I'm interested in evaluating the suitability of a standard-gain pyramidal horn for an ultrawideband (UWB) signal. However, I've had difficulty finding reliable information about the phase response over frequency that would be needed to make this determination.


UWB Definition

In this context, use the FCC's definition of UWB, the relevant parts of which are:

  • Operates in the 3.1 GHz to 10.6 GHz range
  • Bandwidth greater than 500 MHz

The application I have in mind will not need to work over every frequency in that band, just a subset of it that is approx 500-2000 MHz wide.

Considerations w/ Wideband Antennas

There are many types of "wideband" antennas that can easily meet the bandwidth requirement (log periodic, horn, etc.). However, a good UWB antenna should also have these properties (according to [1]):

  • Linear phase over the signal bandwidth
  • Constant phase center over signal bandwidth

These conditions are necessary to prevent the UWB pulse from being distorted during transmission/reception. This is relevant to point out because many "wideband" antennas violate these conditions. E.g. a log-periodic antenna has a phase center that changes with frequency (i.e. as the frequency changes the active region changes), and some wideband antennas may have a relatively flat gain response over frequency, but might not have a linear phase over the entire band of interest.


Does a pyramidal horn antenna meet these requirements?


[1] Stutzman, Warren L., and Gary A. Thiele. Antenna theory and design. John Wiley & Sons, 2012.


1 Answer 1


Now that's a good, hard question, and well phrased! Disclaimer: my antenna engineering experience is 40 years old, but if physics hasn't changed, maybe I can help you some.

First, you're right, for this application, avoid log-periodic antennas in favor of antenna designs with a single, broadband feed point, such as the horn you describe. You will have a constant "phase center" over the horn's operating bandwidth.

But you can think of the horn as a special kind of waveguide whose dimension varies along its length. And the horn, like any waveguide, will exhibit "frequency dispersion" -- that is, waves propagate through the horn at a velocity that varies with frequency, slowing down especially as you get close to the horn/waveguide's lower cutoff frequency. (I'm skipping over the fine points of group velocity vs phase velocity here; you care about the UWB modulation on the wave, which is affected by group velocity.)

Practical thoughts:

  1. If you don't need directivity, consider a broadband dipole instead of a horn.
  2. If you do need some directivity, choose the shortest, least directive horn available, as it will exhibit less dispersion.
  3. Operate as far from the lower-frequency cutoff of the horn as possible
  4. If possible, find a horn with an exponential taper, rather than a linear taper.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.