Month: November 2017

I Love Antennas!

A little history first, then on to the good stuff.

Sometime in the spring of 1999 I found myself sitting in the back of a pickup truck driving down the Las Vegas Strip.  My purpose of sitting in the truck bed was to secure a $20,000 antenna that was destined to be installed at the very top of the MGM Grand the next day.  Our task was to transport it from the shop to the roof and prep the site for the final installation.  This antenna was an omnidirectional antenna, roughly 22 feet long, and if I remember correctly, operating in the 150 to 160 MHz range.  We were so concerned about damaging said antenna that we were afraid to strap it down in case we hit a bump and the antenna flexed and damaged an element inside.  As the new guy, I became the “strap” that could keep it in place and provide real time feedback to the driver to adjust to what the antenna was doing.  How we managed to get an antenna 22 feet long to the roof is a different story for a different time, seeing as it wouldn’t fit in any elevator that the MGM Grand had, but that was the first time I realized exactly how important, and expensive, antennas can be.

Now to present day.

I have been thinking about this post for a while but wasn’t sure how to start and what I wanted it to look like and blah, blah, blah.  Earlier this week I got an email as part of a chain and the sender had a question.  He was designing a Wi-Fi deployment for an area that was designed to be used for conferences and events, had a rather high ceiling, and wanted some feedback on some antennas he had researched.  That got me thinking of the whole process of how you get to the point you are trying to select external antennas.  For me, it’s second nature since I have been dealing with external antennas for almost 25 years but for others, it’s not as intuitive.

During WLPC_EU Lisbon in 2017, Carter Burke gave a Ten Talk where he started out by saying that “it’s all about the antennas.”  I 100% agree with him and if you haven’t watched his presentation, go check it out on YouTube, it’s worth your time.

The reason we feel like this is because in a wireless world, antennas are how the devices see each other.  Everything else can be spot on but if the antenna isn’t great, what does it matter if you have a 1,000 Watt transmitter with a receive sensitivity of -150 dBm and 24 chips on 3 circuit boards dedicated to doing great things if you have a crappy antenna with a bad ground and damaged elements?  Garbage in (due to a bad antenna) WILL result in garbage out.  It’s a simple equation we are all familiar with.  Early on in my Wi-Fi career I was able to convince people I knew what I was doing by simply suggesting new antennas for the existing AP’s.  The AP’s were old and not great, but antennas are always cheaper than the AP and I knew I could make a substantial difference with a minimal financial investment and some installation tweaks.  I was correct and here I sit.

Back to the email from earlier this week.  I realized that for my friend, I needed to back him up a little and start where everything starts, at the basics.  Just as a bad antenna can hinder a great radio; bad design, flawed fundamentals and incorrect hardware selection can’t be solved easily.  My advice to him was start with collecting requirements.  Just because an area is large with high ceilings doesn’t mean you jump straight to the chapter on stadium design and start there.  If there aren’t going to be a lot of devices to support, why use a lot of AP’s with high gain (read: narrow beam widths) if the requirements don’t call for it?  Sure, some areas will immediately scream for external antennas, but without additional requirements defined, how do you know if you need a 27 dBi parabolic dish antenna versus a simple low gain omni?  For the sake of this example, and based on experience, I know that my buddy will be mounting the AP’s in the ceiling (later found out it was a ceiling height of about 50 feet high) with patch antennas pointing down; some straight, others might be at angles depending on location.

Now the scary part.  How many and what type of antenna to use?  Even with all my experience with antennas, I still don’t have enough information to make that call.  Turns out most everyone knows about a tool that makes this easy and it’s called Ekahau.  Unless you are so confident in your abilities to design a space without using such a great tool, use it.  Granted, the parameters will be different than if you are designing a single story, standalone office space, but the output will still be very familiar.  Now, if you are doing a stadium with multiple floors surrounding a vast open space with plenty of crazy angles and the like, I would suggest upgrading to a product called iBwave.  It’s like Ekahau but designed specifically with stadiums in mind.  With our current example, my buddy doesn’t need iBwave, Ekahau will work just great.

When it comes time to define a radiation pattern and gain for the antenna, this is when things get tricky, and honestly I have wasted all this time just to get to this point.  There are a lot of elements (pun intended) to consider when making your selection.  I’m going to try and cover some of the basics and hopefully take some of the mystery out of this equation and lastly, give you my personal opinions on why I make the selection that I do.

