Campbell Manufacturing CR10 Wind, Error, Length, L=500 ft, Direction, L=1000 ft, SECTION 13. CR10 MEASUREMENTS, TABLE 13.3-3.Settling Error, in Degrees

SECTION 13. CR10 MEASUREMENTS

FIGURE 13.3-6. Resistive Half Bridge Connected to Single-Ended CR10 Input

Ro, the source resistance, is not constant because Rb varies from 0 to 10 kohms over the 0 to 360 degree wind direction range. The source resistance is given by:

Ro = Rd+(Rb(Rs-Rb+Rf)/(Rs+Rf)) =

Rd+(Rb(20k-Rb)/20k)[13.3-12]

Note that at 360 degrees, Ro is at a maximum of 6k (Rb=10k) and at 0 degrees, Ro is 1k (Rb=0). It follows that settling errors are less at lower direction values.

The value of Rb for any direction D (degrees) is given by:

Rb(kohms) = (10k)(D)/360 [13.3-13]

Equation 13.3-6 can be rewritten to yield the settling error of a rising signal directly in units of degrees.

Error (degrees) = De-t/(Ro(Cf+CwL))[13.3-14]

Equation 13.3-12, -13 and -14 can be combined to estimate the error directly in degrees at various directions and lead lengths (Table 13.3- 3). Constants used in the calculations are given below:

Cf = 3.3nfd

Cw = 41 pfd/ft., Belden #8771 wire

t = 450µs

TABLE 13.3-3. Settling Error, in Degrees,for 024A Wind Direction Sensor vs. LeadLength

Wind	Error
Direction	L=1000 ft.	L=500 ft.
360°	66°	15°
270°	45°	9°
180°	21°	3°
90°	4°	0°

The values in Table 13.3-3 show that significant error occurs at large direction values for leads in excess of 500 feet. Instruction 4, Excite, Delay, and Measure, should be used to eliminate errors in these types of situations. Using a 10 ms delay, settling errors are eliminated up to lengths that exceed the drive capability of the excitation channel (~ 2000 ft.).

13.3.3TRANSIENTS INDUCED BY SWITCHED EXCITATION

Figure 13.3-6 shows a typical half bridge resistive sensor, such as Campbell Scientific's Model 107 Temperature Probe, connected to the CR10. The lead wire is a single-shielded pair, used for conducting the excitation (Vx) and signal (Vs) voltages. When Vx is switched on, a transient is capacitively induced in Vs, the signal voltage. If the peak transient level, Veo, is less than the true signal, Vso, the transient has no effect on the measurement. If Veo is greater than Vso, it must settle to the correct signal voltage to avoid errors.

13-7