Frequency and Bandwidth Selection to Optimize Performance
- Updated2025-03-31
- 4 minute(s) read
Frequency and Bandwidth Selection to Optimize Performance
You can configure frequency and bandwidth settings of an NI VST to optimize measurement performance for different types of applications.
Basic Behavior
For a simple acquisition, you can set the I/Q carrier frequency and the I/Q rate. When you do not specify a bandwidth constraint, NI-RFSA selects the widest filter for the given frequency region. You can obtain the bandwidth of the instrument for this configuration by querying the Device Instantaneous Bandwidth property. To determine the center of the device instantaneous bandwidth, you can query the Downconverter Center Frequency property.
You can specify the I/Q rate, which configures the decimation filter and, as a result, configures the available passband. When oversampling, the passband is often greater than the actual signal bandwidth present at the port, so specifying the Signal Bandwidth property gives the driver additional information about the requested bandwidth.
The following example illustrates the default behavior.

Specifying Device Instantaneous Bandwidth
In some cases, the default behavior may not be optimal. You can improve the default behavior by providing NI-RFSA with more information about the system. By setting the Device Instantaneous Bandwidth property, you are specifying how much instrument bandwidth is required. The instrument bandwidth is used in filter selection and is only coerced up, so the filters chosen will always yield a bandwidth as wide or wider than requested. You can always read back the coerced or actual device instantaneous bandwidth to find out how much bandwidth is available in the given configuration. For example, if only 160 MHz of bandwidth is needed from an instrument and the next largest filter for that instrument is 1 GHz, NI-RFSA chooses the 1 GHz filter.
Oversampling
In the following example, the I/Q rate is set to 320 MS/s, even though the waveform signal bandwidth is only 160 MS/s. This is referred to as oversampling. The passband is 80% of the I/Q rate, or 256 MHz. In this example, because the passband extends outside the device instantaneous bandwidth, frequency content outside the device instantaneous bandwidth will be present in the time-domain data. This frequency content is usually filtered out by demodulation algorithms but may be problematic if you are analyzing the time-domain data manually, and there is a strong interfering signal in that region.

Specifying a Frequency Offset
To avoid LO leakage present in the passband, you can apply a frequency offset by setting the NI-RFSA Downconverter Frequency Offset property or the NI-RFSG Upconverter Frequency Offset property. This property maintains a constant offset if you vary the I/Q center frequency. You can always read back the actual device instantaneous bandwidth and downconverter center frequency.
Specifying Signal Bandwidth
Another way to provide bandwidth information to the instrument driver is to specify the signal bandwidth. Previously, it was noted that the device instantaneous bandwidth is centered at the downconverter center frequency. Signal bandwidth, however, is centered at the I/Q carrier frequency, which may be more useful for some applications as it follows the signal of interest. Similarly, you can specify the downconverter frequency offset instead of the downconverter center frequency if it is more intuitive for the application.
The following example shows how to programmatically specify the signal bandwidth.

Note that setting the signal bandwidth does not necessarily avoid hardware reconfigurations during in-band retuning, but the settings stay relative to the I/Q carrier frequency, so it may be easier to manage in your application. The following example shows three acquisitions with the same I/Q rate, signal bandwidth, and downconverter frequency and only changes the I/Q carrier frequency for each acquisition.
