Home built fan aspirated radiation shield

[Bashy]

Yeah, i like the solar setup, does it charge a battery for the night time or is it just Daytime only?

[harrysvr]

Yeah, i like the solar setup, does it charge a battery for the night time or is it just Daytime only?

No, but I wish I could do it...(it`s daytime opereted  )

[whatboy]

If a fan needs to be activated only at Daylight, a PC fan could do the trick, but I wonder if this http://www.walmart.com/catalog/product.do?product_id=3331121 could generate sufficent energy to power a 12 volt dc motor, I've never work with solar panels...


Oh and thanks for the instrucctions Breitling...

[weather boy]

Hello,

i have one question! Can i make same radiation shield? I have sensor from Oregon BAR208 HG

[Dave Manley]

Thought I would include a photo of the shield I have just built and installed. The temperature/humidity sensor is 1.23 metres above grass.

[steve26]

Great work !

[selph]

I know this thread is old but I can not help but comment.

This setup is elegant but the temperature measurements will be much warmer than the true air temperature.

First, the position of the fan. The fan is really only circulating air inside the housing. This air will be warmer than the surrounding air especially because the fan motor generates heat. If the fan position is actually pulling in air from outside, it is pulling air from directly under the top plate which will warm up in the sun across across the fan motor (which is hot) and then across the sensor. Ideally the fan should push/pull fresh air across the sensor first and then out.

Second, a daytime only fan will not solve all heating issues. Low sun angles will more penetrate the housing, heating the sensor above ambient temperature. At these low sun angles, a solar panel may not output enough power to turn on a fan resulting in a temperature higher than the true temperature. At most latitudes this would be after dawn for an hour and before dusk for an hour (depending on time of year too).

[Dave Manley]

Thanks for your comments; I would be interested in your views as to how the design can be modified to incorporate your thoughts.

[niko]

selph: Which particular shield design do these comments apply to?

I know this thread is old but I can not help but comment.

This setup is elegant but the temperature measurements will be much warmer than the true air temperature.

First, the position of the fan. The fan is really only circulating air inside the housing. This air will be warmer than the surrounding air especially because the fan motor generates heat. If the fan position is actually pulling in air from outside, it is pulling air from directly under the top plate which will warm up in the sun across across the fan motor (which is hot) and then across the sensor. Ideally the fan should push/pull fresh air across the sensor first and then out.

Second, a daytime only fan will not solve all heating issues. Low sun angles will more penetrate the housing, heating the sensor above ambient temperature. At these low sun angles, a solar panel may not output enough power to turn on a fan resulting in a temperature higher than the true temperature. At most latitudes this would be after dawn for an hour and before dusk for an hour (depending on time of year too).

[Breitling]

Thanks for your comments, but I think you have to read a bit more carefully:

First, the position of the fan. The fan is really only circulating air inside the housing. This air will be warmer than the surrounding air especially because the fan motor generates heat. If the fan position is actually pulling in air from outside, it is pulling air from directly under the top plate which will warm up in the sun across across the fan motor (which is hot) and then across the sensor. Ideally the fan should push/pull fresh air across the sensor first and then out.

Picture #19: "The motor fits upside down on top of #7 tray, propeller at the bottom. We want to aspirate, not blow. The motor is directly on the air flow, allowing a small cooling effect on summer. All possible hot parts have to be above the sensor and in the air flow."

That means the fan pulls fresh air from the bottom part of the shield across the sensor, and push it outside through the top. Motor heating is almost insignificant (1 volt, low rpm), but anyway it is placed over the sensor, in an ascendent airflow.

Second, a daytime only fan will not solve all heating issues. Low sun angles will more penetrate the housing, heating the sensor above ambient temperature.

Picture #2: "You have to make the central cutting on 6 of them. You may want to make a little test to know the cutting extent: trays have to overlap a bit between them, this avoid the horizontal radiation ("A") tho reach the measuring chamber...."

So no direct radiation over the sensor is possible even at sun angles lower than horizontal (see "B" arrow on scheme). Picture #13 illustrate this point in the real shield.

At these low sun angles, a solar panel may not output enough power to turn on a fan resulting in a temperature higher than the true temperature. At most latitudes this would be after dawn for an hour and before dusk for an hour (depending on time of year too).

Pictures #23 & #24: the sun a bit over the horizont and the motor is easily running, it starts at 3 mA current. Even an overcast day with enough diffuse radiation can start the fan.

[selph]

Radiation shields with slats or louvers, even double louvers, will read higher than ambient temperature, especially at low sun angles and any time of the year, not just in the summer. Even though no direct radiation hits the sensor, the sides warm up in the sun and air must pass over those side walls to cool the sensor. This results in, especially at low sun angles, an error with respect to the actual air temperature. By "higher" I have seen errors as large as 2 Celsius in the middle of winter as the sun rises or sets and over 2 in the middle of the summer. If you don't care about a 2 C error than no big deal.

The company Metone, with whom I have no affiliation, has a very unique design that in my experience is significantly more accurate
Here is a pdf description: http://metone.com/documents/076B%20Fan%20Aspirated%20Radiation%20Shield.pdf

Their design has concentric cylinders which could easily be "home made" by using thin walled PVC pipe.... ensuring that the gaps between the walls is also aspirated by the fan and that there are no paths for reflected radiation to go up into the bottom.

There are two issues with the Metone design: power consumption and size. If you have AC power available then the more air the fan moves the better (that usually translates to more power). If available space is no problem then why not.

[niko]

By "higher" I have seen errors as large as 2 Celsius in the middle of winter as the sun rises or sets and over 2 in the middle of the summer.

What was the source of the correct temperature that you used to determine these errors?

[TokKiwi]

That's why I went to a cheap solution that used an inverted plastic narrow bucket with foam insulation layers and a reflective media on the inside together with a duct out the top and a fan in the duct extension outside.  Gives very sensitive values during calm conditions both during the day and at night -also lower peaks in the height of summer (cf local Davis VPII).  It is out in the open beside fields, mounted in a post over grass.

[skyewright]

Even though no direct radiation hits the sensor, the sides warm up in the sun and air must pass over those side walls...

Those sides are typically painted white to help reduce any warm up.

For my own "plant saucer"  based shield I could only get terracotta coloured saucers. Rather than paint them I use adhesive aluminium tape on the outer surfaces. That provides both a very effective radiation block (some plastics are at least slightly translucent) and a very reflective surface. The shield is now over 3 years old and the tape is still in decent condition even in our maritime climate.

[selph]

Niko,
 great question. I had access to several gill-type shields and several aspirated ones.
I measured several days worth of data and the metone, to which I linked before, was the "coldest" all day, so I assumed it was closest to air temperature... The humidity was low, no irrigation, and always above dew points so there was no evaporative cooling effects. To that I compared everything. Other shield types were always higher than this reference with the gill-types (even self-aspirated) being the worst. By "gill-type" I mean shields that allow wind to blow through them, having slats, louvers or stacked "upside-down pie tins".
Heavy material is bad, it absorbs a little heat all day and then gives that off at night.

Another issue is heat loss at night to clear skies.... being colder than air temperature