Choosing Interconnect for RF Test & Measurement
Test fixtures for devices under test (DUTs) are critical elements during RF benchtop evaluation. When assessing DUT performance, the ideal scenario is to directly connect the test equipment to the DUT’s ports. Unfortunately, this approach is often not physically feasible, so, a test fixture is used between the DUT and the test equipment. The test fixture usually includes a combination of connectors, cables, and PCBs. (The DUT could range in complexity from a single chip or component to an entire PCB subsystem.) One of the main goals is for the test fixture is to be as electrically transparent as possible, meaning it should not affect the measurement results across the full bandwidth of the measurement being made. This means that choosing interconnect for RF test & measurement is critical.
A new white paper from a team of RF engineers at Samtec reviews how some types of interconnects (cables and connectors) in benchtop RF test have evolved due to changing test needs over the years. It includes a focus on the important technical challenges for the modern test fixture. (You can read the entire white paper in our technical library.)
Ganged Solderless Interconnects
The white paper includes detailed measured and simulated data on ganged solderless interconnects, an enabling approach to avoid bulky test fixtures in space constrained applications.
A ganged connector offers a high number of channels. In addition, ganged solderless compression systems can improve the speed of swapping out connectors, reliability, and real estate issues. Our research shows (see below) that ganged solderless connectors with cables (see Figure 1) can perform better due to the fact that cable losses are always lower than PCB losses, and, since ganged connectors have increased density, they can be placed closer to the DUT, further minimizing loss.
The benefits of ganged connector solutions can be leveraged even more by changing the plane in which the cables approach the fixture. This becomes particularly important when the test fixture has z-height restrictions. In these applications, bringing the cables in parallel to the PCB is advantageous. There can also be density improvements from unique solutions that allow connectors to be mated on both sides of the PCB, also known as “belly-to-belly” (see Figure 2).
Testing Performance
In the analysis performed in the white paper, the goal was to achieve the lowest insertion loss (IL) and return loss (RL) possible for a test fixture over the frequency range of interest. Figure 3 shows the comparison of trace loss to two ganged cable connectors: one ganged connector using .047″ diameter low-loss cables and the other using .086″ diameter low-loss cables.
The results indicate that ganged cable connectors minimize fixture IL as compared to a stripline approach in a low-loss dielectric. These types of results can give test fixture designers confidence when using ganged compression mount connectors.
Again, you can check out all of the details, including a brief review of the evolution of interconnect technology for RF test, in our new white paper.
Also See:
Connecting Test and Measurement – The Samtec Blog
110 GHz Test Solution Enables 224 Gbps PAM4 SerDes Characterization – The Samtec Blog
