Chapter 3 Hardware OverviewDAQCard E Series User Manual3-6 National Instruments CorporationConsiderations for Selecting Input RangesWhich input polarity and range you select depends on the expectedrange of the incoming signal. A large input range can accommodate alarge signal variation but reduces the voltage resolution. Choosing asmaller input range improves the voltage resolution but may result inthe input signal going out of range. For best results, you should matchthe input range as closely as possible to the expected range of the inputsignal. For example, if you are certain the input signal will not benegative (below 0 V), unipolar input polarity is best. However, if thesignal is negative or equal to zero, inaccurate readings will occur if youuse unipolar input polarity.DitherWhen you enable dither, you add approximately 0.5 LSB rms of whiteGaussian noise to the signal to be converted by the ADC. This additionis useful for applications involving averaging to increase the resolutionof your DAQCard, as in calibration or spectral analysis. In suchapplications, noise modulation is decreased and differential linearity isimproved by the addition of dither. When taking DC measurements,such as when checking the DAQCard calibration, you should enabledither and average about 1,000 points to take a single reading. Thisprocess removes the effects of quantization and reduces measurementnoise, resulting in improved resolution. For high-speed applications notinvolving averaging or spectral analysis, you may want to disable thedither to reduce noise. You enable and disable the dither circuitrythrough software.Figure 3-3 illustrates the effect of dither on signal acquisition.Figure 3-3a shows a small (± 4 LSB) sine wave acquired with dither off.The quantization of the ADC is clearly visible. Figure 3-3b shows whathappens when 50 such acquisitions are averaged together; quantizationis still plainly visible. In Figure 3-3c, the sine wave is acquired withdither on. There is a considerable amount of noise visible. Butaveraging about 50 such acquisitions, as shown in Figure 3-3d,eliminates both the added noise and the effects of quantization. Ditherhas the effect of forcing quantization noise to become a zero-meanrandom variable rather than a deterministic function of the input signal.