Slotted Waveguide Antennas

A couple of weeks ago I attended a wireless workshop, I took along a couple of antennas with the hope that we could do a couple of tests and compare them against an antenna with a known dB rating.
As luck would have it there were a couple of commercial antennas there and they were rated in the range I expected my Slotted Waveguide to be.

I managed to borrow a laptop from one of the guys and connected my 8 slot waveguide to the pigtail then set netstumbler up. at this point I realised that there were no AP's running so Noodles set his iPaq up in Ad-Hock mode.

The test started out well and Noodles watched as a took his iPaq for a wander down to the end of the hall, the communications path had to punch through 2 walls at this stage and we still had no loss of signal.

Next we setup an AP at the end of the room. This was so we could have a fixed reference to test several antennas against.
Below is the netstumbler graph of the test.

All up the test showed that the 8 Slot Waveguide was around 17db and offered a pretty stable signal.

Earlier in the week I contacted a company that have a Spectrum Analyser, their technical guy kindly offered to test my waveguide and see what the SWR was and the resonant frequency.
To my dismay the test bombed. The resonant frequency was somewhere around 8Gig with a harmonic sitting around 2.8Gig. And the SWR was a long way off being useable.

Confused I went home and checked all the measurements and to my dismay found that the first mistake I made was cutting one of the slots about 10mm to long. But after thinking about it for a while it occurred to me that this would not increase the frequency range of the antenna that much. After measuring the lower reflector plate I found that it to be 10mm off making the resonant cavity 10mm less than what it should be.
Bingo!! That is where the problem was.

Rather than try and patch up the mistake I opted for making another 8 Slot Waveguide. I checked, checked and double checked the measurements. I also took the opportunity to make a Jig that would allow me to easily replicate the measurements without to many hassles.

Here is a picture of the finished 8 slot waveguide.


Next was to test the antenna.
An AP with 2db antenna was placed at a distance of 21m away and an elevation of 10degrees above the test point (up my driveway) to provide a stable reference point.

The test unit was a modified USB 802.11 Cameo device. I say modified because these units don?t normally have an external antenna adapter. I have inserted an SMA connector into it. I use a 500mm pigtail

With no antenna attached to the pigtail the results were SNR=38 / Signal= -58

Slotted Waveguide was then attached to the pigtail.
The rotation test was done while the waveguide was in a vertical position.
0 Deg as a reference was having the slots of the waveguide pointed at the AP.

ROTATION TEST
0 Deg (W/G pointed directly at the AP) SNR= 76 / Signal= -24 / Noise= -100
60 Deg (W/G rotated right by 60 degree) SNR= 70 / Signal= -24 / Noise= -100
90 Deg (W/G rotated right by 90 degree) SNR= 70 / Signal= -24 / Noise= -100
120 Deg (W/G rotated right by 120 degree) SNR= 76 / Signal= -24 / Noise= -100
140 Deg (W/G rotated right by 140 degree) SNR= 34 / Signal= -24 / Noise= -100
240 Deg (W/G rotated right by 240 degree) SNR= 72 / Signal= -24 / Noise= -100
300 Deg (W/G rotated right by 300 degree) SNR= 76 / Signal= -24 / Noise= -100

TILT TEST
40 Deg Up tilt (tilted 40 Deg Up at 0 Deg) SNR= 50 / Signal= -24 / Noise= -100
50 Deg Down tilt (tilted 50 Deg Up at 0 Deg) SNR= 50 / Signal= -24 / Noise= -100

Satisfied with these results I decided to take it back to the guys with the Spectrum Analyser. Just for reference the unit they used is worth over $45'000.00.
The resonant frequency of this unit is 2.295Ghz being the optimal frequency at SWR 1:1.2.
being such a small amount off is not going to make to much difference but if one wanted to tweak the resonant frequency this would be done by inserting 2 screws into the chamber. In theory this could be use to get an SWR of 1:1.1 and the frequency bang on.

Certainly the Slotted Waveguide works and is simple. There are many additional bonuses like the effects of lightning on antennas. With a Waveguide it will only ever hit earth unlike with an omni whereas it will hit the antenna.

