Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Antenna Polarity question

Status
Not open for further replies.

mdwebster

New Member
Greetings,

I'm a bit stuck on a problem involving the interpretation of various radiation patterns that are depicted in antenna spec sheets. I get the general orientations (XY, YZ, XZ, or, alternately, Azimuth, Elevation (phi=0), Elevation (phi=90)), but the polarities are throwing me.

A typical spec sheet lists two curves on each of the three orientations, one for horizontal polarity and the other for vertical. I believe that the main determining factor for the polarity is the orientation of the longest element in the antenna (assuming linear polarity to start with).

So, if my assumption above is true, why does the XY-V cut not look like the XZ-H cut? Is it due to the position of and reflections from earth ground?

Thanks for any help in getting my head straight on this,
Mike
 
Last edited:
Greetings,

I'm a bit stuck on a problem involving the interpretation of various radiation patterns that are depicted in antenna spec sheets. I get the general orientations (XY, YZ, XZ, or, alternately, Azimuth, Elevation (phi=0), Elevation (phi=90)), but the polarities are throwing me.

A typical spec sheet lists two curves on each of the three orientations, one for horizontal polarity and the other for vertical. I believe that the main determining factor for the polarity is the orientation of the longest element in the antenna (assuming linear polarity to start with).

So, if my assumption above is true, why does the XY-V cut not look like the XZ-H cut? Is it due to the position of and reflections from earth ground?

Thanks for any help in getting my head straight on this,
Mike

good question.

The physics definition of polarization is the direction of the electric field wrt the earth's surface. now for a dipole held with the arms parallel to the earths surface, it will have an electric field horizontal to the earth. this is horzontal polarization.

but what becomes of a complex LHCP or RHCP antennas? you wouldn't be able to have a single polarization ref. so...

It depends on the user setup. usually on data sheets, the designer will note his frame of reference and the xy, y,z and xz planes wrt to the Rx antenna. He'll announce one position of the AUT as horizontally polarized and then he'll take plane cuts of its radiation pattern. likewise for its vertical position.

The patterns look different because the RX antenna is positioned in a different matter, thus the nulss and lobes of the antenna directivity will change.

so it depends. just read the data sheets. hope that helps.
 
Last edited:
Thanks for the reply.

I guess my main problem is that all of the spec sheets I've been reading don't state the initial assumptions of horizontal vs. vertical polarization. They show the cuts with regards to the antenna, so I assume that the "default" position is horizontal and that rotating the antenna 90° with respect to the ground is the vertical one (like if default were in XY plane, parallel to ground, the "vertical" measurement would be with the antenna rotated into the XZ plane).

Anyway, I've pretty much convinced myself that the big rocky ball underfoot is the culprit in the variations I was worried about. Again, thanks for the help.
 
Thanks for the reply.

I guess my main problem is that all of the spec sheets I've been reading don't state the initial assumptions of horizontal vs. vertical polarization. They show the cuts with regards to the antenna, so I assume that the "default" position is horizontal and that rotating the antenna 90° with respect to the ground is the vertical one (like if default were in XY plane, parallel to ground, the "vertical" measurement would be with the antenna rotated into the XZ plane).

Anyway, I've pretty much convinced myself that the big rocky ball underfoot is the culprit in the variations I was worried about. Again, thanks for the help.

thats strange. i have been working on some chip antennas and the manufacturer gives a diagram of the antenna's horzontal and vertical frame of reference and then gives the plane cuts.
 
I've got an unlisted (i.e. non-Internet) spec sheet from Fractus with no orientation.

There's this one from Johanson Technology with no real indication of H vs V.

There's **broken link removed** from Furukawa. It's not a full spec sheet, but the reference for H/V isn't obvious (to me, at least).

Some others: 1 , 2 , 3 , 4.

Finally! Here's **broken link removed** with the HV orientations listed, but I'm still having trouble interpreting it.

I just had a thought though. Is it possible the H&V listings are for the polarization of the TX antenna (assuming, for the sake of argument, we're taking measurements on the RX side)? Thinking about it, that seems fairly likely since I hadn't considered the "other side" of the link at all. And the TX polarity would have a significant effect on the reception.

Anyway, again, thanks for the help,
Mike
 
What funny little antennas, presumably intended for WiFi toys.
I thought you meant real antennas, nice big yagis with lots of metal in them!

