Directional coupler's test fixture
Directional couplers (fig. 1) are the "heart" of SWR / Directional Power meters. When developing a directional coupler, typically several iterations are required before the desired results are achieved. Each iterated directional coupler (DUT) must be connected to a set of RF test cables in order to be tested. Repeatedly soldering and desoldering test cables to all the iterations is extremely tedious and is also prone to errors. Luckily, the soldering / desoldering steps can be avoided altogether by using a test fixture (fig 2).
Fig. 1: Schematic & fabrication details of a Guanella directional coupler using binocular core
The test fixture adapts a PCB originally intended for a mixer (fig. 2). Mixer PCBs are ideal for adapting to directional couplers because they have more inputs / outputs than amplifier PCBs. The PCB's top layer is the groundplane, whereas the bottom layer consists of the microstrip traces. The opposite ends of the microstrip traces are connected to the DUT and RF connectors, respectively. The connectors are right angle thru-hole SMA.To stiffen the PCB for repeated DUT insertion & extraction, it is attached to an 5mm thick aluminium plate using snap-on plastic spacers (standoffs).
Fig. 2: Photo of test fixture for directional coupler
The PCB uses pin receptacles to connect to the DUT. Pin receptacles were chosen instead of conventional IC sockets because the former protrudes less above the PCB surface, hence minimizes their influence on the measurement (fig. 3).
Fig. 3: Using pin receptacles, different iterations of the directional coupler can conveniently tested without soldering / desoldering
Fig. 4: Photo of test fixture connected to nanoVna
To characterize the fixture's loss & mismatch performances, a 0.4mm wire is used to bridge the IN and OUT traces at their sockets (fig 5). The results indicate that the fixture is usable over DC~300 MHz. The loss and matching are better than 0.2 dB and -22 dB over the aforementioned frequency range.
Fig. 5: Photo of wire bridge used to characterize the fixture and measured insertion and return losses. The results demonstrate that the fixture has adequate insertion and return losses up to 300 MHz