circuit was adjusted to produce an output range of to 10 volts with 

 the resistor varied from to 1,000 ohms. Linearity was determined to 

 be 0.1 percent of full scale over this range. 



Tests were also made to determine the effect of temperature on 

 sensitivity and zero drift. A decrease in sensitivity was noted with 

 decreasing temperature of about 0.03 percent of reading per "Celsius over 

 the temperature range of to 24°Celsius. A zero drift of 2 millivolts 

 was also noted over the same temperature range. A +10 percent change in 

 supply voltage from the nominal +^15 volts produced no observable change 

 in output. If we assume an operating temperature range of +5°Celsius, 

 the maximum error in the wave staff electronics due to the combined ef- 

 fects of nonlinearity and sensitivity variations with temperature is 

 +0.2 percent of reading. Since the primary interest is in a dynamic mea- 

 surement of waves, the zero drift noted will have negligible effect on 

 the experiment since temperature variations of any appreciable magnitude 

 will only occur over long periods of time compared to the wave periods. 



Further calibration tests were conducted using actual wave staffs 

 of 1-inch diameter and 20-foot lengths, and 3.5-inch diameter and 8-foot 

 lengths at various depths of immersion in saltwater. These tests were 

 conducted from a dock at Shilshole Bay on Puget Sound. Because of 

 ripples and waves on the water of the order of 1 inch (peak-to-valley) 

 it was difficult to obtain a highly precise measurement. The output 

 was recorded on a strip chart recorder and it was therefore possible to 

 average these variations to some degree. The readout resolution of the 

 strip chart (and accuracy) is about +1/A of a minor division. Full 

 scale across the chart is 50 minor divisions and, thus, the resolution 

 is about 0.5 percent of full scale. Some nonlinearity is noted near full 

 immersion (see calibration curve). Some offset was expected because of 

 the finite resistance of the saltwater path in the ground return which 

 is not taken into account during initial calibration of the wave staff 

 unit. The initial calibration is made with the wave staff on the dock 

 where full scale and zero are set by making actual contact between the 

 ground wire and the wave staff resistance element at the corresponding 

 ends. However, measurements were made of the resistance of the salt- 

 water path to ground in the same location where the wave staffs were 

 immersed and the value of resistance measured (on the order of 10 ohms) 

 does not account for the offset observed at full immersion. In addition, 

 the offset should occur at all readings and it does not. Therefore, it 

 is believed that the nonlinearity observed is a result of some other 

 phenomenon as yet undetermined. Both units produced highest accuracy 

 near center scale with decreasing accuracy toward either end. Overall 

 accuracy including end points is about +^3 percent. If the range of oper- 

 ation is reduced so as not to use the last 1 foot on each end of the 

 wave staff, the accuracy is improved to about +_1 percent. 



The output from the wave staff electronic circuit is fed directly 

 into a voltage to frequency converter; the frequency output is then 

 counted and stored on separate storage registers, once every 50 milli- 

 seconds. If an 8-bit register is used for the wave staff measurement. 



219 



