The mixer has arrived from EME Kits and the board worked. I have about 5dB net conversion gain from 70MHz to 470MHz with a 400MHz LO. There is a 6dB pad in there so higher gain is possible. It also works at 50MHz with a little more gain.g0mjw wrote:The test boards have arrived from PCBway earlier than expected. Not bad for $16 shipped. The vias are so well implemented I had to use a torch to check there really were holes there. The silk screen looks good. The jury will be out on the design until I get all the parts together. I have already made some improvements to the board design. Even thought about putting the LO on-board using one of those 400MHz LVDS chips, but it's probably best and more flexible left as it is.
As I suspected, the tight output filter design I had intended to kill the LO leakage and any images wasn't on tune with the calculated value parts above. That's most probably due to PCB lead inductance, which is significant (5mm track == 2nH) and PCB stray capacitance. I replaced the resonating capacitors with a Murata SMT trimmers and found it easy enough to tune the output filter. I have already modified the PCB to allow trimmers to fit. Of course the inductors and capacitors could be recalculated to compensate but that would take a lot of fiddling. In practice, the result was pretty much what I wanted covering 450MHz to 470MHz and dropping off fairly quickly either side.