OK referring to the Johanson datasheet, what they are saying is, with their little 2.4Ghz antenna oriented as shown in the diagram, they have gone around it with a test antenna and measuring receiver, measuring the relative field strength, and plotting the results as shown on the polar diagrams.

They went around each set-up twice, once with the test antenna vertically polarised and once with it horizontally polarised.

In the diagram the Z direction is presumably perpendicular to the earths surface, so vertical polarisation is then the electric field response of the test antenna is in the Z direction, and horizontal when it is at 90deg to the Z direction.

OK?

JimB
 
Yep, that's what I thought (or wound up thinking anyway)...

I appreciate the confirmation.

And yeah, they're baby antennas meant for 2.4GHz applications. The particular application I'm working on is only meant to have a min range of 40 feet and a max of 100 feet in a single room. (We actually DON'T want it to go through walls (much anyway, some is expected) to prevent it from interfering with other devices in neighboring rooms). I'm just trying to get my head straight on radiation patterns so I can (intelligently) choose 2 or 3 for prototype testing.
 
I've got an unlisted (i.e. non-Internet) spec sheet from Fractus with no orientation.

There's this one from Johanson Technology with no real indication of H vs V.

There's **broken link removed** from Furukawa. It's not a full spec sheet, but the reference for H/V isn't obvious (to me, at least).

Some others: 1 , 2 , 3 , 4.

Finally! Here's **broken link removed** with the HV orientations listed, but I'm still having trouble interpreting it.

I just had a thought though. Is it possible the H&V listings are for the polarization of the TX antenna (assuming, for the sake of argument, we're taking measurements on the RX side)? Thinking about it, that seems fairly likely since I hadn't considered the "other side" of the link at all. And the TX polarity would have a significant effect on the reception.

Anyway, again, thanks for the help,
Mike

Alright, so let's take a look at the first diagram, the z-x plane. We know from using the right-hand rule that the direction of the magnetic field wraps around the direction of current. i'm assuming the current will flow in the x direction predominatly, so the polarization of the chip antenna is horizontal wrt the earth and the way its stationed on the diagram( y axis is perpendicular to earth).

Since in that manner,the chip antenna is linearly polzarized, unless it has circularly or dual polzarization states, the Rx antenna, is the antenna that switches polzarization states. first the Rx and Tx antennas have the same polzarity, and the Rx antenna scans the azimuth and elevation angle planes. the Rx antenna, not the AUT, then switches polzarization, rotates 90 degrees wrt the normal and then rescans the azimuth and elevation planes. thats how i would interpret the data sheet.

so they're just getting 2-D images of the radiation pattern if any of the antennas are similar or cross-polarized. and that makes sense, since one would want the response of the chip antenna as the whole system undergoes real-time operations.

say for instance, reflections or scattering of the waves exist in a noisy office, that could contribute to a change in the polzarization state or if the user rotates the device so that it changes state, one would need to kow hoe the performance will be affected.
 
Now i have a question. i already know the answer, but this engineer i talked to seems like he doesnt know wtf he's talking about. or at least i haven't heard of it in school and the industry.

We know radiation patterns can tell us the relative field strength or power in some direction, xy, yz, zx or theta or phi. the pattern can be normalized, which it usually is, and it can be linear or it can be log based.

what am i getting at?

This one engineer told me, this one chip antenna is referenced to an isotropic antenna. fine, so its gain is below 0 dBm. so its lossy. so the company gives a picture of the orientation and then diagrams of the pattern ploats at different cuts.

However, he said the pattern on this one plot is a plot of constant gain around the antenna. I don't understand, i thought radiation patterns were plots of field strength around the antenna. Depending on the antenna type, its response will be stronger or weaker in certain directions.

I think it has to do with the surface charge density since, an area that has more surface charge induced on it, will have greater electric field strength. Also, the polarity(+/-) of the lobes dictate how and where lobes might form-theory. The impedance along some point on the surface also would have an affect. i have to read some more.

but anyways, he started blabbing on that to get the circle or so of constant gain, you have to move non-AUT antenna farther or closer in. i believe that, but thats not how tests are done right? I have done some work in anechoic chambers, and what they do is have the scan antenna, move in geometric patterns-planar, cylindrical, spherical or starburst.

