To balun or not to balun...

[Greg Balmer]

We're starting out on BLE project with the nRF51822 and trying to get a prototype design together having completed a POC with the developer kit and PCA10004, and are considering using a trace antenna initially.

Looking at the 2 reference designs for the QFN package as hopefully a good place to start, I'm wondering what the advantage of using the one with the ST balun is or if it is even relevant when using a trace antenna?

Sooner or later I guess we will migrate to a chip antenna as we would like to achieve decent range performance in a domestic scenario, but I fear our current skill-set and resources are not up to designing the necessary additional circuitry atm... (the antenna tuning guide looks a bit like black magic to me)

Any advice much appreciated,

Thanks,

Greg.

P.s Are the reference designs available in Diptrace format? or in some other portable encoding? as Altium is pretty pricey for an early days startup...

[FormerMember]

Hello,

1) The adventage of using the balun is that it is smaller than the discrete matching network, so that is makes the design slightly more compact. The performance of the balun and the discrete matching network is the same.

2) The performance of a chip antenna and a PCB antenna is the same. The main differences between them are the size and the price. Both a chip and PCB antenna have to be tuned; the chip antenna with a pi-network and the PCB antenna with a shunt component, like described chapter 4 in the "Antenna tuning" white paper. For design of a PCB antenna I recommend you to read the "Quarterwave printed monopole antenna for 2.4GHz" white paper. It may also be useful to take a look at this post.

Regards, Kristin

[Greg Balmer]

Thanks Kristin, from your response and reading those papers it's apparent that we will probably need the shunt for the PCB antenna. If I look to the PCA10004, and how it deviates from the ref design, for inspiration, it looks as though a single shunt cap is added at C17 - is this correct? Is its value documented anywhere? (I understand that the antenna impedance will vary with PCB design but I'm looking to get into the ballpark for the first prototype...)

If I was to use the nrf51822 ref design with a shunt in a similar position to C17 (As per PCA10004), in combination with the bent Antenna design @ Figure 1b of "Quarterwave printed monopole antenna for 2.4GHz" white paper, can you propose a good initial value for this capacitor? I'm not quite sure how else to begin with the design process - perhaps 1.0 pF capacitor (as per the value in "antenna tuning" white paper... section 8.2).

It would just be great to get some good advice on getting a prototype off the ground without having to spend time initially in an RF lab....

Thanks again,

Greg

[FormerMember]

Yes, the difference between PCA 10004 and the reference layout is that C17 is added. The reason that C17 is not in the reference layout, is that C17 is more related to the antenna than the chip. In addition, what kind of matching network that should be used depends on the antenna. I agree that it can be a cunfusing. The value of C17 should be found during tuning, so while the antenna is not tuned, it is not necessary to mount any component on C17. Normally, there will be some range when the antenna is not tuned. However, by tuning it, the range will often be increased significantly.

[Greg Balmer]

OK, we'll provision pads for C17 and cook the board without the cap initially and go from there. Thanks for the help, hopefully we'll get some range with the antenna untuned...

[pavan]

Hello All,

Not able to open this link Quarterwave printed monopole antenna for 2.4GHz