Shop OBEX P1 Docs P2 Docs Learn Events
Pyranometer — Parallax Forums

Pyranometer

Sal AmmoniacSal Ammoniac Posts: 213
edited 2012-07-01 19:36 in Accessories
I'd like to add a pyranometer to my Propeller-based weather station, but the commercial models I've looked at cost several hundred dollars.

Has anyone built their own out of commonly available parts? I don't need research-grade accuracy, but would like something capable of being calibrated in W/m^2.

Comments

  • Tracy AllenTracy Allen Posts: 6,658
    edited 2010-07-03 01:16
    Just about any small photo diode will make a good wide-band pyranometer. The µA into a short circuit is what you need to monitor to get a linear response proportionaly to W/m^2, so the photodiode should operate at virtual ground into an op-amp configured as an I->V converter. For weather protection, mount the diode inside a box with a light pipe or diffuser. There is some talk about this in the Applied Sensors book.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Tracy Allen
    www.emesystems.com
  • Sal AmmoniacSal Ammoniac Posts: 213
    edited 2010-07-03 01:46
    Thanks. I'll take a look. I thought that perhaps a ping-pong ball cut in half would make a good hemispheric diffuser.
  • ercoerco Posts: 20,255
    edited 2010-07-03 04:26
    @Tracy: Once again, your technical knowledge and instantaneous, helpful reply about something I never heard of is most impressive.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    ·"If you build it, they will come."
  • Tracy AllenTracy Allen Posts: 6,658
    edited 2010-07-03 15:35
    Sal, good idea about the ping-pong ball. It would "color" the response, but maybe in a good way. Pyranometers are supposed to have a wideband response to measure total W/m^2 over the whole visible band and also near IR. Photodiodes peak in the near IR, so the instruments often use a "blue enhanced" diode and a bit of filtering, usually through an acrylic diffuser. Reference pyranometers use a different principle, thermocouples under an absorbing black absorptive coating compared to thermocouples under a white reflective coating, under an evacuated quartz dome.

    Erco, thank you for your kind words! My company is a VAR for LI-COR, one of those companies that makes the expensive pyranometers and other scientific light sensors that are used for assessing solar panel installations, or heat load on buildings, or energy for plant growth, efficiency of lighting, things like that. We make amplifiers for interfacing the sensors to data loggers and SCADA systems. The sensors have µA level signals that need to be conditioned up to Volts or 4-20 mA.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Tracy Allen
    www.emesystems.com
  • kf4ixmkf4ixm Posts: 529
    edited 2010-07-03 16:08
    @Sal
    Just curious, what software do you use for your weather station? Or let me backup, is yours a standalone all in one station whereas the prop does all the calculations and displaying of the data or does it parse the data from the sensors and forward it to a pc where you use software like Weather Display...
    http://www.weather-display.com/index.php

    I have used weather display in the past with one-wire sensors from hobby boards...
    http://www.hobby-boards.com/catalog/index.php?cPath=22&osCsid=2a3c30acba1961201ccc275c72c137b3

    I have been wanting to incorporate a prop into a weather station using one-wire and just curious on your set-up.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    I'm just 7*1027 atoms floating through time and space in close formation. -KF4IXM
  • Tracy AllenTracy Allen Posts: 6,658
    edited 2010-07-03 18:27
    Here is a light sensor I DIY'd, not a pyranometer, but a low-light photometer for correlating the activity of bats with twilight and moonlight. The cap is a gasketed lexan lens made by VCC (from Mouser) and the sensor is the MicroSemi LX1973, which has a photometric voltage output proportional to the 4th root of lux, so that it can cover a large dynamic range.

    ▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
    Tracy Allen
    www.emesystems.com
    moonlight1.jpg
    483 x 361 - 34K
    moonlight2.jpg
    297 x 231 - 17K
  • Sal AmmoniacSal Ammoniac Posts: 213
    edited 2010-07-04 22:08
    @kf4ixm
    The Propeller side of the station, which sits outside in a plastic box inside my wooden instrument shelter, runs software I wrote myself. It collects data from the various sensors and sends the data via an RS-232 link to a netbook computer running a C# program that I wrote to display and log the data. The sensors I currently have are: temperature via an LM34, humidity via an SHT15, and barometric pressure via an SCP1000. I plan to add a pyranometer, anemometer/wind vane, and tipping bucket rain gauge in the near future. I do all math on the Propeller in fixed-point format and do a final conversion to floating point strings for transmission to the indoor box. The current set up is a proof of concept and has worked well.

    Up until now, I've been storing data collected by the C# program in a text file and graphing it by importing it as a CSV file into Excel. I plan to add data plotting and graphing abilities to the "indoor" C# software as soon as I finish the hardware design of the upgraded outdoor component. I'm also thinking of replacing the LM34 sensor with an NTC thermistor. The LM34 accurate to 1 degree F, but I think I can do better with a thermistor. Not that I need great accuracy, but because it sounds like an interesting experiment.


    @Tracy
    Thanks for the description of your bat light sensor. I'll take a look at the LX1973 and the lexan lens and see if I can put something together that meets my needs. Calibrating the final result to W/m^2 will probably be the biggest challenge.
  • legoman132legoman132 Posts: 87
    edited 2010-07-08 02:46
    The Applied sensors text mentions making a pyranometer on page 84 (page 92 of the PDF) using the AD592 (Parallax P/N 28130) in an RC time circuit.

    Post Edited (legoman132) : 7/8/2010 2:51:27 AM GMT
  • sven98desven98de Posts: 8
    edited 2011-05-05 05:03
    Perhaps this could be a solution, unfortunately it's in german:
    http://www.umnicom.de/Elektronik/Projekte/Wetterstation/Sensoren/Pyranometer/Pyranometer.htm
  • JosephineJosephine Posts: 1
    edited 2012-06-15 01:57
    A solution that is free, if basic, is to download this pyranometer App to an iPhone or similar device.

    It works reasonably well for single solar irradiance measurements, which will give you hemisphere measurements in W/m^2.
  • Tracy AllenTracy Allen Posts: 6,658
    edited 2012-06-20 08:31
    That's interesting about the iPhone app from Hukseflux. They manufacture high end scientific light sensors. I'm in making a Prop interface for the Hukseflux NR01 net radiometer, one that will be deployed in the River of No Return wilderness in Idaho. The interface has four channels and measures both incoming and reflected radiation energy, at both short (visible) and long (heat, far infrared) wavelengths. Also there is an internal RTD that monitors the temperature of the instrument itself. The Prop reads the voltages with an 8-channel, 16 bit converter.
  • Prophead100Prophead100 Posts: 192
    edited 2012-07-01 19:36
    I used the TAOS TSL230R on a Propeller project for a similar purpose. It's output can be directly converted to W/M^2. You may have to develop a calibration offset to match the spectra but implementation-wise it was a great way to go. The other thing to consider is the orientation of the sensor and accounting for the cosine effect from the sun' motion. You could ignore it with a flat sensor using a per surface area reading, account for it with a cosine function based on the sun's location or track the sun with a servo or stepper motor. The Prop has objects to help with that aspect and I am sure if you look hard enough there are STAMP versions as well.
Sign In or Register to comment.