11/04/2015 at 04:31 •
I'm downgrading from the $6.50 Molex SMA connector to the cheaper AliExpress, eBay variety with 0.4mm gap between the GND pins. Some 20 of them finally arrived from an eBay order today to verify the PCB footprint I made, and it fits!
Note: ignore those GND plane via gaps. They are actually directly connected in the PCB. I needed to remove the poly connection to expose the SMA footprint. For some reason, the print out version of the top layer isn't showing the SMA pads, but it's fine on the Gerber file.
Now, all that is left is wait for those pesky 1.27mm single row headers to arrive to verify that there will be no component collisions. I'm not getting good results from the 3D models, maybe it's inaccurate models from the www but won't know for sure until I get my hands on the mainstream supply.
10/16/2015 at 13:24 •
I haven't had a slightest idea whether to continue working on this project, so I've uploaded the new design files, and will be uploading the Gerber shortly after a week or two once every tingling gut-feel to fix board bugs checks out.
At this stage, the board has been shrunk to 24x24 mm which can be panelized easily in a 50x50mm PCB for cheaper manufacturing. It should be simple enough to change the balun filter IC for another according to Texas Instruments application note SWRA250A for operation in 433MHz, 868 MHz or 915 MHz individually.