Accuracy

Signal Characteristics

Knowing the characteristics of the signal under consideration helps you to choose the correct settings to maximize measurement accuracy. Such characteristics include:

• Peak-to-peak value—This parameter, in units of volts, reflects the maximum change in signal voltage. If V is the signal voltage at any given time, then V_pk-pk = V_max – V_min. The peak-to-peak value affects the vertical sensitivity or gain of the input amplifier. If you do not know the peak-to-peak value, start with the largest input range, and decrease it until the waveform is digitized using the maximum dynamic range without clipping the signal. Refer to the specifications for your digitizer for the maximum input range. The following figure shows how different ranges affect the resolution of a 600 mV_pk-pk signal.

  

  

  

• Source impedance—Most digitizers have a 1 MΩ input resistance in the passband. If the source impedance is large, the signal is attenuated at the amplifier input and the measurement is inaccurate. If the source impedance is unknown but suspected to be high, change the attenuation ratio on your probe and acquire data. In addition to the input resistance, all digitizers and probes present some input capacitance in parallel with the resistance. This capacitance can interfere with your measurement in much the same way as the resistance does.
• Input frequency—If your sample rate is less than twice the highest frequency component at the input, the frequency components above half your sample rate will alias in the passband at lower frequencies, indistinguishable from other frequencies in the passband. If the highest frequency of the signal is unknown, you should start with the digitizer maximum sample rate to prevent aliasing and reduce the digitizer sample rate until the display shows either enough cycles of the waveform or the information you need. Refer to the Nyquist Theorem for more information.
• General signal shape—Some signals are easy to capture by ordinary triggering methods. A few iterations on the trigger level may finally render a steady display. This method works for sinusoidal, triangular, square, and saw tooth waves. Some of the more elusive waveforms, such as irregular pulse trains, runt pulses, and transients, may be more difficult to capture. The following figure shows an example of a difficult pulse-train trigger.

  

  Ideally, the trigger event should occur at condition one, but sometimes the instrument may trigger on condition two because the signal crosses the trigger level. You can solve this problem without using complicated signal processing techniques by using trigger holdoff, which lets you specify a time from the trigger event to ignore additional triggers that fall within that time. With an appropriate holdoff value, the waveform shown in the previous figure can be properly captured by discarding conditions two and four.
• Input coupling—On many digitizers, you can configure the input channels to be DC-coupled or AC-coupled. DC coupling allows DC and low-frequency components of a signal to pass through without attenuation. In contrast, AC coupling removes DC offsets and attenuates low-frequency components of a signal. This feature can be exploited to zoom in on AC signals with large DC offsets, such as switching noise on a 12 V power supply. Refer to the specifications for your digitizer for input limits that you must observe regardless of coupling.