Welcome to the new home of the Cumulus Support forum.

Latest Cumulus release v1.9.4 (build 1099) - Nov 28 2014
Latest Cumulus MX release - v3.0.0 build 3043 Jan 20 2017. See the Wiki for download

Homebrew Davis Solar sensor

For discussion of DIY weather equipment - sensors, accessories, improvements to existing kit etc
User avatar
mcrossley
Posts: 5210
Joined: Thu 07 Jan 2010 9:44 pm
Weather Station: Davis VP2
Operating System: Stretch Lite rPi
Location: Wilmslow, Cheshire, UK
Contact:

Homebrew Davis Solar sensor

Post by mcrossley » Tue 16 Apr 2013 2:36 pm

I am in the final stages of making/calibrating a home made solar sensor for the Davis Vantage Pro2. It is now in its final housing, and propped between a couple of bricks on my shed roof whilst I check the calibration.

Quite easy to make, and it cost maybe £10 - and that gives me enough bits to make a second spare circuit board - the weather proof housing was the most expensive bit at about £4 - and a third cct board to 'play around' with UV.

The current graph can be found here, click on the solar button. (also on the 'gauges' and 'now' pages)

Note that the theoretical value has stepped down this afternoon, I did have the transmission factor set at 0.85 as part of the tinkering, I reset it today back to the default 0.8.

There are lots of white clouds around today, so in theory the sensor should be over-reading, but it seems to be hovering around the max in full Sun, so maybe it will need tweaking up a bit.

Edit: Actually the value seemed to jump up as the Sun went in and out of clouds. Trees started shading the sensor from direct Sun around 16:30 :(

User avatar
Werk_AG
Posts: 182
Joined: Sun 13 Jan 2013 8:04 pm
Weather Station: WeatherDuino Pro2
Operating System: Windows 7
Location: Cercal CDV - Portugal

Re: Homebrew Davis Solar sensor

Post by Werk_AG » Tue 16 Apr 2013 4:23 pm

Very interesting. What type of sensor used? Can you give more details about the hardware?

Thanks
Image

RayProudfoot
Posts: 2620
Joined: Wed 06 May 2009 6:29 pm
Weather Station: Davis VP2 with Daytime FARS
Operating System: Windows XP SP3
Location: Cheadle Hulme, Cheshire, England
Contact:

Re: Homebrew Davis Solar sensor

Post by RayProudfoot » Tue 16 Apr 2013 6:24 pm

Well done Mark. You are very resourceful and practical. The exact opposite of me! :lol:

Any possibility I could pay a visit sometime and have a gander? I'm on leave next week but retire from the MoD mid-May. Any time that's convenient to you really.
Cheers,
Ray, Cheshire.

Image

User avatar
mcrossley
Posts: 5210
Joined: Thu 07 Jan 2010 9:44 pm
Weather Station: Davis VP2
Operating System: Stretch Lite rPi
Location: Wilmslow, Cheshire, UK
Contact:

Re: Homebrew Davis Solar sensor

Post by mcrossley » Tue 16 Apr 2013 8:44 pm

Werk_AG wrote:Very interesting. What type of sensor used? Can you give more details about the hardware?

Thanks
Just a photo-diode and an op-amp (well two op-amps actually, but only one is required). The diode I used is the BPW34, but it is a bit too large really, the current draw of my sensor is a little more than the Davis - but only significantly (10's of mWs) when the sun is shining, so the solar charger is working at its best anyway.

I post up some details when I'm done.
RayProudfoot wrote:Well done Mark. You are very resourceful and practical. The exact opposite of me! :lol:

Any possibility I could pay a visit sometime and have a gander? I'm on leave next week but retire from the MoD mid-May. Any time that's convenient to you really.
Thanks Ray, I'll PM you with my whereabouts over the coming weeks.

Edit: I noticed the sensor read a little high just before sunset, so I've knocked it down a bit tonight, lets hope its sunny again tomorrow :)

User avatar
Werk_AG
Posts: 182
Joined: Sun 13 Jan 2013 8:04 pm
Weather Station: WeatherDuino Pro2
Operating System: Windows 7
Location: Cercal CDV - Portugal

Re: Homebrew Davis Solar sensor

Post by Werk_AG » Wed 17 Apr 2013 4:04 am

mcrossley wrote:Just a photo-diode and an op-amp (well two op-amps actually, but only one is required). The diode I used is the BPW34, but it is a bit too large really, the current draw of my sensor is a little more than the Davis - but only significantly (10's of mWs) when the sun is shining, so the solar charger is working at its best anyway.

