2
$\begingroup$

I have a discone antenna intended for use in the 25-1300 MHz frequency range, but it uses a SO-239 connector.

Now, we know that the so-called "UHF" i.e. the PL-259 male plug and SO-239 female socket connectors are not great even in the 30–300 MHz (VHF) range, much less at the 300–3000 MHz frequencies range we now officially call the UHF range. So it seems like an unfortunate choice of connector for an antenna that works up to 1.3 GHz. And indeed other wideband VHF/UHF antennas (e.g. the CLP-5130 log periodic which also goes to 1.3 GHz) do use a "proper" N connector instead.

But since is the "end of the line" anyway, where the coax ends and the antenna begins, is there really any practical difference to the choice of connector here? If it's okay to solder antenna elements directly to coax would a short bit of whatever intervening metal be much worse, especially if it's yes primitively but yet still (maybe?) somewhat designed for higher frequency RF?

Likewise, what difference does the UHF connector on the radio make? Many base stations that I've seen have a sad little piece of coax, half of which has been peeled apart so it can be soldered to the PCB and to the connector itself. This is usually justified against the "1/10 wavelength" rule of thumb, where the impedance of short traces/leads isn't considered particularly important.

So by the same rule of thumb, do the poor specs of a UHF connector really matter in the UHF range? A tenth of a wavelength at 3 GHz is 1 cm, which is about the depth of a typical PL-259/SO-239 threaded connection. Or to put a picture (found via http://wa6pzb.blogspot.com/2014/09/220-mhz-ground-plane-antenna.html but looks to be credited to AI4JI originally) on this question:

Ground plane antenna mechanical drawing, with a radiator soldered onto the center of a coax socket and four radials attached to the corners of its mounting plate

If I were to make this ground-plane antenna with elements targeting a 1.3 GHz resonance, would it make any difference if built it around a "proper" N or BNC socket vs. any old UHF socket? And if so, why? It seems to me like the transition from "coax" to "antenna" is going to be pretty rough regardless!

$\endgroup$

3 Answers 3

2
$\begingroup$

If you're planning to tune the antenna for best VSWR anyway, then you'll compensate for any transformation that happens in the mated connector pair.

A PL-259 - Barrel - PL-259 combination in the middle of a cable might significantly raise your VSWR at 1300 MHz, in a way that's hard to fix. But right at the feedpoint, it's just another imperfection in the antenna that you'll adjust for.

You might see a small change in antenna performance if you tune it with Brand A PL-259 on RG-58 cable, and then install it with Brand B PL-259 on RG213, but this won't be a disaster.

Depending on where it occurs, 1/10 of a wavelength is quite a lot, it's almost half of a quarter-wave, the most significant length of wire. If the opened-up coax has an impedance of (say) 80 ohms instead of 50, a quarter wave will transform to 130 (causing 2:1 VSWR) while 1/10 wave will cause perhaps 1.5:1 VSWR. The rule should really be "much less than 1/10 wavelengths".

$\endgroup$
3
  • $\begingroup$ "just another imperfection in the antenna" — ok, yes that's helping confirm a realization that started dawning halfway through writing up my question. I.e. that a well-done antenna design would have to include the feedpoint in its purview, selecting a socket and carefully engineering any necessary transition between it and the actual "elements". $\endgroup$ May 13, 2022 at 19:36
  • $\begingroup$ (So maybe the designer found a good SO-239 and overcame its limitations, or maybe they just whacked on a cheap part they knew their market would accept/prefer? I'm not here to judge my the manufacturer of my discone; I picked their product primarily on price and my antenna system as a whole is still mostly limited by its control operator and/or the detrimental terrain surrounding it :-) $\endgroup$ May 13, 2022 at 19:48
  • $\begingroup$ I answered mostly thinking of the actual monopole groundplane thing in the picture. In general UHF connectors will give more toruble at this frequency, and should be avoided. (Just use SMA, they're cheap on ebay, buy 10 of every kind for your junk box). For this antenna it won't matter too much. And with care you can get away with things... for example we once did direct soldering of RG213 (8 mm braid diameter) to the feed dipole of a 5.8 GHz yagi (20 mm wide) with VSWR reliably below 1.5:1. $\endgroup$
    – tomnexus
    May 14, 2022 at 5:03
2
$\begingroup$

Addressing the "sad little piece of coax, half of which has been peeled apart so it can be soldered to the PCB and to the connector itself" part...

Yes, in HF radios it is common to see the coax shield and center conductor split between the back side of the SO-239 and the PCB. But at HF frequencies, that will still be far less than 1/10 wavelength.

If you look at the specs of the expensive high end SMA (and others) connectors designed for PCB at 5GHz and up, they actually have very explicit instructions on how to solder the connector directly to the PCB, including the shape of the pads on the PCB you are suppose to connect it to. As the frequency goes up, these things become more and more critical.

$\endgroup$
1
$\begingroup$

A geometric change in a transmission line geometry that is far smaller than 1/10 wavelength can still cause an impedance mismatch, thus reflections at GHz frequencies, thus increase SWR. There are (expensive?) CAE tools that can extract (spice, EM simulator, et.al.) parameters from CAD geometry for items the size of connectors (and far smaller). For some antenna and feed line designers (for high volume cellular mobile devices, etc.) these parameters and the trade-offs involved do make a difference.

$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .