PCB via fences around RF modules: pitch rules that move isolation (and cuts that kill them)

【Cross-site note from rfforum editorial / rfforum_ops】 Lab and instrument boards next to radios often add a via fence around an RFIC and still fail isolation or desense. Isolation is a return-current problem before it is a “more vias” problem. Checklist we use (soft promo = 0): 1. Fence pitch starting near λ/20 at the highest aggressor is a start, not a guarantee — keep the fence close to the paddle edge. 2. Stitch every ground layer the part actually references. 3. Connector or USB cuts in the return plane defeat the fence; bridge them densely or redesign. 4. Measure isolation/desense before vs after fence changes on the same fixture. Companion: https://www.rfforum.com/t/253 Question: for Wi-Fi / SDR / instrument front-ends, what paddle-to-fence distance and pitch survived your last desense bring-up? — rfforum_ops / https://www.rfforum.com

【Cross-site note from rfforum editorial / rfforum_ops】 Lab and instrument boards next to radios often add a via fence around an RFIC and still fail isolation or desense. Isolation is a return-current problem before it is a “more vias” problem. Checklist we use (soft promo = 0): 1. Fence pitch starting near λ/20 at the highest aggressor is a start, not a guarantee — keep the fence close to the paddle edge. 2. Stitch every ground layer the part actually references. 3. Connector or USB cuts in the return plane defeat the fence; bridge them densely or redesign. 4. Measure isolation/desense before vs after fence changes on the same fixture. Companion: https://www.rfforum.com/t/253 Question: for Wi-Fi / SDR / instrument front-ends, what paddle-to-fence distance and pitch survived your last desense bring-up? — rfforum_ops / https://www.rfforum.com