I post up some details when I'm done.
I'm also trying to do a Solar Sensor and a UVI meter.
For now I'm only doing tests with one UV sensor, connected to an OP and read by an Arduino.
For Solar Sensor I'm planing to use a Light Sensor. So, any info will be helpfull.

Thanks
Image

User avatar
mcrossley
Posts: 5210
Joined: Thu 07 Jan 2010 9:44 pm
Weather Station: Davis VP2
Operating System: Stretch Lite rPi
Location: Wilmslow, Cheshire, UK
Contact:

Re: Homebrew Davis Solar sensor

Post by mcrossley » Fri 24 May 2013 10:26 am

I finally have the UV-I sensor built and connected to the Davis SIM. It is still being calibrated, but I need some solid sun shine to do that - I expect the calibration to be complete sometime in 2015 :lol:

The readings are on my 'gauges' page and 'recent' graphs.

I have used a Reyax UVI-01 diode in the sensor. My sample seems to be way over reading according to the spec., and I'm not sure why - I'm in the process of winding the gain down in the circuit down day by day!

Credit again goes to Steve on the WD forum for providing the guidance on the electronics side

User avatar
Werk_AG
Posts: 182
Joined: Sun 13 Jan 2013 8:04 pm
Weather Station: WeatherDuino Pro2
Operating System: Windows 7
Location: Cercal CDV - Portugal

Re: Homebrew Davis Solar sensor

Post by Werk_AG » Fri 24 May 2013 11:50 pm

I'm also using this sensor and also get over readings. It easily reaches UVI 10 two 12 in not very bright days.

In my case I think the resistor I used, can not have exactly 1M, but a little more, which causes a significant increase in output voltage.
For amp I used a single OP from an AD822, adjusted to output 1,1V (Arduino Internal VRef) with an input of 60mV.

Actually I'm testing another UV sensor (UVM-30A) wich seems to give more realistic readings, and it's more easy to read with an Arduino. It's output is already between 0 and a litle more than 1V, and works with 3 to 5 Volt (it´s like a module).
The gauges (your script, :D :D ) in my page actualy use values from this sensor.

The small one on top is the Reyax UVI-01, at bottom is the UVM-30A.

Image

Image
Image

User avatar
mcrossley
Posts: 5210
Joined: Thu 07 Jan 2010 9:44 pm
Weather Station: Davis VP2
Operating System: Stretch Lite rPi
Location: Wilmslow, Cheshire, UK
Contact:

Re: Homebrew Davis Solar sensor

Post by mcrossley » Sat 25 May 2013 10:03 am

I am using the UVI-01 into a 3M ohm resistor, we reckon the output is at least double what the spec. sheet says. According to the spec it should require a 10:1 gain for the Davis, but in practice that has proved to be way too much. It is sunny today, so I should be able to get the calibration somewhere near around solar noon, then this evening I'll take the sensor down and measure the actual gain used.

User avatar
Werk_AG
Posts: 182
Joined: Sun 13 Jan 2013 8:04 pm
Weather Station: WeatherDuino Pro2
Operating System: Windows 7
Location: Cercal CDV - Portugal

Re: Homebrew Davis Solar sensor

Post by Werk_AG » Sun 26 May 2013 12:34 am

Interesting you be using this sensor with a Davis ...
I wish you success with the calibration
Image

TgT
Posts: 127
Joined: Thu 26 Jun 2008 5:51 am
Weather Station: Davis VP2 + Solar
Operating System: A20 TV box + CumulusMX
Location: Slov.Konjice, Slovenia
Contact:

Re: Homebrew Davis Solar sensor

Post by TgT » Sun 26 May 2013 11:16 pm

IIRC he just needs to send the correct voltage data to an ISS. :roll:
Image

User avatar
mcrossley
Posts: 5210
Joined: Thu 07 Jan 2010 9:44 pm
Weather Station: Davis VP2
Operating System: Stretch Lite rPi
Location: Wilmslow, Cheshire, UK
Contact:

Re: Homebrew Davis Solar sensor

Post by mcrossley » Mon 27 May 2013 8:16 am

Correct, the solar and UV interfaces to the SIM are just analogue voltage levels, so a simple circuit to provide the correct voltage range is all that is required.

