Analog Output

If the DUT inputs are floating, use either the pseudodifferential or differential configuration.

If the DUT inputs are grounded or ground referenced, use the differential configuration. Using the pseudodifferential output configuration on a grounded DUT creates more than one ground-reference point. This condition may allow ground loop currents which can introduce errors or noise into the measurement. The 50 Ω or 1 kΩ resistor between the negative input and ground is usually sufficient to reduce these errors to negligible levels, but results can vary depending on your system setup.

Configure the channels based on the signal source reference or DUT configuration. Refer to the following table to determine how to configure the channel.

The PCI/PXI-4461 is automatically configured for differential mode when powered on or powered off. This configuration protects the 50 Ω resistor on the negative pin.

The differential output impedance between positive and negative signal legs is approximately 50 Ω. There is no high impedance idle output channel configuration for the USB-4431. However the output can be configured to maintain its existing value when idle or return to zero when idle.

The differential output impedance between positive and negative signal legs is approximately 22 Ω when you generate a waveform. When you are not generating a waveform, configure the AO.IdleOutputBehavior property for one of the idle behavior options listed in the following table.

A sampled signal repeats itself throughout the frequency spectrum, as shown in the following figure. This figure shows how the signal repetitions begin above one-half the sample rate, f_s, and, theoretically, continue up through the spectrum to infinity. Images remain in the sampled data because the data actually represents only the frequency components below one-half f_s (the baseband). The device filters out the extra images in the signal in three stages.

First, the data is digitally interpolated by a factor of 2ⁿ, where n is a positive integer from 0 to 6 (USB-4431) or 0 to 7 (PCI/PXI-4461). Therefore, the effective sample rate (f_es) is 2ⁿ × f_s. The interpolation factor must be sufficient to move the effective sample rate into the 51.2 kS/s or higher range (USB-4431) or the 102.4 kS/s or higher range (PCI/PXI-4461). The following figure shows an example of four-times interpolation and the resulting images. A linear-phase digital filter then removes almost all energy above one-half f_s.

Second, the DAC resamples the data to a new frequency (f_DAC). The frequency f_DAC is eight times higher than f_es. The following figure shows the resulting images.

Some further (inherent) filtering occurs at the DAC because the data is digitally sampled and held at eight times f_es. This filtering has a sin x/x response, yielding nulls at multiples of eight times f_s, as shown in the following figure.

Third, a multi-pole analog filter with a fixed cut-off at 89 kHz (USB-4431) or 243 kHz (PCI/PXI-4461) filters the remaining images, as shown in the following figure.

Output filter delay—the time required for digital data to propagate through the DAC and interpolation digital filters—varies depending on the update rate for DACs. For example, the filter delay at 10 kS/s for the PCI/PXI-4461 is 38.5 update clock cycles. This signal experiences a delay equal to 3.85 ms. This delay is an important factor for stimulus-response measurements, control applications, or any application where loop time is critical. You often might want to maximize the sample rate to minimize the time required for a specific number of update clock cycles to elapse, since it varies with frequency, as shown in the following tables.

The interpolation filter adds additional output filter delay depending on the update rate. The following tables provide more information about how the interpolation filter affects the output filter delay.