Building the 8+8 Slotted Waveguide

We used 100mm x 50mm rectangular hollow box section aluminium with 3mm thick walls. This gives us an internal size of 94mm x 44mm. We chose to build the antenna for 2.442GHz, or channel 7, as this is close to the center of the 13 channels available to us (US designs use channel 6 (2.437Ghz), as they have 11 channels). The internal width (the wide side), determines the operating frequency. The internal height (the narrow side) determines the impedance. We used Rob Clark's 5 wavelength design over Trevor Marshall's 4.75 wavelength design simply because we could cut the 5 wavelength design down to get the 4.75 wavelength one if our first try didn't work. Trevor's design also has wider slots and our slots could be made wider if we messed them up too much. We used Paul Wade's spreadsheet (Slotantenna.xls)to calculate the slot sizes and placement (The "Improved from Elliot" figures with a 2% reduction in slot length for square ends (see his text)). I have included a key lengths calculator in the form below.

Frequency GHz
Wide Sidemm (Inside Measurement)
Short Sidemm (Inside Measurement)
Material Thicknessmm
NSlotsBoth Sides
 

Free Space
Wavelength
mm
Waveguide
Wavelength
mm
Probe Lengthmm
Slot Lengthmm
Slot widthmm
Slot Offsetmm (From Center to Slot Center)
Slot Spacingmm (centers)
Slot to Topmm (center to top Plate)
Slot to Probemm (center to probe center)
Probe to Bottommm (Probe center to Bottom Plate)
Inside Lengthmm (Top to Bottom Plate)
Outside Lengthmm (Top to Bottom)

The final product only needs to be accurate to about 1mm as this will only shift the frequency within the range of the 802.11 spectrum. A smaller slot length will shift the frequency response of that slot to a higher channel. A longer slot length will shift it down. Changes in the slot spacing affects the vertical signal pattern. A smaller spacing (higher channel) results in the lower channels having a slightly downward virtical pattern. A larger spacing (A lower channel) will result in higher channels having a slightly upward tilted vertical pattern. Slot width variation affect the frequency response. The distance of each slot from the center varies with design. I don't know what this affects. One design deliberately changes this distance for each slot so as to reduce the side lobes of the vertical radiation pattern.

Methods and madness

I measured to .1mm using vernier callipers, on the assumption that errors in cutting might bring that closer to the 1mm error mark. I measured and marked up the front first and used a square to copy the lines around the side faces and onto the other side to ensure the slots on both faces lined up. I did in fact manage to get most slots cut to a few tenths of a mm in length. I used a drill press to make holes at either end of each slot and a hand held router with built on guide to cut between them. I then used a file to square the ends of each slot and to clean the edges up. I did discover how easily a router with can eat holes in 3mm aluminium. Some of my slots are slightly wider in places. Tungsten router bits and tungsten circular saw blades will happily cut aluminium (and you).

Misc Notes

Step by Step pictorial guide

A view from the Front. Note that you can see through the slots (i.e they are opposite one another).

For a step by step pictorial guide, follow me. The same contruction technique is common to both the single and double sided wave guide antenna.