After yesterday I am happy that the UV reading is somewhere close, it peaked at 6.6 which I turned down to 6.4. The forecasts in general were saying 6 (integer forecasts) or 6.2, I'd rather it over read slightly than under read as it is supposed to be safety advise.

Interesting though that the UK met office forecast is always lower than everyone else's, yesterday they had 5 predicted.

User avatar
mcrossley
Posts: 5210
Joined: Thu 07 Jan 2010 9:44 pm
Weather Station: Davis VP2
Operating System: Stretch Lite rPi
Location: Wilmslow, Cheshire, UK
Contact:

Re: Homebrew Davis Solar sensor

Post by mcrossley » Tue 04 Jun 2013 8:00 pm

My first full day of Sunshine. The UV and Solar sensors seem to be working OK. They are shaded in the current 'accessible' position before ~7am and after 6:30pm.
Solar.png
Interesting that the UV sensor shows some 'steps' in the curve, I think this may be due to the flat topped cylindrical package they use. The value steps up as the Sun elevation increases and allows it to directly hit the diode junction through the top rather than the side. Or it may be that the diode needs to poke through the housing body a little more.

The Solar sensor is perhaps reading a little low at around 85% of theoretical. It also over-reads at low Sun angles, I think this is because I am using a transparent bodied diode and an over size diffuser.
Solar_percent.png
So, on to a mark II solar sensor I think, with a smaller opaque bodied diode...
You do not have the required permissions to view the files attached to this post.

User avatar
Werk_AG
Posts: 182
Joined: Sun 13 Jan 2013 8:04 pm
Weather Station: WeatherDuino Pro2
Operating System: Windows 7
Location: Cercal CDV - Portugal

Re: Homebrew Davis Solar sensor

Post by Werk_AG » Tue 04 Jun 2013 8:41 pm

mcrossley wrote:Correct, the solar and UV interfaces to the SIM are just analogue voltage levels, so a simple circuit to provide the correct voltage range is all that is required.
Just for curiosity, for UV and Solar Radiation, which are the maximum input voltage, and reading values ​​that match?

Seems that your work with the home made sensors is going well.
Image

User avatar
mcrossley
Posts: 5210
Joined: Thu 07 Jan 2010 9:44 pm
Weather Station: Davis VP2
Operating System: Stretch Lite rPi
Location: Wilmslow, Cheshire, UK
Contact:

Re: Homebrew Davis Solar sensor

Post by mcrossley » Tue 04 Jun 2013 8:52 pm

Here is the interface info...
ISS Solar Input
===============
0V = 0 W/m2
~1.67V = 1000 W/m2
2.977V ~ 1780 W/m2
step per W/m2 = 1.67mV

Input line is diode clamped to 0V and 3.3V, there is a 'switched' 100k pullup to 3.3V, 'protected' with a 100R resistor.
(Requires a rail to rail op-amp)

Samples every 50 seconds for 120ms (power on) - varies between 110-135ms

2 - Yellow - +V
3 - Green - signal
4 - Red - Ground
5 - Black - Ground


ISS UV Input
============
0V = 0 UVI
zero offset = 70mV = 0 UVI
2.395V = 16 UVI
3V = 16 UVI
step per 0.1 UVI = 14.3mV

Input line is diode clamped to 0V and 3.3V, there is a 'switched' 100k pullup to 3.3V, 'protected' with a 100R resistor.
(Requires a rail to rail op-amp)

Samples every 50 seconds for 2 seconds (power on)
Samples every 5 minutes when dark (?)

2 - Yellow - +V
3 - Green - signal
4 - Red - Ground
5 - Black - Ground

User avatar
Werk_AG
Posts: 182
Joined: Sun 13 Jan 2013 8:04 pm
Weather Station: WeatherDuino Pro2
Operating System: Windows 7
Location: Cercal CDV - Portugal

Re: Homebrew Davis Solar sensor

Post by Werk_AG » Tue 04 Jun 2013 9:41 pm

Nice info.
Many thanks Mark

I'm also using rail to rail op-amp (AD822) with my Home Made UV / Solar sensors.

I'm doing:
step 1 - 30 consecutive solar readings at 150ms interval, a software filter excludes the 10 high and 10 low reads, averages the 10 left over (The author of this method calls it "mode filter" and states that it is optimal to remove spurious peaks)
step 2 - do the same for UV.
- repeat step 1 e 2 for 2 times, averaging.
- send values by RF to Arduino receiver at each 23s

Edited: I said 1,5s interval and meant 150ms interval
Image

Post Reply