HPSDR Alex Filter Bank Grounding
In the original Alex measurements, we noticed a significant hump in the frequency response at around 110 MHz. This was due to lack of a low impedance ground between the Alex RX and TX boards. Adding a ground strap significantly improved the ultimate rejection; this plot shows before-and-after frequency responses of the prototype unit:
On the production Alex that I tested, I added ground wires at two points between the RX and TX boards, and extended them to tie into the enclosure as well.
I elected to drill holes in the enclosure cover and use #4 screws to ground the case. The easier way is to wrap the ground wires around the mounting screws that attach the cover plate (or to the Pandora enclosure) but that seemed suboptimal for a couple of reasons. It was easy to pop the holes in the cover, and relatively easy to get the ground wires tightened down to them.
I initially put a single strap on the front edge of the boards (the edge with the ribbon connectors) but that still left a long ground path between the RX and TX coax connectors and the response hump was still present. The grounds near the BNCs were necessary to get optimum performance.
Here are photos:
