type lithium batteries. I was hoping to go solar powered with a solar-charged battery backup, rather than some individually powered sesnros.
Some of the 1-wire sensors can be "parasitically" powered, as I am sure you know.
-Tony
This does look like a good system. I just wonder if having the ZX be the 1-wire master would make interfacing easier (other than to an ethernet lan).
It looks as if many of the units are powered by coin-type lithium batteries. I was hoping to go solar powered with a solar-charged battery backup, rather than some individually powered sesnros.
Some of the 1-wire sensors can be "parasitically" powered, as I am sure you know.
-Tony
I have some 1-wire temperature/humidity sensors in my system - for tangential purposes like the fan temperature of the server in the garage and front yard temperature. But in reality, I've found that you want just one outdoor temperature sensor, since the location of the sensor affects readings by 5 degrees or so. Family members don't want the extra readings - too much data.
An attic temperature sensor can be used as part of a home automation system such as I have, to influence attic vent fans and so on.
But with the all-in-one consumer made-in-China weather stations from LaCrosse and Oregon, the DIY weather station has to be just for the fun of it.
If I were doing one, I'd inteface all the sensors to a diskless micro and have it dispense web pages and some sort of data feed by TCP so a database system can pull and archive.
The WuHu Weather Underground guys have an active forum. They have weather stations as I discussed above, in the sub $100 area, and some have the next step up Davis weather stations. The push weather data to a web server that produces statistics and graphics.
An attic temperature sensor can be used as part of a home automation system such as I have, to influence attic vent fans and so on.
But with the all-in-one consumer made-in-China weather stations from LaCrosse and Oregon, the DIY weather station has to be just for the fun of it.
If I were doing one, I'd inteface all the sensors to a diskless micro and have it dispense web pages and some sort of data feed by TCP so a database system can pull and archive.
The WuHu Weather Underground guys have an active forum. They have weather stations as I discussed above, in the sub $100 area, and some have the next step up Davis weather stations. The push weather data to a web server that produces statistics and graphics.
I had always imagined a fair degree ofg home automation. (not planned to be a real world thing in MY home), just as you suggested.
Attic and exterior temperatures, and interior temps are measured. Then an attic fan can be used to suck out hot air from the house to be replaced by cooler exterior air.
Then as the outside temp rises, the attic fan shuts off and the windows close up.
What I _do_ envisions as a real possibility is to have a very smart thermostat. It can learn the speed of interior temp rise and fall based on the outside temp and the time of day (or sunlight strength), then it can project the interior temps into the future and turn the heat/AC on or off so that the peak and minumum temperatures always fall within a specified range.
I will check the weather underground.
I was thinking about getting one of the Davis systems, and then trying to read the wireless signal directly from a 900Mhz receiver to the ZX.
Can you believe that the Davis solar radiatios sensor costs $140!!!! It refits a few percent a year, so it is only a relative measurement pretty quickly.
I am pretty sure I could come up with a nice photodiode circuit for 1/10th the cost!
-Tony
Attic and exterior temperatures, and interior temps are measured. Then an attic fan can be used to suck out hot air from the house to be replaced by cooler exterior air.
Then as the outside temp rises, the attic fan shuts off and the windows close up.
What I _do_ envisions as a real possibility is to have a very smart thermostat. It can learn the speed of interior temp rise and fall based on the outside temp and the time of day (or sunlight strength), then it can project the interior temps into the future and turn the heat/AC on or off so that the peak and minumum temperatures always fall within a specified range.
I will check the weather underground.
I was thinking about getting one of the Davis systems, and then trying to read the wireless signal directly from a 900Mhz receiver to the ZX.
Can you believe that the Davis solar radiatios sensor costs $140!!!! It refits a few percent a year, so it is only a relative measurement pretty quickly.
I am pretty sure I could come up with a nice photodiode circuit for 1/10th the cost!
-Tony
Davis is premium grade!
The LaCrosse and Oregon units have a serial port data interface. The WS3600 that I have is not like the WS2300 and othera, in that the TD and RD lines aren't used (!). The weather station and host (PC) toggle DTR/RTS/CTS/DCD to send data in a clocked serial manner. I think they did this because the WS3600's microprocessor had no serial port.
Anyway, the open3600 software is C source and it shows how to talk to all the LaCrosse units. So that's what I use. You could do that with a microprocessor instead of a PC - which is what I've wanted to get up the gumption to do for some time. Have that micro on the LAN and Internet. But since I have a home automation/web server/ftp server and TV guide PC on 24/7 anyway, may as well host the software on it.
The LaCrosse and Oregon units have a serial port data interface. The WS3600 that I have is not like the WS2300 and othera, in that the TD and RD lines aren't used (!). The weather station and host (PC) toggle DTR/RTS/CTS/DCD to send data in a clocked serial manner. I think they did this because the WS3600's microprocessor had no serial port.
Anyway, the open3600 software is C source and it shows how to talk to all the LaCrosse units. So that's what I use. You could do that with a microprocessor instead of a PC - which is what I've wanted to get up the gumption to do for some time. Have that micro on the LAN and Internet. But since I have a home automation/web server/ftp server and TV guide PC on 24/7 anyway, may as well host the software on it.
-
weatherman
- Posts: 4
- Joined: 10 January 2009, 15:26 PM
I discovered the zbasic web site today while looking for information on how to work around RAM limitations on the BX24. I have built a solar powered, wireless weather station using the BX24 that works very well. However, I am looking forward to enhancements that will be possible with the ZX-24a.
