The Phidgets Solar Powered Weather Station


Yes, it’s a weather station, one of those things that records data from a suite of sensors for a compact and robust way of logging atmospheric conditions. We’ve seen a few of these built around Raspberry Pis and Arduinos, but not one built with a Phidget SBC, and rarely one that has this much thought put in to a weather logging station.

This weather station is designed to be autonomous, logging data for a week or so until the USB thumb drive containing all the data is taken back to the lab and replaced with a new one. It’s designed to operate in the middle of nowhere, and that means no power. Solar it is, but how big of a solar panel do you need?

That question must be answered by carefully calculating the power budget of the entire station and the battery, the size of the battery, and the worst case scenario for clouds and low light conditions. An amorphous solar cell was chosen for its ability to generate power from low and indirect light sources. This is connected to a 12 Volt, 110 amp hour battery. Heavy and expensive, but overkill is better than being unable to do the job.

Sensors, including temperature, humidity, and an IR temperature sensor were wired up to a Phidgets SBC3 and the coding began. The data are recorded onto a USB thumb drive plugged into the Phidgets board, and the station was visited once a week to retrieve data. This is a far, far simpler solution than figuring out a wireless networking solution, and much better on the power budget.

Via embedded lab

14 thoughts on “The Phidgets Solar Powered Weather Station

  1. “This is a far, far simpler solution than figuring out a wireless networking solution, and much better on the power budget.” I dont think so. Why is it every time it comes to environmental datalogging so impossible to keyboardhack an old e.g. nokia which great battery (or e.g. at-command if it has to be more suffisticated) and turn it every few days on for a bunch of sms (which are also free in charge if you use some in the same net services)??? This is really no great thing and enables researches to place these stations somewhere they could achieve the best measuring results and dont have to compromise the location because of the accessability by the “loggerfarmers”. (see picture station is placed to fr to the trees…)

      1. If you don’t mind quick-n-dirty with MS Windows, the first places to start are to read up on UIView32 and AGWPE. Overkill, to say the least, but it’ll get the job done.

        73 de kd0gzj

  2. The built looks excellent and the write up is great.
    But everything had to be up sized because of the 2W SBC that does nothing most of the time. I mean there are smaller micros with usb host that can take pictures and put them on an usb memory which would consume virtually zero between the measurements….
    Not to mention that the whole concept of getting the data by swapping the card seems rather old fashioned. Is there no mobile internet available there?

    1. Not to mention this is as expensive as it gets for a data logger. $150 a board? Honestly, couldn’t this have been done with simpler, less costly solutions? Don’t tell me it’s a “because we can” project.

  3. Idea: If you hooked a propeller to a generator, could you make an anemometer by measuring the current generated, AND use the power created to run the weather station on cloudy days?

    Kill two birds with one stone!

    1. Yeah it might work… I think the problem is you must then have a tall mast for the wind generator. You also lose sensitivity. An anemometer is supposed to be very sensitive and give accurate readings even for a slight breeze. If you use a wind generator as an anemometer, you are going to get a reading of 0 for any wind lower than, say 10km/h.

  4. yah.. typically when building a system like this, if done well, you would find that the LEDs are becoming your biggest power suck, and you then cut those off too.
    I hate to say “arduino” (actually.. I love the little thing), but there is nothing here that could not be done by one, and with far less power consumption (assuming you use a pro mini, remove the LEDs, log to SD card, and use sleep modes appropriately).

    I am currently building a system which reads wind speed, direction, temp, humidity, light levels.. will add pressure and particulate contamination. If not for wind speed making it hard to keep it in sleep mode for any length of time, Id be using a fraction of a percent of the fidget board.

  5. I built a standalone “quite-less” featured weather station using Renesas GR Sakura a while back. I used a 3w solar panel and a buck conv to charge the Li-Ion battery and a boost conv to run the Sakura board and other electronics! Used mainly PVC for the pole in order to make it easy to deploy! The data for temperature, Panel Voltage, Battery Voltage,Soil humidity and Light intensity were logged onto the SD Card.
    There was also a Web interface over Ethernet routed thru wifi to check the state!

    Have a look :

  6. I would think that panel hanging off the side of the unit (and thus partially shielding several sensors) would render some sensor readings somewhat at variance from reality. Is this a concern here?

    1. From experience of having a similar setup, yet. Having it that close to the vane and anemometer (in any configuration) probably throws off the wind readings.

      OTOH, shielding the thermometer will help with the accuracy somewhat.

  7. IMHO and it is my opinion only, something like an SBC, in this case a Phidget with all the bells and whistles can be a much more stable and reliable system long term than using “experimenter boards”. They each have their place.

    I’m not into the whole “sneaker net” thing so some kind of GSM for interrogating the data would be needed. Kind of the difference between buying your own solar panels vs. making them.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.