Weather Station Project #27 - Took some samples through material

This morning I was out in the yard setting up an experiment.  I connected the Solar Radiation/UV Sensor setup to a POE enabled cable connected to a Ethernet over Power Line adapter.  The other Power Line adapter I plugged into my computer downstairs in the computer room.  This adapter was plugged into a Pers VLAN connector on one of my GS108Ts.  By doing this, I was able to connect to the sensor setup using my laptop on the porch through wi-fi.  I am going to leave the connection downstairs so that I will be able to use the Power Line adapter in other VLANS as necessary.  It was neat to get it working.  Here is a picture of the parts on my porch table.

The values of the data that I took was as follows:

results with no cover in bright sunlight

{ "tslLUX":"65536.00", "uvVIS":"1677.00", "uvIR":"13354.00", "uvUVRaw":"761.00", "uvUV":"7.61" }

reading from site was 12.0 for UV

Kingstowne says 7.4 UV, 835 watts/m^2

results with plastic cover in bright sunlight

{ "tslLUX":"65536.00", "uvVIS":"1594.00", "uvIR":"12876.00", "uvUVRaw":"716.00", "uvUV":"7.16" }

results with polycarbonate in bright sunlight

{ "tslLUX":"65536.00", "uvVIS":"1600.00", "uvIR":"13228.00", "uvUVRaw":"720.00", "uvUV":"7.20" }

Kingstowne says 6.9 UV, 796 watts/m^2

results with soda glass in bright sunlight

{ "tslLUX":"65536.00", "uvVIS":"1627.00", "uvIR":"12333.00", "uvUVRaw":"734.00", "uvUV":"7.34" }

Kingstowne says 7.2 UV, 833 watts/m^2

results in the shade with no cover

{ "tslLUX":"2152.00", "uvVIS":"312.00", "uvIR":"962.00", "uvUVRaw":"30.00", "uvUV":"0.30" }

{ "tslLUX":"2168.00", "uvVIS":"315.00", "uvIR":"976.00", "uvUVRaw":"31.00", "uvUV":"0.31" }

Kingstowne says 6.7 UV, 817 watts/m^2

This UV data shows a loss of about 5 and a half percent in the visible spectrum, 4 and a half percent considering the IR spectrum.  The overall percent difference between my local site and the site that I am using for calibration appears to be about 2.8 percent; my value is higher.  The distance between us is about 4.13 miles so I have been assuming that the values should be about the same.

According to this data, I should not have a problem if I used the lid to my sandwich container (plastic cover in the test).  I just have to compensate for the loss; there doesn't appear to be a loss associated with the UV sensitivity as far as I can tell.  So I should multiply my UV readings by 1.0628 to get the same as the Kingstowne values?  I should multiply my uvVIS values by 0.4979 to get the watts/m^2 equivalent to the values from the site where I am comparing my values.

Weather Station Project #26 - Some more thoughts

After looking through the raw data from my sensor, I discovered a number of points where the data was "pegged".  The only thing that I can think of is that the integration time on the sensor was set too high.  This is going to require me to redo the sketch that was loaded onto the Arduino.  Thinking more about this, I should also attempt to get the final packaging completed for the sensor setup.  This means a package that is waterproof and uses POE to power it so that I have but one cable from the POE source to the sensor package.  I already have the packaging and I have a few POE splitters on order that will work well with this setup.  Now I need to come up with the final specifics.

The list of parts (from here) with some modifications are:

• Raspberry Pi B+
• Arduino Uno
• Arduino Ethernet Shield
• Arduino Proto Board Shield
• TP-Link TL-POE10R PoE Splitter w/power connector cable
• Power splitter cable (1 female to 4 males - 5.5mm x 2.1mm)
• Power adapter plug (1 female 5.5mm x 2.1mm to ISO block)
• Power plug 5.5mm x 2.1mm to micro USB cable
• Ethernet socket adapter (for mounting on wall of sandwich box)
• Clear sandwich box (see-through for RPi camera)
• RTV to pot up the holes so that water can't get in
• Mounting platform and hardware to mount sandwich box on outside

The Raspberry Pi is used to sample from the Arduino.  I have proven that I can set up a web service on the Arduino, so accessing it from within my network is a no brainer.  I have also proven that I can write a python script which is able to sample both from the Arduino based sensor set and from a Weather Underground station with the sensors that I am trying to emulate.  Ultimately the extra sensor base will include:

• Lightning sensor 
• Moisture sensor (Transpiration) 
• UV monitoring sensor
• Solar radiation sensor

I have the parts for the Moisture sensor but I am missing the lightning sensor parts.  Not sure where to find that.

Update: I got in some POE splitters yesterday so I will probably be building up the package over this weekend.  Right now I am just going to use the plastic lid.  Will need to get a crystal glass lid so that it will pass UVA and UVB through.  I don't think that UVB will be able to pass through the lid at this point.