We do this at a fixed distance, usually some where in the far field at some predicted field strength. and scan the antenna at that fixed distance. we don't move it in, becoz that'll throw off the readings anyways.

for example, let look here

on page 5, it gives the oreitnation and plane cuts of this particular antenna. so lets look at the aziumth, theta =90. although in my EM text books the azimuth is given in phi, but alas. anyways, the plot looks fairly uniform at 0 dBi. at 0 degress, the gain is 0 dBi but at 180 degress, it is a bit greater than -2.5 dBi. and by greater, i mean that a field strength at -1.5 dBi is greater than one at -2.5 dBi, not -3.5 dBi is greater than -2.5 dBi, which some folks get confused about the number line and log values.

so thats how i interpret it. the circle is not at constant gain. so am i reading it right? i just want some clarificatin. yeah thats alot.

Thanks though.

i'll go read JPUG.
 
Heres one more question/double check:

antennas are usually referenced to a dipole antenna or isotropic radiator.

I have computed the gain of a half-wave dipole to be 1.61-1.64 dB.

If I build an antenna, and it has a gain of 4dB :

1: if I reference it to dipole which is 1.61 dB. is the gain 4 dBd or 5.61 dBd or 2.39 dBd? some normalize the gain of a dipole to 0 dB and then

2: if i want it in dBi, i have to add 2.14 dB to it. so my antenna would have a gain of 6.14 dBi? if its gain =4dBd?

Its d@mn confusing and i already talked to a fellow engineer, and he gave me some bull$hit. i read books, and white papers and they bull$hit.

I want a straight concise answer with a few examples. also a why and how about why we would want to reference to a dipole and iso would be great but not required.

Thanks.
 
Well, I'm certainly no expert, but I'm inclined to agree that the engineer you were talking to was mistaken. I can somewhat see that the radiation pattern is proportional to a line of constant gain (where the line denoted distance), but it isn't measured that way at all. In fact, since signal drops off as an inverse square, and the scale is generally in dBi, you'd have to scale it by 1.059^(-signal) (where signal is in dB) to get distance.

As to your second post, my understanding is that you never get more power than you put in, but you can shape it so that you get more power in a particular direction than an isotropic antenna would deliver (or receive).

*Edited out stuff about isotropic radiator* You know, I'm not sure how they get the isotropic radiator baseline. Maybe they just use a 1/4 wave dipole then weight the expected signal strength to get back to an isotropic reference line.

*Added speculation* I'm guessing that they probably use an "omnidirectional" antenna, which is an antenna that has a constant gain pattern in one of its planes. As long as they line up the omni antenna so that the transmitter lies on its constant gain plane, it should effectively act like an isotropic antenna.

So you take your measurements around the isotropic antenna as a baseline then replace it with the AUT. Then you take the measurements again and relate it back to the isotropic antenna to get the dBi for that point. So the initial isotropic measurement is *defined* as 0dBi for every point where it's measured and the AUT is stacked up against that. If the reception is 2dBm better than the isotropic antenna at point A, then the AUT is given +2dBi at that point. If it's 10dBm worse, it's -10dBi. This also helps to throw out the variations inherent in the measurement conversion, the amplifiers, and any attenuation in the feedline.

So, long story short, keep in mind that your gain is only in a given direction. Your "real" measurement out of your feedline is going to be in milliwatts, converted to dBm (0 dBm = 1mW). So, if you have an antenna that, in a particular direction, picks up 4dBm where, in the same location, your dipole picked up 1.61dBm, then the relative gain of your AUT would be 2.39dBd.

Now, if the initial 1.61 "dB" dipole measurement were in dBi, then you could reference the AUT back to dBi by adding in the dipole gain.
 
Last edited:
Heres one more question/double check:

antennas are usually referenced to a dipole antenna or isotropic radiator.

I have computed the gain of a half-wave dipole to be 1.61-1.64 dB.

If I build an antenna, and it has a gain of 4dB :

1: if I reference it to dipole which is 1.61 dB. is the gain 4 dBd or 5.61 dBd or 2.39 dBd? some normalize the gain of a dipole to 0 dB and then

2: if i want it in dBi, i have to add 2.14 dB to it. so my antenna would have a gain of 6.14 dBi? if its gain =4dBd?

Its d@mn confusing and i already talked to a fellow engineer, and he gave me some bull$hit. i read books, and white papers and they bull$hit.

I want a straight concise answer with a few examples. also a why and how about why we would want to reference to a dipole and iso would be great but not required.

Thanks.


It's easiest to understand antenna gains considering a transmitting antenna.

The isotropic antenna is an idealised beast that simply radiates the energy equally in all directions. The power spreads out uniformly over the sphere, so if you transmit P watts, at a distance R the isotropic radiator results in P/(4*pi*R^2) watts per square metre.

