Loss due to polarisation error

The most suitable previous topic was closed having reached 100…

Re the attenuation due to using “the wrong polarisation” I was “reflecting“ * on this situation this morning and worked out a few basic calculations.

The attenuation is basically due to the wave front intersecting the antenna and only creating a fraction of the incident energy available. Well, what fraction?

A model for the situation is an antenna that is inclined at an angle, call it theta to the angle of the incident wave. The antenna will receive a fraction of the energy that can be calculated as sin(theta). This will give the current in the antenna that is induced by the incident wave.

At 45 degrees, sin(45) = 0.7, causing 0.7 of the available voltage and current from the wave to flow in the antenna elements. This is converted to db by calculating the ratio of the reduced energy to the original energy which is 20 log (0.7) = -3db, which is intuitively correct as titling the antenna by 45 degrees should result in the same loss for both polarisations.

So my next thought was to ask how much signal degredation should result from a serious mismatch, say 89 degrees (almost completely the wrong polarisation). In this case we would say the angle of incidence is 1 degree.

So going through the same calcs, 20 log (sin(01) = -35db. This is a lot of loss.

And an intermediate result for a 10 degree incidence: -15 db.

Reversing the process to calculate a 10 db loss: 18 degrees. And while that polarisation has 10 db loss, the other polarisation is only 0.4 db off the ideal, which is essentially unmeasurable by someone on a hill.

By calculating for each angle of a certain antenna tilt, you could accept a certain amount of loss on one polarisation (say for FM contacts with locals) while preserving almost all of the antenna gain for dx contacts on ssb.

This works for a normally “vertical” antenna being used to work horizontally polarised stations, with say an 18 degree tilt being enough to work those stations with only 10 db loss. At the same time the vertical antenna would still work the FM locals with only 0.4 db of loss.

Conversely, a beam that is needed to work ssb stations at 400 km can be tilted by 18 degrees, causing only 0.4 db of loss, and accepting that the locals on FM (who are always strong anyway) will be 10 db weaker than they would if you changed the antenna to vertical polarisation.

Finally, regarding the 35 db of loss from the 89 degree offset, it is quite difficult to be that accurate. At a 2 degree error, the loss drops to 30 db. At 3 degrees it is 25db.

So it is entirely feasible that due to errors in antenna setup, the sensitivity of the antenna to polarisation could be well below the theoretical maximum which is infinite.

Submitted to my learned friends for consideration.

Andrew VK1DA/VK2UH

Footnotes: * couldn’t resist** this one…

** see *


Reflecting upon your thoughts, and directing[2] mine to consider the practicalities of this, it might be instructive to consider that antennas oftentimes have some vertical and horizontal components to their radiation pattern, no matter their orientation.

[2] This is going downhill quickly.

Indeed, and this whole topic is not relevant to dual polarised antennas as they aren’t as critical in their polarisation selectivity. A diamond shaped quad for example has dual 45 and 135 degree polarised waves and works equally well on both polarisations, due to the 3db effect noted earlier.

10 points for the footnote, you picked up the thread nicely! :wink:

I should add that this analysis does not apply to HF or to any frequency where ionospheric reflection has occurred.

Do not worry about antenna polarisation on HF. Signals reflected by the ionosphere (E or F layer) come back to earth in multiple polarisations according to the experts.

I’ve been advocating slant polarisation for mixed mode activations for quite a few years now, so I am pleased that you have confirmed my reasoning. I think that there are two little problems that stand in the way of more general acceptance of the idea. Firstly, many if not most antennas sold today have square section booms, which makes fitting to a mast a matter of ingenuity. Secondly there is the problem of slant to slant communication, get it wrong and the losses become intolerable. Still, it seems that slant works well for contacting both horizontally and vertically polarised stations on line of sight paths.

1 Like

Surely the slant to slant scenario is only the same as a horizontal to vertical mismatch. I had always taken that to be -20dB, though Andrew’s calculation suggests a worse situation. Of course, these are theoretical figures based on a free-space set up which will never occur in practice where the situation could be better or worse since the topography on the path between the two stations will have an effect.

