I note what seems to be a minor conflict re: the copper/FR4 keepout region for the imp-002 antenna. Figure 12.3 of the -002 spec would seem to indicate some encroachment of the FR4 into the rectangular region shown in figure 11. (Normally not a big deal, and I’d just go with figure 11 … but just trying to understand if it’s ok to have a bit of add’l FR4 as shown in fig. 12.3)
Thanks,
Larry
(On another note: also just saw the prelims on the muRata module. Pretty neat … something else to read up on and look forward to).
This is by no means a recommendation, but we encroached on the antenna on some of our PCBs accidentally because of misinterpretation of the requirements, and saw no ill effects. On one board, we only had 1.5mm clearance on the sides of the antenna (completely free of FR4 underneath) and its hard to tell if there is any WiFi performance degradation at all. However, you could end up coupling additional RF energy into your circuit if you have tracks or power planes in this area so performance (analog especially) may suffer because of increased noise. It may cause issues for EMC compliance testing if that’s a direction you will be following.
I guess what I am saying is that to avoid headaches, figure 11 is the one to go for, but I have experienced that you can get away with major encroachment on the keep-out zone…
That’s a good point! Figure 11 is the gold standard to follow for the absolute best antenna performance. I’ll try and clarify a bit what you’re seeing in the screen shots on recommended layout.
It’s very important not to have any FR4 under the antenna. The dielectric constant for FR4 is different than for air, and the antenna is tuned for air. It’s not a notable problem to have some FR4 in the bottom corners of the keepout area, as long as they’re not under the antenna element or right next to the ends of the antenna trace.
It’s also very important to avoid putting any copper in the keepout area. You’ll note that in the screen shots, while there is some FR4 in the bottom corners (really just for mechanical reasons and some aesthetics), the copper planes have been pulled back so they don’t encroach on that keepout. It’s also very important to keep components out of that area.
Thanks for your feedback, as well as observations and explanations, deonsmt and T. Really appreciated … (and I’m sorry for the delayed response).
Yea, makes sense that best practices would be to use the maximum recommended keepout area for FR4, copper, and components … it is interesting to note however that the performance appears robust enough to permit some (perhaps even significant) encroachment. (I wouldn’t be doing any compliance testing at the moment).
In general though, I’ll plan to adhere to the ‘gold standard’ for keepout.