Education

(Misc. topics)

2025-07-10 Off-freq Antenna Gain reduction

2023-08-25 Reactive Loads

2022-07-08 Free Book: WNDW

Getting / Upgrading / Renewing Your Amateur Radio License

2025-07-10 Off-freq Antenna Gain reduction

Inflated gain figures

Antenna manufacturers love to brag about their gain. Some things to reduce your excitement over their numbers:

Some out-n-out lie (No, that simple shortened 1/4-wave wire doesn't have 10dB gain!)
They like to quite dBi (subtract 2dB for dBd)
They give overall gain, not max gain at a particular angle (concentrated where you want it, or spread out?)
They don't tell you what the takeoff angle of max gain is. (Straight up isn't helpful for VHF/UHF!)
They quote max gain but don't tell you how quickly it drops off away from the center design frequency (i.e. what the bandwidth is for a particular drop)
They don't quote realized gain, i.e. after subtracting impedance mismatch loss.

Points 3 & 4 are well addressed by them showing a plot from a computer model.

Meshtastic

As I'm selecting antennas for Meshtastic antennas, I was especially concerned about this because I'm paying good money to put high-gain antennas (mostly omni-directional collinear arrays) in key locations, and don't want to end up w/ "cheap antenna" performance.

Initially I was concerned about overly broadband antennas (which are unlikely to have usefully high gain, gain generally being inversely proportional to "Q") that covered both the European "EU868" (863-873 MHz) band and the American "US915" (902-928 MHz) band, a 65 MHz swath centered on 895 MHz and outside of the band of interest, so probably not working that well.

Then, I realized that the default Meshtastic channel, er, frequency slot, #20 is down at 906.875 MHz (actually, I was erroneously thinking 902.something), so it's far from the band center, meaning that an antenna designed for 915 isn't optimal either (but a "universal" antenna isn't as bad at 907 as it would be for 915 MHz, either).

Gain-reducing factors

Antennas are usually designed to minimize SWR at the design frequency. Increasing impedance mismatch means increasing loss. Additionally, the radiation pattern changes as the frequency moves. For some simple low-gain antennas like a horizontal dipole, the change from the classic "bloated donut" shape to a more complex shape w/ sharp lobes can mean lucky gain increases in certain directions (at the expense of other broadside gain), but when you're talking about a vertical antenna where maximum horizontal gain is desired, no deviation from the design pattern is likely to be good.

Arrays are even more sensitive to radiation pattern changes; they depend on exact spacing of elements causing interactions that create nulls in the EM field and push more energy into the desired direction. At other frequencies, gain can fall apart. (Similarly, beware of vegetation or other nearby objects that can become parasitic elements in your antenna system and dramatically alter the pattern. I once field-strength-tested a Yagi that, due to antenna damage and poor placement, seemed to have the most gain 90 degrees away from where it was pointing!)

Gimme some numbers

I searched the web and only found one quantitative example of gain reduction vs. frequency change: an old military publication that gave 3 gain numbers for a particular collinear array at 155, 160, and 165 MHz. The numbers were scary, but then I extrapolated those +/-~3% values and found that they correlated to 886-944 MHz -- twice as wide as the 915 MHz ISM band -- so percentage-wide, 915 isn't too large so the worst degradation would be less than their testing.

I pulled up a simple antenna modeling program (MMANA) and found that it came with a simple 3-element collinear. (It used 5/8-wave elements, not 1/4-wave which is also popular) and doesn't necessarily closely represent any particular antenna that I'm using, and there are lots of other variables in antenna modeling that I'm not worrying about, but it's close enough to give a useful idea of what to expect.

I ran simulations on 17 different frequencies, from 94 to 110% of the design, translating the design frequency (145 MHz) to 915 MHz, and recorded the predicted feedpoint SWR, the overall gain, the gain at an 8 degree take-off angle, and the value& angle of maximum gain. The resulting table is shown below. The 5 center rows bound the US915 band, so worst-case is less than the outer values there.

Bottom line? Not as bad as I feared.

I also found it interesting that the radiation pattern didn't differ nearly as much at lower frequencies as it did at higher ones. Note how that max gain is actually at 8% high -- but the useful gain (at 8%) is cut in half.