Disregarding specialist installations such as those for EME and satellite work, there will never be many fixed stations using anything other than horizontally or vertically polarised antennas. I would therefore suggest that using 45 degree slant polarisation in the field is unlikely to be a problem and the 3dB loss is more than compensated for by the convenience of being able to switch mode without having to adjust the antenna… as indeed I have found to be the case in practice.


1992-ish RSGB 432MHz FM contest, located Cyrn-y-Brain GW/NW-043 with a Tonna 21 ele horizontal and Tonna 19 ele vertical with remote switching between antennas and switchable masthead “hot-as-hell” GasFET preamp at about 15m.

FT736 with squelch adjusted so it was just flickering between open and closed. Being FM contest, normal search pattern was to leave vertical selected. Station called in from Birmingham way and squelch flicker rate changed so it was open more of the time. That was the only indication there may be a signal. Preamp off and flicker return to usual pattern, preamp on and squelch more open than closed. Tweak rotator but still no discernable signal other than flicker of squelch.

Flip to horizontal and there was an audible click as the FT736 S-meter hit the end stop and other station continued calling us. Other station had a single long horizontal Yagi running c. 25W up moderately high and could not hear us when we were on vertical but a local told him our QRG and he could hear others work us and called us blind-ish. He was S9++++ with the correct polarisation. Looked like the path and it was LOS so with both antennas well clear of of the ground there were few reflections so essentially no workable vertical component to signals.

Maybe our vertical system was pants… well if we didn’t win the contest we were top 3 so probably quite good.

Such path losses can occur in reality. For stations located in a built up area, there is normally sufficient local scatter from nearby roofs etc. that you don’t see the extreme so often.

I was thinking more of the special case of summit to summit contacts. A loss of 20dB, let alone 35dB, is liable to cost you the contact.

I had this Vertical / Horizontal issue demonstrated to me on 23cm. Normally I use Vertical because most of my QSOs are to a handi. I tried to work GI6ATZ (mega power Horizontal) from Dent G/LD-045 not a squeek, as soon as I changed to Horizontal 59+++. I had always assumed my poorly constructed antenna would not be that pure in its radiation pattern but not so.

I use a SG-Labs transverter on the back of the 23cm antenna. The driver is normally a 2m Handi.

73 de

Andrew G4VFL

Well VHF S2S contacts are a rarity in my activation logs, so I would not be concerned about that problem. If the idea was taken up by everyone, then there would be an issue to be contended with, but I very much doubt that will happen.

Indeed, my handheld quadruple quad proved to be an excellent antenna to facilitate a quick polarisation change when I ran 23cm from many a Lancashire and Cumbria summit. I must get back up there with the kit sometime (Standard C710 triband transceiver).

What you need is crossed slant polarised antennas. LOL.

1 Like

That are infinately variable…

1 Like

The general solution is a circularly polarised antenna with switchable circularity. That is, two antennas with a switch to change the offset from a quarter wave to three quarters.

Two antennas with a switch seems simpler.

Andrew VK1DA/VK2UH

It’ll need a few stepper motors as well I reckon…


The circularly polarised antenna wouldn’t need stepper motors.

Wait - which “with a switch” were you quoting?

Anyway I suspect you are pulling my leg, as usual…

I get around this problem by plugging the vertical into an HT and connecting the horizontal to the 2m ssb rig. Then turning down the HT so I can hear the dx.

73 Andrew VK1DA/VK2UH


I have a pair (or is that 2 pairs?) of Maspro circular polarised antennas for satellite use, VHF/UHF. There can be an amazing (S1 to S9) difference on the sat receive when I press the switch to change polarisation from L to R.
Just waving around an Arrow style yagi or pulling on a J pole till the signal is optimised seems to work on a summit.

Stepper motors for your legs perhaps??


My steppers have no motors. Is that the problem?

3dB loss to linear polarised users.

Whether horizontal or vertical, though, the same effect as slanting at 45 degrees. Basically a sacrifice for occasions like using SSB on V/UHF when you can hope for calls from people who only have vertical antennas, for instance it might be useful for the 70cm flavour in July.

Slant only works for one station. Unless all others use a single polarisation, when you have multiple stations using slant, you have once again, 50% probability of high attenuation when you both rotate 45 degrees clockwise. Your clockwise is the other guy’s anticlockwise.

1 Like