| Sensor Modifications | |
| The
Temperature Sensor is
provided with a small and inefficient Stevenson Screen in an attempt to stop heating from direct sunshine. A larger screen is much better and further improvement can be achieved with a small fan to produce a slight air-flow over the sensor unit. Result - a proper temperature reading of "in the shade" temperature. The Rain Gauge supplied has a small shallow funnel allowing rain that falls on the funnel to be blown out again in a strong wind or to dry off before enough collects to register. A larger, steep sided funnel not only helps rain water stay in the gauge but also provides a greater rainfall sensitivity and resolution. The funnel shown increases the rain resolution by about 10 times. The new resolution must be set in the calibration part of the weather station software. |
| Stevenson Screen |
|
 |
 |
| My first attempt at a better screen (see Archive) worked reasonably well but I was not quite satisfied and decided to build a more traditional Stevenson Screen. The main frame is wood with plastic angle for the loovres and inverted plastic picnic trays to form the roof. This way was a lot less work than the traditional all wooden construction. The screen was painted brilliant white on the outside to reflect the sunlight and matt black inside as per the latest Stevenson Screen specification. The "box" was then mounted on a fence stake at about 5ft off the ground. |
|
| Rain Gauge Improvements |
 |
 |
| This is a further improvement on the first modification (see Archive) with a larger funnel and improved thermostatically controlled heater to combat frost and allow snow to melt and be counted. The original rain gauge was screwed to a piece of wood. Above this is a large (10 inch diameter) funnel with a bit cut off the bottom to allow for it to sit on top of the original rain gauge. Two holes were cut in the base board below the tipping bucket ends to allow discharge of water. These were covered with a fine plastic mesh (as used for fly screens) to keep insects and spiders out. The mesh was attached to the wood with hot melt glue. The outer casing is a large plastic bucket which used to contain cattle lick and obtained from a local farmer. This is used to support the funnel and shelter everything from the elements. Most of the bottom was cut out leaving a narrow ledge on which to rest the funnel. The inverted bucket was secured to the base board with plastic blocks and screws. The bucket was lined with pieces of polystyrene packaging to provide insulation and reduce heat loss from a thermostatic heater used to prevent freezing up in the winter. The ceramic resistor(s) used as heater(s) and the thermostat circuit were screwed to the base board. To allow for the increased rain sensitivity, a correction factor about 9x must be applied in the software settings. Calibration against a manual rain gauge shows that each bucket tip now equates to 0.036mm of rainfall instead of the original 0.3mm. Theoretically, the correction factor is the ratio of the original area presented to the rain, to the new area. The original rain gauge is 55 sq.cm being a rectangle of 11 x 5 cm. This large funnel is 25cm diameter giving an area of 491 sq.cm, about 8.9 times the original. This compares with 8.3 measured against the manual gauge, so the original gauge was quite accurate (within the accuracy of the manual rain gauge). |