Sensors on my current build include:
3 temperature probes (thermistors)
barometer (MPXAZ4115A6U)
humidity sensor (Humirel)
Tipping bucket rain gauge (Rainwise)
anemometer and wind direction indicator (Davis)
solar radiation sensor (scratch build)
Based on what I am seeing in data from other personal weather stations on the the weather underground web site and in the published specs from Davis and others, the home brew station performs right up there with the best of them.
I would be very interested in comparing notes with others who have attempted such a project.
Sensors on my current build include:
3 temperature probes (thermistors)
barometer (MPXAZ4115A6U)
humidity sensor (Humirel)
Tipping bucket rain gauge (Rainwise)
anemometer and wind direction indicator (Davis)
solar radiation sensor (scratch build)
Based on what I am seeing in data from other personal weather stations on the the weather underground web site and in the published specs from Davis and others, the home brew station performs right up there with the best of them.
I would be very interested in comparing notes with others who have attempted such a project.
I was sconsidering such a thing, but then I went with a commercial station (Davis). I did, however, build my own solar sensor for the station since the Davis sensor is over $100 just for the sensor!weatherman wrote:I would be very interested in comparing notes with others who have attempted such a project.
What method(s) do you use to communicate from the sensors to the CPU? I2C, One-wire, analog?
-Tony
-
weatherman
- Posts: 4
- Joined: 10 January 2009, 15:26 PM
Davis makes an excellent product and would also be my choice in a commercial product under $1500.spamiam wrote:I was sconsidering such a thing, but then I went with a commercial station (Davis). I did, however, build my own solar sensor for the station since the Davis sensor is over $100 just for the sensor!weatherman wrote:I would be very interested in comparing notes with others who have attempted such a project.
What method(s) do you use to communicate from the sensors to the CPU? I2C, One-wire, analog?
-Tony
Sensor communication - Sensor to BX24 (soon to be ZX24):
I use multitasking with wind speed and rain gauge running under their own dedicated task responding/reporting in real time as the events occur, while all other sensors are polled at intervals in a separate task.
1. Anemometer - magnetic switch closure is detected with the available interrupt on pin 11 (WaitForInterrupt)
2. Rain Gauge tip is detected using button push style of logic. The duration of the tip (switch closure) is extended to ~0.75 sec using a 555 one shot circuit.
3. All other sensors are analog and are read using either the on board ADC or external ADC. For the latter I use a Microchip MCP3304 (12 bit, 8 channel). Analog signals are buffered through a unity gain OP AMP circuit as the MCP3304 prefers low impedance signals.
Wireless Communication - BX24 to PC.
I use the Maxstream XCite modules on both ends. This makes for very simple and reliable wireless communication as the pair constitute a wireless serial port, eliminating any need to deal with the complexities of RF protocols and such. I choose to send mostly send raw analog data to the PC for for processing in my own program and posting to the weather underground http://www.wunderground.com/weatherstat ... =KKSSTILW2.
I would like to add a good soil moisture sensor. The gypsum block types are inexpensive but provide only qualitative data. I would love some suggestions.
-
weatherman
- Posts: 4
- Joined: 10 January 2009, 15:26 PM
Would you mind sharing some details on your solar sensor?spamiam wrote:I was sconsidering such a thing, but then I went with a commercial station (Davis). I did, however, build my own solar sensor for the station
-Tony
My own design was inspired from the design I found at http://www.pages.drexel.edu/~brooksdr/D ... ometer.htm. I used the same Digikey PDB-C139-ND photodiode but used a simple transimpedance amplifier circuit to boost the voltage to a value that could be used by my ADC.
The biggest challenge was with the physical design in mounting the teflon difusser that would withstand constant exposure to the elements.
How did you calibrate your unit? As with all calibrations one really needs a reference standard. There were a couple of PWS with Davis sensors in my area already reporting to the weatherunderground I cheated and used one them as my reference standard. These two Davis stations report significanly different values but my home brew sensor has remained in sync with the one I used for calibration for almost a year now.
I also used a transimpedance amplifier. I used an MCP602 opamp. I forget what photodiode I used. It is one from siliconix as I recall.
I make a small pcb from a radioshack solderable breadboard pcb. I mounted it in a small case. I used a translucent rod to conduct the light into the box directly in contact with the photo diode.
I was not very interested in the W/sq. meter so the sine response was not necessary. I was more interested in the "brightness" overall, from any direction. So the translucent rod was cut so it would stick out as high as its diameter.
Overall it works well. I did not "calibrate" it. I just adjusted the output so that I could get as high a reading on the Davis base station as possible without saturating the reading. I can't tell if the sensor is saturating or if the brightness stays pretty constant for hours per day in the summer.
-Tony
I make a small pcb from a radioshack solderable breadboard pcb. I mounted it in a small case. I used a translucent rod to conduct the light into the box directly in contact with the photo diode.
I was not very interested in the W/sq. meter so the sine response was not necessary. I was more interested in the "brightness" overall, from any direction. So the translucent rod was cut so it would stick out as high as its diameter.
Overall it works well. I did not "calibrate" it. I just adjusted the output so that I could get as high a reading on the Davis base station as possible without saturating the reading. I can't tell if the sensor is saturating or if the brightness stays pretty constant for hours per day in the summer.
-Tony