This reference architecture serves to demonstrate the use of NI PXI VNAs with NI PXI switches for multi-port measurements.
VNAs are often used in conjunction with switches to increase their total number of ports. This reference architecture should be used as a foundation for beginning switch-based VNA measurements as it incorporates many important practices for robust, reusable, and scalable applications. Many components have been optimized for performance, notably measurement time per port tested.
Because of the sensitive nature of VNA measurements, care must be taken to separately calibrate every switch path (typically one per UUT socket) to compensate for factors such as switch cross-talk, isolation, and insertion loss, which are unique to each path. Once calibration data is obtained for every UUT socket, it must be applied to the raw uncorrected measurement data in correlation with each corresponding switch path. Loading new correction factors onto an instrument each time the switch path changes adds unnecessary test time (2-3 seconds per commit), slowing down overall throughput. Since this instrument loading dominates an overwhelming majority of the total test time, improving this aspect has drastic performance implications for multi-UUT applications.
To decrease this test time, we can leverage the software-defined advantage of PXI by configuring the VNA to return uncorrected data. Identical sweeps are performed across all UUT sockets and system vector error correction is applied in software post-acquisition without having to re-commit any settings to the VNA between sweeps. This way, we avoid the performance hit taken from loading new correction factors for each UUT socket. The more ports being used, the more you benefit from this optimization.
Please note: Current support is only implemented for reflection measurements on VNA port 1.
Relay control
This subVI sets your switch path for the current measurement.
Note: Edit this subVI to support your switch setup. Current support is for 2597 and 2543 RF switches.
Initiate sweep, wait for finish, and fetch measurement
Perform a single sweep for one switch path. Then get the current measurement data from the VNA after the sweep is complete.
Note: The VNA should be configured to return uncorrected data.
Perform error correction and format data
This subVI uses the correction factors associated with the current switch path to apply vector error correction to the measurement data. Corrected data is then formatted for display according to user selection.
Update measurement data
Add current measurement data to results for display.
This execution flow can be modified such that the Get Correction Factors step is only called once, after which the Measurement step can be repeated for multi-batch systems (user prompt may be required to pause execution while DUTs are swapped in/out). This is the key where test time improves, because subsequent measurements do not require committing any settings to the VNA, even when changing switch paths with different correction factors.
Example code from the Example Code Exchange in the NI Community is licensed with the MIT license.
Hi
As there are two method for large signal measurement systems: The sampler-based technique and the mixed-based approach.
Sampler based technique is used in LSNA and mixed-based approach in NVNA.
Kindly correct me If I am wrong, I am new to this field.
Is there any application notes from where I can understand the working of both so that I can better understand.
Thanks.....
Hi gctripathi,
This example is for general purpose VNA. Switching in this case is to increase the total number of ports for reflection measurements.
For general questions, our RF Academy would be a good starting point to learn about fundamental RF concepts. While it may not contain specific information regarding LSNA and NVNA, it is continuously updated and I would suggest looking out for new content.
Julia P.
Applications Engineer
National Instruments