WIFI Antenna Design

I have designed a dual band WIFI antenna (2.4-2.5GHz and 5.1-5.7GHz). In 2.4GHz it's completely omnidirectional but with negative gain and in 5.6GHz it's somehow directional but with very good positive gain. why is this?? Where have I been wrong?

• It would help if you explained the design, or better yet, including a helpful graphic. – PearsonArtPhoto May 3 '15 at 1:04
• thanks for pointing it out. I've added graphs. Hope it helps to clarify my question. – user4621 May 3 '15 at 8:17

Just to make it clear - if you're wondering why the directivity changes from one configuration to the other, there is no clear answer. The directivity of an antenna varies a lot, depending mostly on the shape of the antenna.

If you're wondering why you have this particular link between directivity and gain, it is actually logical that you would find this:

$G = E_{antenna} D$

Where $E_{antenna}$ is the antenna efficiency (you can look it up on wikipedia). If you have an omnidirectional antenna, you have a lower directivity, and thus a lower gain than for a directive antenna.

This being said, you might also have a different antenna efficiency for those two frequencies, explaining a bigger gap than what you would obtain with only the directivity thing.

Decibels measure ratios. It looks like the horizontal line in your graph is around -10 dB. That translates to:

$$-10\:\mathrm{dB} = 10^{-10/10} = {1 \over 10}$$

In English, "-10 decibels" means "one tenth". So, one tenth of what? There is nothing in the graphs you have provided to say.

So when your graphs indicate negative gain, this means "radiation in that direction is some fraction of radiation in that direction compared to some reference which is not specified." Which may or may not mean you've done anything wrong, depending on what that reference is.

• The top graph isn't gain, it's reflection coefficient *. The left hand axis says S11 if you look closely. * (sometimes erroneously called return loss) – tomnexus May 4 '15 at 19:23
• @tomnexus That link is dead – endolith Oct 12 '17 at 15:31

The gain seems to be low in one of the plots, it looks like the peak is about -5 dB (dBi I suppose). The matching of your antenna looks reasonable, so your antenna is absorbing power within the two bands. The pattern also looks like that of a dipole.

So where is the power going? Do you have any lossy elements in the simulation? Any printed circuit boards, real metal conductors, lossy dielectrics, or is it all perfect conductors and insulators? If there's anything with loss, that is where the power is going. Ff you can get HFSS to plot the structure losses or dissipation, that might tell you where it is.

I don't think -5 dBi is very surprising at 5 GHz if you've used real materials in the simulation.

Perhaps post a picture of the simulation model and we can see where the power might be going.