Does anyone have any experience selecting chip antennas for the P0 module?
Unfortunately it doesn’t seem like the Advanced Ceramic part used on the photon is readily available (can’t find it on Digikey/Mouser etc.)
I do understand that products require FCC certification however at this stage I just need to whip up a PCB for a pitch pretty quickly (within a month) so was hoping that somebody might know of an alternative that was likely to work if I copy the RF layout of the Photon!
I also realise that I could use the P1, however it’s a little too big for my application.
I offer this… you’ll find a negligible difference among chip (on PCB) antennas., They’re all too small versus the wavelength at 2.4GHz to be good radiators. And they’re not good omni-directionally either.
A 1/4 wave length or so of wire connected to the U.FL will do better.
I’ve been researching since I posted this and I can see a lot of work goes into these designs.
If I was to go for a chip antenna, do you think selecting one with parameters as close as possible to the Advanced Ceramic one, and copying the Photon layout and RF inductors would yield a reasonable chance of success? I realise signal strength might not be optimal but that’s ok, if I go for FCC compliance later it will be refined.
Given your response, do you think it would be equally possible for me design a 50ohm 1/4 length (31.25mm) trace antenna down one side of the board? Or would I need to use a network analyzer to fine tune this?
It is very unlikely you will get a good match to a ceramic antenna simply copying the design, there are just too many variables, and small changes in the layout can have large consequences.
However, it is equally true that unless you need significant range or high performance, you can probably get away with just about anything, and the various protocols will usually do a pretty good job of hiding all but the most egregious problems.
Thanks for the suggestion @AndyW, I could but I’m anticipating potentially going into production with this device so I suppose I should start on the right track, even if I don’t get great performance first time around. Interestingly I downloaded the spark component library and found a trace antenna in there, I suppose Spark may have experienced with this early on, although I can only guess that it’s performance wasn’t as good as the current chip antenna. Perhaps I’ll design in a chip antenna with footprints for a matching/Pi network and try and iron out the details later when I have access to a network analyzer.
Ah, sorry - I was keying off your “whip up a PCB for a pitch pretty quickly” comment. I would not go to production with a patch antenna, was just trying to help you out in a pinch.
Without the right test equipment, or help from a vendor/friend (who has the right test equipment) - you’ll really have absolutely no idea how good the match is after even the most seemingly innocuous layout change to an existing design.
No worries, it was a useful link thanks, it probably sounds like I’m trying to get the best of both worlds, in reality I don’t mind if I only get 10m or so line of sight reception on the first go, it’s just I’m pitching a wearable device and I didn’t really want an antenna dangling off it, by I didn’t explain that previously
I’ve read quite a few articles and reference design notes on the internet and some say it’s possible to achieve ‘adequate’ performance by designing a trace F antenna slightly too long and trimming it while watching the RSSI values. I realise this would probably make an RF engineer cringe.
Whatever I end up doing, I’ll post back here the results, just in case it’s useful to someone at a later date.
Better to use a 1/4 wavelength wire oriented optimally.
PCB antennas take up a lot of board space. If you do that, be sure to exactly duplicate a respected reference design… down to the details: board material, copper thickness, exactly. Even so, it’ll be inferior due to directionality.
