That RF trace looks like it wanders a long way (relatively speaking) before it gets to the SMA connector. It needs to be closer to the RF pin of the radio module to avoid problems.
You're exactly correct, thanks. When I resized the board (it was much wider with a DIP Atmega) I flipped the radio 180 degrees and forgot to move the SMA connector. D'oh!
Rather, it would wander a bit, if it were actually connected...
Looks like you got bitten by the eagle "feature" where it looks like a wire is connected, but it isn't actually connected. the RF out isn't actually connected to the radiometrix footprint.
And this is even worse! So the antenna connector went straight to ground. I was surprised in my initial testing that while my home APRS station could receive the packet, the local iGate (5miles away or so) didn't see it. I had put this down to me using a small rubber duck for my test and that I would try it again later with a decent antenna.
The problem is so obvious too looking at the finished board. I'm embarassed. I quickly went back thru the dozen or so board variations I'd created and was relieved that this was a new "enhancement" that I'd added.
It really does pay to go thru a schematic and "shake" each part with the move tool to make sure it's connected exactly as you expect - this isn't the first time I've been so bitten. I believe there is a similar problem near the VR as well <sigh>
Since you appear to already have your boards fabbed, you can use a short piece of small coax (e.g. RG174) to jumper from the header to the SMA.
Good idea. But I did only have 3 made ($14 worth) and I wanted to find a better mounting solution for the GPS at any rate: I'd like to find a pair of connectors I can use to mount the GPS on-board as well as inserting a cable in-between (and something a little more robust than a pair of 0.1" connectors).
BTW do you think it would be wise to increase the size of the RF trace?
Thanks again for your feedback,