Author Topic: Active-Passive temperature and humidity sensor.  (Read 3861 times)

0 Members and 1 Guest are viewing this topic.

Offline rossw

  • Senior Moderator
  • Hero Member
  • *******
  • Posts: 878
  • Karma: +35/-0
  • Grumpy-old-Unix-Admin
Active-Passive temperature and humidity sensor.
« on: March 04, 2012, 01:43:27 am »
Having recently moved in to my new place, it was time to get a temporary weather station up.
The toy commercial unit I got has a silly temperature/humidity probe, but putting that out in the sun... seriously???

So here's the explanation of my idea for an "active" self-ventilated housing for the temperature and humidity sensor.

1. There is no "international standard" for radiation shields for meterological use.
2. Stevenson Screens, the old standard, are horribly difficult to make, very labour
intensive to maintain, and tests show results up to 1.7 deg K higher than other
screens implying significant self-heating in such enclosures.
3. "ventilated" (either naturally or artificially) sensors seem to be gaining acceptance.

So, what I've tried to create, based on my old design of 25 years ago, is a sensor inside
a vertical tube made of a material that doesn't conduct heat very well. (In this case,
40mm stabilised PVC pipe).

This tube is coated on the outside with a reflective and low emmissivity surface (polished aluminium foil) glued all 'round with some water-based, neutral-cure sealer, the foil then burnished and polished as far as practical.

This tube has a 45 degree bend at the bottom to prevent reflected radiation from the ground around it striking the sensor directly. There is another bend at the top, similarly to prevent sun shining down the tube onto the sensor. There is a length of
tube on an angle upwards and away from here, and another 45 degree bend back to vertical.
(We don't want wind blowing down here and forcing hot air over the sensor! That's why the sensing tube returns to vertical, rather than horizontal off to the side, where it would either act as a venturi and suck air out (that would be ok) but the opposite effect if wind was comming from that direction, and facing down would result in hot air trapped
in the sensing tube.)

The top bend, and tube, are both painted black so any solar radiation will produce local heating at the top of this tube only, and create an upward draft, drawing real "ambient" temperature air in from the bottom of the tube and over the sensor which is located about 1/3 the way up the tube, resulting in more accurate readings.

The sensor:

held centred in the vertical tube by 3 screws through the inner tube. (I couldn't find my nylon screws, so will replace these later).

The outer tube (solar shield) is also lined with a reflective, low-emmissivity coating, again, the polished aluminium. Low emmissivity is desired here, because we don't want to radiate heat from the inside of the shield to the sensor tube. The shield is made from 100mm sewer pipe, although a thinner material with less mass would be preferable.

The inner tube is held in position by six screws through the outer shield. 3 at the top, 3 at the bottom. (The easiest way to mark the 120 degree points is to wrap some wire or string around the tube, mark the circumfrence. Then, straighten the string, measure it and mark 1/3 and 2/3 along its length. Wrap it around the tube again, and drill at the 3 points). I will probably replace these screws with nylon screws or standoffs later too.

The outside hasn't been coated yet, but will be painted white. We want something that is highly reflective, but that has a high emmissivity this time, so it can re-radiate the energy it receives.

The whole assembly is then mounted to a post, set vertical with the tube inlet at about 1.5 metres above the ground.