  1. There is one fundamental truth that it pays to remember.  To increase the gain of an antenna, something else has to give.  For patch/directional antennas, that means a reduction in beam width, both horizontal and vertical.  For omni antennas, it means a reduction in “vertical” beam width since the horizontal pattern is defined in the name.
  2. Take the time to learn how to read the spec sheets for not only the radiation pattern, but for also the physical characteristics.  For outdoor antennas this means wind loading, temperature rating, and water resistance.  Most antennas that are outdoor rated pretty much have the same type of mounting solution (pole mount) but pay attention to the size of the pole the mount will accommodate.  For a simple Rohn Tower this generally isn’t an issue but for other types of locations, this can become an issue.  For indoor applications, aesthetics immediately come in to play and most solutions are driven off of this requirement, not what you need to do to make a great system/experience for the end users.
  3. Don’t be afraid to ask questions!  If something doesn’t seem “correct” about the product you are investigating, ask the vendor!  Remember, when using external antennas making an incorrect selection can destroy everything else that you have done to this point and everything else that comes after it; up to and possibly including someone purchasing the correct antenna and then swapping them out for the original hardware selection.  For a small job, and depending on the contract, that could wipe out any and all profit for that job.  Not a good feeling my friend.

Omni Directional Antennas

All antennas are gauged off what’s called an Isotropic Radiator, pictured below.  An Isotropic Radiator is theoretical perfection of a radiated signal.  Key term here is theoretical.  It doesn’t exist outside a lab or in the cubical I resign myself to occupy on a more constant basis.

Isotropic Radiator

It does, however, give us something to compare the antennas that are used in the real world.  Ever hear the term “dBi”?  Generally used when defining the gain of an antenna, it’s referring to the differential when compared to an Isotropic Radiator.  Since the majority of us don’t live in a clean room or Faraday Cage, we use actual real world antennas that differ from the picture above.  For this section on omni’s we will use the following examples:

From left to right, we are looking at a low gain omni, a high gain omni, and then a semi-hemispherical antenna pattern chart.  While most people like the cool graphics with the colors, I prefer the boring charts above the pretty pictures.  Either way, these are good representations of what happens when the radiating element(s) is (are) manipulated during the design process to accommodate different requirements.  It might be hard to see, but the image on the right, a semi-hemispherical antenna, shows what happens when you want the same idea as a low gain omni directional antenna but want to push most of the power in one vertical direction (think ceiling mount pointing down).  The difficulty in envisioning this antenna installed is the chart shows the antenna sitting on a table pointing up, not down.  The takeaway from these images are as the gain is increased, the elevation (vertical) plane is pressed down in the center to push the edges out further.  I think most people are pretty comfortable with the concept of omni directional antennas, so I will move on.

Directional Antennas

While studying for my CWNP CWDP exam, I ran across the term “semi-directional” and it hurt my brain.  I asked around to some of my radio buddies and they hated that term as much as I do.  Everyone agrees there are omni directional antennas (at least I think everyone agrees) and then there are antennas that are “not omni directional.”  If we negate the omni (omni meaning “in all ways or places”) then we are left with directional.

That’s it.

There is a defined direction the signal will propagate.  If you accept the term “semi-directional” now you need to define a term to define an antenna that isn’t omni and isn’t “semi-directional.”  That’s just stupid and I refuse to accept that.  What term are you going to use, “mostly directional”?  What’s after “mostly directional”, “all directional”?  All directional already has a term, omni, and since our not omni directional antennas are what we now trying to define, we can’t use that term.  I do, however, accept and use the term semi-hemispherical, as seen above.  I will now put away my soapbox and get on with things.

Directional antennas come in all sizes, shapes, and radiating patterns.  For most people, this is where things get confusing and they punt and select an antenna from the middle of the list, thinking it’s not the worst but not the best, so it “should” work.  If you have made it this far through my ramblings, have hope!  This is where it gets good and fun!  When I select an antenna, and antenna vendor, I look at the line drawings they provide.  Most AP’s that are deployed today have more than one antenna port, meaning the antenna you select should have more than one element.  If the vendor can’t or won’t provide a chart showing each element individually, how much do you trust their design or understanding of the product they are wanting you to pay money for.  I like charts that show me all the elements.

I know, some of you will puff out your chest and say “but Jim, I have locations where I have a 4 or 6 port AP and I only used one or two single element antennas!”  Bully for you, I have done that to, but you better be sure you know what you are doing if you want to pull that off.  If there is enough interest, I can go into that as well, but I ask one question.

Why?

My reasons generally center around existing single element antennas that were kept in place due to budgetary concerns but for new installations, either pick a different AP / radio type or design the installation site to use all available antenna ports.  It’s like buying a nice new truck with a V8 engine and then asking the mechanics to disable 2 cylinders before you drive it off the lot.  If you only wanted / needed a V6, buy the V6!  If you want a 2 stroke single cylinder engine (like my lawnmower), but one of those.  It won’t do much to get me to the office each day but it mows my lawn quite nice!

