UM-975-001-WSD1 Manual –V2.1 9 | P a g e
Speed component of the sensor (anemometer)
The use of this design also allows the anemometer to be used in circuits down to zero voltage and
current, without reducing the life expectancy of the reed switch.
In the circuit diagram shown in Figure 1 a 100 ohm resistor is fitted into the wiring. This is because in
long cable runs the capacitance between conductors is appreciable. When the switch closes the
capacitance is discharged across the contacts; without the resistor fitted this could lead to shortening
of switch life and generation of transients in the other wires. Although only needed in longer cable
runs, this resistor is fitted as standard within the sensor.
Direction component of the sensor (vane)
This is a low-torque 1k ohm potentiometer design using a 3mm shaft and associated bearings (3 off).
In the circuit diagram shown in Figure 1 a 100k ohm pull down resistor is fitted into the wiring in the
sensor. This is fitted to remove the gap at north by tying the wiper (T2) to the low terminal (T3). The
analogue inputs on some loggers will float, giving odd readings when there is no input. This value of
resistor should eliminate this. Linearity is slightly affected, however, this is only 0.01% which can
usually be ignored due to the potentiometer's 0.5% linearity.
General Cable information
Cable supplied as standard is 3 metres in length and may be shortened or lengthened as required. If
the cable is lengthened, please ensure a good quality environmental connector, or suitable waterproof
junction box, is used. Extension cables used must be of a similar specification; however, separate
cables for the speed and the direction components could be used. The cable is a Belden 9503 or
equivalent. The conductors are 7/0.2mm (24awg) stranded tinned copper, twisted pairs (3 pairs).
Long cable runs on the sensor
There are no problems when long cable runs are used. However, with the direction element of the
sensor the longer the cable is, the longer the signal rise time and the longer it is before the input has
settled. This error is greater for the larger angles of wind direction. If the logger allows for this (for
example, as Campbell Scientific loggers do) then use a delay to eliminate this; 10 mS is usually enough
for most situations, but this does depend on the logger used.
Wind direction averaging
If a simple averaging system is used on wind-direction a problem arises around the North direction.
Example: If 355oequals a 2V full scale reading and 0oequals a 0V output.
If the wind is hovering either side of North, and a number of readings are obtained, some would be
around 2V and some around 0V. If these were averaged, a result of around 1V would be obtained,
equivalent to South.
To solve this problem, it is necessary to use a vectoring system to produce the correct results.
