Tech: DIY Parabolic Reflectors / Wi-Fi Antenna Boosters

First of all, the phrase "parabolic reflector" is just damn cool. "So what'd you do today?" "Eh, just made some homemade parabolic reflectors to boost my wi-fi network signal."

Secondly, parabolic reflectors LOOK damn cool.

Third, they really work.

While researching how to make my wireless network more reliable I came across a link with information on making cheap, effective antenna boosters.
first see: Deep Dish Cylindrical Parabolic Template
then see: Ez-12 Parabolic Reflector Template

The idea behind a parabolic reflector is that it reflects signals back to a specific focal point - the antenna. Any signal that hits the reflector (assuming it's not an extreme angle) will be redirected towards the focal point. This has the effect of collecting a broad signal and concentrating it at the antenna, making it easier for the antenna to receive. This is an inherent property of parabolas, if you remember your Algebra 1 or Algebra 2 days in high school.

It also works in reverse. A signal emanating from the antenna will strike the parabolic reflector and be bounced back out in the direction the reflector is pointing. This has the effect of minimizing the strength of the signal behind the reflector (useful for wi-fi security - why send a signal into your neighbors' house or out onto the sidewalk?). But because it's a reflector (as opposed to an absorber), it also strengthens the signal in the direction of the reflector. All that energy is bounced away from the undesired direction and towards the intended direction.

The end result is that you get a stronger signal where you want it, you enhance security by containing your transmissions, and you enhance the antenna's ability to collect and receive your incoming requests. Quite a miracle all-in-one solution!

I began with Erskine's EZ-12 template linked above (the EZ-12 is an enhancement on his earlier design, but it helps to read about the old design first since he doesn't duplicate the info on the newer EZ-12 page). It's a very easy solution. You basically print out his template onto a piece of cardstock paper and cut out the two sections. You wrap the rectangular section with aluminum foil and then curve it around the funny-shaped parabola guide/spine. Cut out the focal point Xs and feed your antenna through the holes.

It does work and can be confirmed by an increased signal measured by NetStumbler (see the earlier post about wi-fi networks).

The downside is that it's a bit fragile and imprecise. The shape of the parabolic curve and the position of the focus point (i.e. antenna) are crucial. The card stock spine just isn't strong enough or precise enough - it tends to sit crooked on the antenna and the parabola its contour traces out is a bit compromised by the six insertion slot tabs. The top and bottom edges of the reflector also might deviate from the proper curve as there is no guide to shape them.

Therefore I created my own Do-It-Yourself parabolic antenna design, using the EZ-12 template as my starting point.

I traced out two copies of Erskine's parabolic curve separated by about 1/8" onto a section of cardboard. It's also important to remember to draw in the focal point.

I then used an X-acto knife to cut out the curve and the cardboard piece's exterior border. I duplicated this process and ended up with three pieces of cardboard that were all the exact same size and shape.

I measured the width of my wireless router's antenna and made appropriate mounting holes in two of the cardboard guides (the antenna has two parts - a fat bottom and a slender top. I measured such that one cardboard piece would sit at the bottom of the fat end of the antenna while the middle piece would sit at the base of the slender portion of the antenna). You have to be careful to make sure that the focal point sits at the center of the hole you cut for the antenna.

My original plan was to then connect them together using disposable chopsticks as a spine. But then inspiration hit.

I realized I could just cut notches into the cardboard and build a notched spine that would interlock and hold the structure together. This has the benefit of simplicity and makes for a clean, strong connection. The problem with taping chopsticks to the cardboard is that there isn't a good way to achieve a strong perpindicular connection between the two. The rigidity and right-angled nature of cardboard make it perfect for this application.

Once assembled the whole thing looks rather awesome. It was surprisingly rigid - I'd cut the interlocking spine notches just right so that the fit was just tight enough. I didn't even need any tape or glue to hold it together. The angles were mostly perpindicular, though there was some error factor.

At this point I took the EZ-12 design's rectangular reflector piece (though elongated to make it taller) and printed it out on a piece of cardstock. It's important to note that you must match the size of the parabola traced onto the cardboard with the proper width for the rectangular reflector (otherwise it won't fit width-wise). And it's not enough to just measure the width of the curve from endpoint to endpoint - the reflector will end up being too small if you try that.

If you use the EZ-12's curve (take the top edge, ignoring the three mounting tabs, all the way from edge to edge, stopping when it curves back inwards), the resulting parabola will be perfectly fitted to the rectangle piece in the template. However, give yourself some extra wiggle room as nothing is ever quite so precise. And I found the EZ-12's curve to be a bit imprecise. I ended up using the curve provided in his earlier template. Just remember that the rectangular reflector must match the size of the curve you trace - I used Photoshop resizing to match the first template's curve with the width of the rectangular reflector from the second template. Either get the sizes correct or just guess and check the reflector width against your cardboard guides.

Wrap the rectangular reflector piece with alumninum foil. Pre-curve the cardstock before you wrap the foil. Foil doesn't stretch, so it won't take to bending very well if you apply it flat and then try to curve the reflector. You can apply either the front or the backside of the reflector. The back is easier but it looks much cooler if you do the inside.

I opted for the inside. I taped down the left side of the foil to the pre-curved cardstock. Then I laid the foil down inside the curving paper. I didn't have any kind of glue that would dry flat (you don't want to use glue that will create ridges in the foil) so I had to find some way to tape the foil down without wrinkling it and make sure it stayed flat against the inside of the curve.

As you can see in the picture I cut vertical slats in the top and bottom of the foil up to the point where it met the cardstock. I then wrapped each slat around the cardstock. The slats allow the foil to fan out as it's wrapped over the edge of the curving cardstock. Try it without the slats and you'll see what I mean. They also keep each section of the foil tight to the inside of the curve (more or less). From there it's an easy job to tape down the remaining right section of the foil.

That accomplished, it's time to feed the reflector through the cardboard guide array. Spiffy, eh? I love this photo - I inadvertantly captured the light bouncing off the reflector and collecting at the focal point! It's especially obvious in the middle cardboard piece.



I made two of these reflectors and mounted them on my wireless router. The antenna mounting holes were just the right size to make for a snug, secure fit.



After some testing and re-aiming, I realized a definite improvement in my wireless signal. In truth the simple EZ-12 reflectors also did an excellent job, but my design ended up being a lot more precise and a lot easier to make small aim adjustments. My overall signal gain was about 8dB - not quite the 12dB advertised in the EZ-12, but I'm measuring through the floor and at least one wall (my router is upstairs in my bedroom. The wireless digital music server will be downstairs in my living room).

The biggest difference between the EZ-12 and my design was signal stability. Though the signal was improved by the EZ-12, the strength meter bounced around quite a bit. With my design (and lots of careful, precise aiming) the signal has steadied significantly.

The other nice thing about this design is that it's easy to just slide new reflector materials into the array. I'd eventually like to replace the flimsy tin foil with a solid sheet of alumnimum roofing, mesh screens, or other similar material. As long as I can cut the width correctly I should be able to experiment with all sorts of different possibilities.

Anyway, I'm very happy with the results and I'm thrilled with the way these reflectors turned out. I love building things myself and I really didn't anticipate this by-the-seat-of-the-pants design to turn out so well.