Dual Single Line

For the next couple of minutes, I will use the pattern shown above.  There are some basics that you can tell by this drawing, let’s walk through them.  All of these directions are based off the center front of the antenna, and most antennas should have an arrow that defines which direction is up.  Imagine yourself standing in the middle of the circle, holding the antenna up against your chest like a spotlight, with the up arrow pointing up.  The chart then explains what the patterns will look like.  The numbers around the outside of the circle are degrees, like on a compass, and the gradients inside the circle represent the drop in RF signal from the maximum (always at zero) and these numbers are in dB.  Actual transmit power is NEVER defined since it’s all based on the transmitter, just know that zero is maximum and as you spin left and right, it starts to drop off.

  1. Read the top line.  This is from Cisco, but all diagrams should define which band the drawing is from.  If it doesn’t, punch out now!  They really don’t know what they are doing and who knows if anything else is correct and what will happen when you install it.
  2. Figure out what the color lines are.  In this example, they combined both planes into a single drawing.  While not my favorite way of doing it, I can live with this presentation.  The azimuth plane defines horizontal, or left to right, and the elevation plane defines just that, the elevation or up and down.
  3. When manufacturers define the parameters of an antenna, they use a common figure to define the numbers they print on the specification sheet.  This figure, -3 dB, is the standard and is also referred to as “3 dB down.”  This is when the radiating signal drops 3 dB below the maximum RF level.  In the above example, the maximum RF level is at zero degrees, so directly in front of the antenna.  When looking left and right of the center line and along the blue line, we can only guess that it crosses the -3 line (or 3 dB down) at roughly 50 to the right, and 310 to the left.  The absolute difference between these numbers, transitioning through 360 degrees, is 100 degrees.  This defines the horizontal beam width of the antenna at 100 degrees.  Do the same thing for the red line, the elevation, and we can compute that the vertical beam width of this particular antenna is about 60 degrees, a rectangle much like the shape of the plastic radome that covers the elements.  Now, if this vendor had put more effort into the drawing, and trust into the designers who will add this antenna to their system, they would have added a line specifically at -3, not made you guess.

Roll Off

Now that we know the published specs for the antenna, lets talk about one more element that is easy to see in the drawing but hard to consider in real life.  My terminology for this element is the “roll off” of the antenna.  While 3 dB down is a great metric to measure off of, it doesn’t mean the signal stops at the 3 dB down mark.  It keeps going.  If you take a look at the blue line above, it keeps going farther to the left and right before it starts to pull back towards the antenna.  For this particular antenna, you get almost all the way back to the radiating elements before it bothers to cross the 10 dB down line!  At 10 dB down, this is actually closer to a 160 degree antenna.  At 2.4, where most people run with fewer transmitters and a little higher power, this isn’t good.  Consider a transmit power of 20 dB (100 mW) and the gain of this antenna is 6 dBi.  Assuming no loss in the cable or connector, we now have an effective radiating power (ERP) of 26 dB.  10 dB down from that is 16 dBm.  Using the numbers from item 3 above, 3 dB down is at 50 degrees and 10 db down is at about 80 degrees.  This differential gives us a roll off of about 30 degrees.  The smaller the roll off number, the tighter the radiating cell, and the more isolation the antenna provides in relation to it’s neighboring antennas.  In an environment with mobile devices, this antenna can add almost 60 degrees to it’s published horizontal band width specification and keep clients connected to AP’s that you don’t want them to.

Look, I really love external antennas, but using them blindly or at the recommendation of someone who doesn’t know any better can really mess up what you planned out to be a great RF environment.  The scary part is for the pattern above, the 5 GHz band is worse on roll off in the horizontal plane.  It’s almost a 180 degree antenna at 10 dB down!  Sadly, the elevation plane is pretty tight so I could really envision a deployment where I would install this antenna rotated 90 degrees out of the vendor recommendation, just to keep the cruff on the horizontal plane from interfering with it’s neighbors in the same elevation.  Now, this antenna has horrible mounting solutions if you want to rotate it 90 degrees, so that sucks.  The solution, after market antennas!

My buddy Mike once told a vendor that until the support people can remote or SSH into our antennas, he will use whatever antenna he feels best and not bother to mention that we aren’t using “their” antennas.  It adds a little more effort on our part, but he has a point.  The workaround is to build a chart that matches the radiating specs of the antenna we install to the vendor antenna so when they ask us which antenna we installed, we give them their part number.  Since they can’t SSH to the antenna, and they aren’t going to fly out here to verify it, no problem!

Now that we are moving to more advance stuff, let’s look at some drawings from some after market antennas.

