in distance from the water surface, and variation in gage output due 

 to changes in the I. F. frequency. Neither of these relationships 

 are linear, but can be shaped to make the output of the gage relative 

 to its distance from the water surface a linear relationship. 



The relationship of I. F. frequency vs. the output of the gage 

 is a curve determined by the frequency response of the gage circuits 

 from the output circuit of the mixer to the detector stage. If 

 this curve can be shaped so that it will compensate for the curvature 

 of the R. F. frequency vs. the distance from the water relationship, 

 the output curve of the gage, as voltage vs. the distance of probe 

 from water, will be a straight line. In Figure 3 a family of the 

 latter curves is shown, in which curve No. 10 is almost a straight 

 line. Each curve shown in Figure 3 represents use of a different 

 frequency from the fixed oscillator. This frequency change of the 

 fixed oscillator changes the voltage output of the gage for a fixed 

 distance above the water surface. This frequency can be regulated by 

 adjusting the small trimmer (marked (1) in Figure 2) in the fixed 

 R. F. oscillator. 



The performance of the gage as presently constructed is satis- 

 factory, however, it is believed that some improvements can be made. 

 The instrument drifts with temperature changes as it warms up but 

 if allowed to warm up for 45 minutes it is very nearly stable. When 

 used in a building where temperature does not change rapidly the 

 gage can be used very satisfactorily but will have to be adjusted 

 occasionally. Calibration drift due to temperature change is shown 

 in Figure 4 to be about 0.067 volt per degree centigrade of temperature 

 change over a range of 0.19 foot. Tuning and adjustment of the gage 

 for linearity must be carefully done; the adjustment that effects 

 linearity the most seems to be in the probe coupling condenser and a 

 larger trimmer in this circuit might help. 



The power supply has a voltage regulating (V.R. 150) tube across 

 the B+ and a constant voltage transformer input to eliminate effects 

 of line voltage changes. 



The probe used in this gage was made from a brass rod 1/16 inch 

 in diameter and 5 inches long, with a circular plate 3/16 inch in 

 diameter soldered to the end next to the water surface and the other 

 end bent in a loop for fastening to a ceramic feed-through. Other 

 sizes of probes could be used but the gage circuits would have to be 

 readjusted. 



To adjust the gage for linearity a family of curves as shown in 

 Figure 3 must first be made. This family of curves is obtained by 

 adjusting the Probe coupling, I.F. transformers, and the trimmer 

 in the fixed R. F. oscillator tank circuit. By varying the tuning 

 and coupling condensers a curve approaching a linear relationship, 

 such as No. 10 on Figure 3 t can be found. This curve then is the 



36 