Any real antenna sends more power in some directions than others. The gain in a particular direction relative to an isotropic antenna is simply how much power the antenna manages to send in that direction relative to the power an isotropic antenna would send in that direction. Antenna gain is like adding a reflector to a torch bulb, you send more power in some directions at the expense of other directions. Those preferred directions have more gain.

Antenna gains are typically measured by comparing the power transmitted by the test antenna compared to a reference antenna. If you used a dipole as the reference antenna you would measure gain relative to a dipole (e.g. 4dBd), if you add the dipole gain (2.14dBi) to it you get the antenna gain as 6.14dBi.

You need some reference, both from a measurement and also usage point of view. The most common usage is to work out the power received from a receiver antenna when a given power is transmitted - this is an important part of radio engineering. See Link budget - Wikipedia, the free encyclopedia

Peter
 
Now i have a question. i already know the answer, but this engineer i talked to seems like he doesnt know wtf he's talking about. or at least i haven't heard of it in school and the industry.

We know radiation patterns can tell us the relative field strength or power in some direction, xy, yz, zx or theta or phi. the pattern can be normalized, which it usually is, and it can be linear or it can be log based.

what am i getting at?

This one engineer told me, this one chip antenna is referenced to an isotropic antenna. fine, so its gain is below 0 dBm. so its lossy. so the company gives a picture of the orientation and then diagrams of the pattern ploats at different cuts.

However, he said the pattern on this one plot is a plot of constant gain around the antenna. I don't understand, i thought radiation patterns were plots of field strength around the antenna. Depending on the antenna type, its response will be stronger or weaker in certain directions.

I think it has to do with the surface charge density since, an area that has more surface charge induced on it, will have greater electric field strength. Also, the polarity(+/-) of the lobes dictate how and where lobes might form-theory. The impedance along some point on the surface also would have an affect. i have to read some more.

but anyways, he started blabbing on that to get the circle or so of constant gain, you have to move non-AUT antenna farther or closer in. i believe that, but thats not how tests are done right? I have done some work in anechoic chambers, and what they do is have the scan antenna, move in geometric patterns-planar, cylindrical, spherical or starburst.

We do this at a fixed distance, usually some where in the far field at some predicted field strength. and scan the antenna at that fixed distance. we don't move it in, becoz that'll throw off the readings anyways.

for example, let look here

on page 5, it gives the oreitnation and plane cuts of this particular antenna. so lets look at the aziumth, theta =90. although in my EM text books the azimuth is given in phi, but alas. anyways, the plot looks fairly uniform at 0 dBi. at 0 degress, the gain is 0 dBi but at 180 degress, it is a bit greater than -2.5 dBi. and by greater, i mean that a field strength at -1.5 dBi is greater than one at -2.5 dBi, not -3.5 dBi is greater than -2.5 dBi, which some folks get confused about the number line and log values.

so thats how i interpret it. the circle is not at constant gain. so am i reading it right? i just want some clarificatin. yeah thats alot.

Thanks though.

i'll go read JPUG.


Hi Q
I just read this thread and wanted to say something about your post. I too am puzzled about the reference to a constant gain plot. The term constant gain is often used in work with the Smith Chart, in, for example, the plotting of an amplifier's constant gain circles. In all my years of antenna work, I have not seen a "constant gain" plot of an antenna. In fact, it would seem that plotting constant gain would require 3-dimensional visualisation. Can you provide a copy of that controversial plot for us to interpret?

I also wanted to point out a little error in that you used the term " so its gain is below 0 dBm". This term doesn't make any sense, since gain is never described in units of dBm.

regards
Ron
 
Hi Q
I just read this thread and wanted to say something about your post. I too am puzzled about the reference to a constant gain plot. The term constant gain is often used in work with the Smith Chart, in, for example, the plotting of an amplifier's constant gain circles. In all my years of antenna work, I have not seen a "constant gain" plot of an antenna. In fact, it would seem that plotting constant gain would require 3-dimensional visualisation. Can you provide a copy of that controversial plot for us to interpret?

I also wanted to point out a little error in that you used the term " so its gain is below 0 dBm". This term doesn't make any sense, since gain is never described in units of dBm.

regards
Ron

The only plot i had for reference was the fractus chip antenna. Thanks for the clarification, I got my powers mixed in with my gain plots.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top