Single Line drawingsDirectional Horizontal patternsDirectional Vertical patterns

These are small, and they suck like this, but I have a point.  The set on the top are decent, but some graphics designer got a hold of them and made them look hip, and I can tell you from looking at the full size of them, you can’t tell anything from them.  The drawing is almost pointless!  The shading and multiple lines make it hard to read and the lack of degrees around the outside again make it hard to read.  The fact that the company let this happen is discouraging!  Antennas aren’t sexy, but to those in know we don’t need pretty to do our jobs!  I also want to point out that of all the examples I’ve posted so far, only one has shown multiple lines even though all 3 antennas are 4 element, dual band antennas with roughly the same horizontal and vertical beam width specs.  Only one of the three vendors thought enough about the designer who is making the antenna selection to give them a chart that makes sense!  The top example shows it all the way down to the -50 dB marker, and that’s just stupid!  It’s pointless and only muddies the picture.  My favorite is the second set, let me explain why.

  1. The company thought it through and actually drew a 3 dB down line on the chart for us, no more guessing.
  2. Instead of 90 degree or 45 degree markers around the outside, they gave us 30 degree markers, easier to read and not as much guessing.
  3. They included all 4 elements.  The dirty truth is not every element is perfect, and enclosing 4 in one space means they feed off of each other, even if they aren’t touching.  To assume they all work exactly the same is naive and not showing all 4 makes me wonder what they are hiding.  Is every element perfect in these drawings?  No.  Do I care?  A little bit, but I at least have all the information to make my own decision.  This vendor isn’t hiding anything by creating an average of all 4 elements so I can’t see the massive null in the 5 GHz vertical plane from Lead 1 and Lead 2.  They offset each other but now I know how they adjusted the design to compensate for the null.  This vendor trusts their product so much they are showing the actual sweep, and I respect that.

What’s a null you ask?  A natural “hole” in the pattern that happens in antennas that aren’t isotropic radiators.  This null is also what creates what are known as lobes; both side lobes and back lobes.  Anytime the null comes significantly in towards the radiating element it creates “wings” in the pattern and the wings are known as lobes.  If the lobe occurs forward of the center line (in this case the +90 and -90 line) it’s known as a side lobe, behind that center line it’s known as a back lobe.

Top Tips

  • Each vendor does their drawings different, especially when it comes to labeling and layout.  Not sure which one you are looking at, try these cheats:
  1. They will always show a vertical and horizontal plane.  Figure out one label (“E Plane”?  No idea) and the other one is the other one.  H Plane is horizontal because nothing else makes sense so E Plane must be elevation (or vertical) plane.
  2. Still confused, find the 3 dB down line on the drawing and figure out the degrees of that element.  3 dB down at 55 degrees means it’s probably a 110 degree graph.  Go to the specs and figure out which spec is 110 degrees and match it up from there.
  3. Crazy pictures like the last set are real life, embrace it and learn to love it.
  • When in doubt run your own tests.  Install 2 different types of antennas with similar specification and walk around with a device that measures signal.  As you expand your distance from the antenna, write down the signal.  When you see the significant drop in signal, you’ve now reached the edge of your cell.  Since we discussed roll off earlier, you have now discovered that in real life.
  • Not everything that is good in one location will be good in another.  Treat each location as it’s own and re-evaluate on it’s own merit.  Internal antenna AP’s don’t give you much flexibility in this department so the planning and workload isn’t as much.  External antennas can solve a lot of problems but not taking the time to select the correct antenna for each location can also induce a lot of problems.  Antenna isolation and roll off might work well in one environment but may be problematic in a different environment with a different use case.
  • Never forget to follow best practices when designing with external antennas.  Yes, it gives you a lot of flexibility but skipping basic design steps like identifying use cases and device type will come back to haunt you and leave you answering a lot of questions
  • This is a repeat tip, but ASK!  Reach out on Social Media, ask the vendor, ask the manufacturer, phone a friend, doesn’t matter.  Physics define how most of this works so most answers should be similar in nature, if not even the details.
  • Consider the mounting options from each vendor.  If you think the antennas are different, think about the places you are asking someone to install them, if it isn’t you climbing the ladder. Multiple and flexible mounting solutions means less specialized hardware to order and greater options when your original plan flies out the window while you are 30 feet in the air in a cherry picker.
  • Consider cable type and length.  An extra 10 feet of Cat6 cable doesn’t have as much loss as an extra 10 feet of low grade coax cable.  When we are talking about receive signals of -76 dBm, a couple dB means a lot!

Conclusion

This little post was never intended to be the end all be all about antennas.  This was my attempt to take some of my knowledge about antennas and get it written down to help out a friend.  I could go on further but I think this works for now.  The antennas I envisioned while typing this up were the more common types I use but I have and use a myriad of other types.  If I could make one recommendation for you to take away from this it would be this: respect antennas, but don’t be afraid of them.  Get some different types of antennas and do your own testing.  Figure out what works best and then go give it a shot.  There might be some stumbles along the way but the more comfortable you get with these the better the end result will be.

What’s your experience or stories with external antennas?  I would love to hear them!

Advertisements