Sparkfun seems to be selling several nRF24xxx based boards with different types of antennas and layout. There’s a ceramic antenna, a straight trace on the PCB, as well as a circular PCB trace.
Being a CS guy I don’t know a thing about antenna design - which type should I use for my own PCB design based on that chip ? The application notes suggest a large ground plane this however would interfere with some of my sensors. Is it mostly important to have the ground plane on the backside of the chip for shielding or does it have to be large to work as part of the dipole ?
The first one is a folded dipole, it will probably give the greatest range, but is directional. Thay can be easier to match to the output as they are balanced.
The monopole will be fairly non-directional if both antennas are vertical. It won’t be all that efficient because the antenna doesn’t have a true ground plane.
I have also found some nice application notes by Nordic regarding antenna traces. They suggest a slightly longer antenna (30%) for small ground planes and then “cutting it to size”.
Expensive test equipment is really needed for designing and testing antennas for optimum performance at 2.4 GHz, trial and error doesn’t work very well.
Would it be possible for someone to either give us a good guess as to the two inductors and the 3 caps that are used in the modules, or a schematic and BOM. Given that the sparkfun modules seem to work the nordic reference is obviously more complicated than it needs to be.
“The antanae design is not for amatures” junk doesn’t help much when there is a decent reference design.
Ditto to what Leon said, best to try to design staying as close to a reference design as possible, since unless you have the equipment to tune the RF path, you won’t be able to do it (case and point, I do this for a living, and am under-equipped with a $17,000 spectrum analyzer/sweep generator and a $1000 directional coupler… Occasionally I travel to a partner company that has a VNA to do a more accurate job of circuit tuning).
When you build a reference design for RF, it won’t perform as good as it would with tuning (since simple things like different PCB materials, different PCB thickness (HUGE effect on performance), and even different manufacturers of supposedly the same component will not be the same as the reference design) but it should usually get you “close enough” to work “decently”, although likely nowhere near what the reference design supporting documentation says it can do.
Check the Nordic website, they have some pretty bare bones reference designs in addition to more complicated RF-paths.
RF, especially 2.4 GHz is half a black magic-voodoo… Even if you know what you’re doing, unless you have a way to measure it, you just taking a shot in the dark.
Usually its the mathematician that is supposed to starve to death because he cant get to the food. The engineer is supposed to be able to figure it out that you can get close enough and in this environment hopefully show people how to get there as well.
My point is that there are 3 examples which have extremely simple layout and parts count. (just no published values) I will try to find the nordic samples that are not as complicated as the layout and matching network. The one in the datasheet is more expensive than the part.
I’m not saying it’s not possible, and in fact I don’t want to discourage you from trying… Far from it. I’m just trying to point out that “simple” RF design isn’t simple, especially in high frequencies like 2.4 GHz… So trying to stay as close to a reference design as possible will usually yield the best results. II suppose I could have worded that intent better, gotta love the Internet for that.
Check the app note for the NRF2401A used in a USB dongle. That has a simple Balun… I’m not entirely sure of Nordic’s product selection, but if they have similar RF front ends, that particular output design might work for some of their other chips as well.
Anyways, like I said, the last thing I want to do is discourage you from giving it a shot… You won’t learn anything from giving up based on someone on the Internet saying it’s too hard to do… If you need help, I’ll be glad to provide insight, along with I’m sure several other people on this forum.
I am working with the NRF2401A at the moment and the usb design appears to be much closer to what I was hoping to find. I am actually hoping to incorporate the design of the smaller sparkfun modules into a pair of single board designs using avrs (one with an atmega and one with an 90usb162).