Noise is unwanted signals present on the output channels that can affect devices connected to the output channels.

Noise can be characterized as normal-mode noise or common-mode noise. Regardless of its characterization, noise is meaningful only when it is specified with an associated bandwidth.

  • Common-mode noise—Noise present between the Isolated GND terminal and the chassis or earth ground. In this sense, the equivalent circuit is a current noise source connected across these two terminals. When you connect an impedance between the output common/ground and chassis or earth ground, a noise current can flow in the impedance, resulting in an unexpected offset or other undesirable error.
  • Normal-mode noise—Noise present between the +6, +20, or -20 terminal and its corresponding return terminal (GND or Isolated GND), appearing either in series (constant voltage mode) or parallel (constant current mode) with the output of the device. Normal-mode noise can be expressed as voltage noise or current noise, depending on the control mode of the output channel.

    AC-to-DC rectification causes ripple, a type of periodic normal-mode noise.

    • Verifying Output Noise Specifications

    Exercise care when verifying the noise specifications of an output device, such as a power supply or SMU. When verifying the specified wideband noise of a device, the effects of ground loops, unnecessarily long probe ground leads, and electrically noisy environments can combine and skew your measurements.

    Observe the following recommendations when verifying the output noise specifications of a power supply or SMU:

    • Connect the probe directly to the terminals of the power supply or SMU. Do not use long leads, loose wires, or unshielded cables.
    • Limit the probe ground lead to 2.54 cm (1 in.) at most. Connect this lead directly to the output common/ground terminal of the appropriate channel.
    • Set the bandwidth of the measurement device to the bandwidth of interest.
    • Exercise caution when making measurements in a modern laboratory environment—with computers, electronic ballasts, switching power supplies, and so on—to avoid measuring the environment noise instead of the device noise.

    AC and DC Noise Rejection

    You can manipulate the aperture time of measurements made with SMUs and power supplies to reject specific AC noise frequencies in DC voltage and current measurements.

    Each measurement that an NI-DCPower instrument returns is an average of one or more higher-speed samples. All instruments return a multiple of 50 Hz and 60 Hz to enable rejection of power line noise.

    You can reject AC noise by adjusting the number of samples in a measurement so that the aperture time of the measurement is a multiple of the AC noise period. You can configure the number of samples to average to reject AC noise.

    Rejecting AC Noise in DC Measurements with Measurement Averaging

    The PXI-4110 allows you to adjust the number of samples averaged in each measurement to influence the aperture time of your measurements, which allows you to reject specific AC noise frequencies. You can set the number of samples to average programmatically using the niDCPower Samples To Average property or the NIDCPOWER_ATTR_SAMPLES_TO_AVERAGE attribute.

    The number of samples you choose to average in a measurement implicitly controls the aperture time based on the following relationship:

    Aperture Time = Samples To Average / Maximum Sample Rate

    where

    • Samples To Average is the number of samples, as determined by Samples To Average, that compose each measurement; the default value and valid range of samples to average depend on your instrument
    • Maximum Sample Rate is the maximum sample rate of the instrument as documented in the specifications for your instrument

    The overall hardware measurement rate of the instrument is the reciprocal of the aperture time.

    Complete the following steps to reject AC noise frequencies by manipulating the aperture time of your measurements with measurement averaging.

    1. Based on the frequency f (Hz) you need to reject and the maximum sample rate of your instrument as indicated in the specifications, calculate the number of samples you need to average to reject the frequency.

      Samples To Average = Maximum Sample Rate / f

    2. Set Samples To Average to the value you calculated.
      Tip To improve noise reduction while keeping frequency rejection, you can use integer multiples of Samples To Average, as long as that multiple is a supported value of Samples To Average for your instrument and the longer aperture time that results is appropriate for your application.
      Note If you set the Samples To Average property in the Running state, the output channel measurements may move out of synchronization.

      If this occurs, set Reset Average Before Measurement to True before calling Measure Multiple in your program. You can set Reset Average Before Measurement to False after Measure Multiple runs.

    Determining Measurement Rate

    Although the measurement speed of the PXI-4110 is 3 kS/s for all voltage and current measurements, the measurement rate of the PXI-4110 can vary depending on the setting of the niDCPower Samples To Average property or the NIDCPOWER_ATTR_SAMPLES_TO_AVERAGE attribute.

    The default value of the niDCPower Samples To Average property and the NIDCPOWER_ATTR_SAMPLES_TO_AVERAGE attribute is 10. As expressed in the following equation, the PXI-4110 returns 300 measurements per second using the default value:

    3000 ⁢   samples second × 1 ⁢   measurement 10 ⁢   samples = 300 ⁢   measurements second

    If no measurement averaging is used (Samples To Average = 1), the PXI-4110 returns 3,000 measurements per second.

    While measuring without averaging yields the fastest measurement rate, noise from the environment (for example, the 50 Hz or 60 Hz noise introduced by cabling) increases measurement uncertainty.

    Adjust the niDCPower Samples To Average property or the NIDCPOWER_ATTR_SAMPLES_TO_AVERAGE attribute as necessary to optimize the noise performance and measurement rate for your application.

    Note Measurement rate refers only to the hardware measurement rate and does not include software latency.

    Rejecting Noise

    If you know the noise frequency, you can reject it from the signal. To determine the number of measurements necessary to reject noise from a signal, divide the measurement speed of the PXI-4110 by a full wavelength cycle of noise.

    Table 11. Number of Samples to Average Based on Frequencies Rejected
    Number of Samples to Average Frequencies Rejected for 3 kHz Sample Rate
    1 3 kHz
    50 60 Hz
    60 50 Hz
    300 50 Hz and 60 Hz

    For frequencies not listed in this table, use the following formula:

    N = 3000/Fr

    where N is the number of samples to average

    Fr is the frequency rejected

    Note To improve noise reduction while keeping frequency rejection, set the number of samples to average to N, 2N, 3N, and so on. The maximum allowed samples to average is 511.
    Tip Set the niDCPower Samples To Average property or the NIDCPOWER_ATTR_SAMPLES_TO_AVERAGE attribute to 300, an integer multiple of both 60 kHz and 50 kHz, to actively reject both noise frequencies.