Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5

Davis solar radiation sensor
#1

   
.pdf vtb 5041.pdf Size: 118,67 KB  Downloads: 438
.pdf vtb8440.pdf Size: 231,87 KB  Downloads: 377
[attachment=242][attachment=243]Hi looking at these solar radiation sensors, I attach a very basic schematic (as far as I can see it.) and a few pictures of the layout of a Davis 7821 sensor. There are a few other things to consider.
a) The diffuser (any ideas on material.?)
b) The double housing which allows an air flow around the sensor inner housing.
Other info:
http://www.davisnet.com/support/weather/...sp?dtype=2
7821
Can this, or similar work on weatherduino?


Attached Files Image(s)
   
Reply
#2

Hi,

Yes, in principle, that circuit diagram could be suitable, if you can fill in the missing information and reproduce and calibrate the hardware adequately. But that might be easier said than done.

Firstly, the Op-Amp is a TLC 2272 with Pin 1 on the right-hand side of the diagram. It seems rather "over-specified" for the task, but not too expensive so you may as well use it if available cheaply (otherwise a normal Op-Amp should be good enough for this simple "shunt feedback, virtual earth input" design).

However, the VTB5041 Photodiode seems very expensive, with a relatively "incomplete" Specification. It is described as having an "extended blue response" but this is not quantified and the peak-response is still in the IR region. Also it has only a +/- 45 degrees half-power beam angle (not the 67 degrees of a "Sine" response) which is presumably the reason for using a "diffuser". Personally, unless more data is available, I would use something like a BPW34 photodiode which has a similar detection area (~7 mm2) but close to a Sine response, and not attempt to construct a diffuser.

Unfortunately, I can't read the colour codes on the resistors in the photo; it looks as if R1 might be 47k or 4k7 but both of these values seem to be (much) too high for the photocurrent delivered by a 7 mm2 photodiode in direct sunlight. So maybe the "diffuser" is also a strong (i.e. low transmission) optical filter? Do you have any more details of the resistor values or the physical appearance of the diffuser? The other component values don't seem too critical, but do use a high stability "cermet" pot.

Cheers, Alan.
Reply
#3

The reading on my website www.ayrshireweather.org is from my old Davis Groweather using sensor 7821 http://davisnet.com/support/weather/supp...sp?dtype=2 and the photos are of an old one of these bought as box of bits some years ago, don't forget this sensor is now at least 15 years old. However Davis are very good and a complete Davis Groweather would have cost at least £1500 in 2000. They are very good stations and mine continues to work with few problems.

However I am going to attemp to build a weatherduino station with the bits and sensors I have collected over the years.

some links of interest:
http://archives.sensorsmag.com/articles/.../index.htm
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3291930/
http://www.eppleylab.com/solar.htm
www.kippzonen.com
http://www.instesre.org/construction/pyr...ometer.htm
The solar sensor in cars is often located in the top part of the dashboard at the base of the windscreen .
[IMG]file:///G:/image/199001482.image[/IMG]
1 - Solar sensor
2 - Output current lighting typical diagram
3 - Diagram showing operation of solar sensor

Just a few more pictures which previous post missed.

Also re cost;

2272 op amp: £2.81 http://www.ebay.co.uk/itm/TEXAS-INSTRUME...5aef81a167
VTB8440 photodiode £5.10 (nearest spec I could find to old version above. http://www.ebay.co.uk/itm/VTB8440BH-Phot...0873564672


Attached Files Image(s)
                       
Reply
#4

Hi,

Thanks for that information which is very interesting. I haven't read through all the links, but IMHO some of that information is not relevant to Meteorological measurements. For example, it appears that car (aircon) sensors are designed to have an increased horizontal response (corresponding to light/heat coming in through the windows), whilst a meteorological sensor has a "Cosine" response (which has a zero horizontal response).

(30-05-2015, 10:14)JT118 Wrote:  VTB8440 photodiode £5.10 (nearest spec I could find to old version above. http://www.ebay.co.uk/itm/VTB8440BH-Phot...0873564672

Beware that that listing is for a "BH" suffix version which includes an IR filter. As far as I'm aware, Meteorological sensors are designed to peak in the (near) Infra Red region because about half of the insolation (watts/m2) is actually received in the Infra Red range.

Cheers, Alan.
Reply
#5

(24-05-2015, 11:47)JT118 Wrote:  Hi looking at these solar radiation sensors, I attach a very basic schematic (as far as I can see it.) and a few pictures of the layout of a Davis 7821 sensor. There are a few other things to consider.
a) The diffuser (any ideas on material.?)
b) The double housing which allows an air flow around the sensor inner housing.
Other info:
http://www.davisnet.com/support/weather/...sp?dtype=2
7821
Can this, or similar work on weatherduino?

The WeatherDuino Solar Radiation input, (pin A0 on the TX board) expects a signal with a voltage between 0V and 1.1V.
The TX unit reads this voltage at ADC0, and send it as an integer value to the RX unit, which in turn will convert it into a value correspondent to a solar radiation between 0 and 1400 W/m2.

I'm using the internal 1.1V voltage as a reference for the ADC0, but changing the code can allow other options.
There is no other math involved. It just rely in the "cosine" response of the solar cell when placed in horizontal position, which may be questionable, but the results are at least very realistic.

Reply
#6

Might be of interest.
.pdf Pyranometer.pdf Size: 243,64 KB  Downloads: 807


+ some detail of Davis Solar Radiation sensor, note "double skin" one out black section and one inner, both mounted on common set of screws and springs which allow the sensor to be levelled using the bubble level and the whole to be mounted clear of the base to allow air flow.


Attached Files Image(s)
   
Reply




Users browsing this thread: 1 Guest(s)