Thanks to .Rob Clark and Trevor Marshall for the information they published on their websites.

Great stuff shadow, now you just need to setup a sweat shop to build wave guides

Yeah well done. That looks really good.

Beautiful machining too. Looks well made. Good to have someone making the quality stuff here. The figures are certainly impressive.
The best think is they don't look like aerials. Make more. Sell some to us!
I wonder how long it will be till we can all scrape together enough to have a Spectrum Analyser for the whole group to use to aid antennae testing/ design.
That is what a user group does. It shares equipment for the advance of the whole group. Too bad they are so costly.

I need one of these!

ShadowX - Can you supply antennas or help me (and us all)
with making one.

Since you have made a jig, maybe it would be easier to knock off
a few at a time. I'm sure there are plenty of people here
keen on one.

Cheers,
Bigted

Bigted

One step ahead of you I have 3 units that should be available next week.

I have not officially set the pricing but I am still working through the costing and trying to scrape through a minimum labour charge.

I may also release a 4 Slot Waveguide for those who are in close proximity but what the stability offered by the waveguide.

All the waveguides have their slots covered by a plastic type material so as to keep the bugs and weather out. Also they come with an N type female socket on the front.
There is a 90 to 100mm space at the bottom for mounting however the 16slot needs some additional bracing and I am designing a better bracket that will be sold separately.

Hi Guys,
Well the tests are not official. I still want to run a few more so as to ensure that the figures are/were not being interfered with by any strange signal reflections and i am in the process of building a better test rig so as to be able to map the signal a little better.

VSWR: For those who are not sure what this is in breif it is the amount of the signal you are transmitting that gets reflected back into your trnasmit equipment. This would be one of the most important figures as if your VSWR is way off you can cause permanent damage to your transmitter.
Here is the plot showing the VSWR for the waveguide at different frequencies.

M1 is the figure we are interested in as it shows the VSWR for the 2.4Ghz range.


Next up was to test is against a known referance antenna. For this i used a 15db Hills Grid. A referance point was established and a reading was taken from the same point for both antennas. The test was performed over a range of around 200m (length of 2 rugby fields) using the ZyXel AP with a 2dbi antenna. A referance signal was attained from the Hills Grid. The wageguide was then connected and placed in exactly the same point we actually got 2 readings from the waveguide and for this test i used the worsed of the two so as to be on the convservative side. The plate with the radians on it has a very slight curve that caused the base of the waveguide not to sit perfectly on it, the better signal was afheived with a 1mm uptilt.
However by deducting the Hills Grid reading from the Waveguide reading we had a difference of 4.4db So 15db + 4.4 = 19.4db !
Just for kicks i did the same with the same with the higher reading. 20.9db

OK now for a rotation test. the referance is 0° pointing directly at the referance signal from the AP.
Here is the chart.

and for the horizontal plane.


The big lobes at the base (rear of the antenna) weare a little strange. Im not sure as to where or why they are there but at some stage will do some more tests to ensure they were not from some unusual reflections.

I am a complete noob to all this, is this in reference to internal antenae only or are they able provide a hotspot connectivity as well?

Maximus,

These antennas are Outdoor antennas. I guess you could use them indoors if you were keen but their design is for bilding Hotspots and connecting to hotspots.

Cool be keen to see some pricing, I am seriously looking at moving to Papakura to anyplace that is fibre fed

Shadow,

All i can say is..... BRAVO!


We take our collective hat off to you!

Wookie

Hello.
I just joined the forum today so please excuse me if this subject has already been covered.

What I have is a quick method of building a very precise slotted waveguide antenna out of easily obtainable cheap materials and a free design program hlssa61e from Australia http://apache.airnet.com.au/~fastinfo/wire...ss/slotted.html

The basic materials are Styrofoam (the high density exterior house insulation type - typically blue) and heavy duty aluminum foil (the kind you wrap turkey's in for baking/BBQ).

With a veneer blade fitted to my circular saw, I cut the Styrofoam into precise 30mm by 90 mm strips and cut to length for the number of slots you want.

The aluminum foil is attached to the Styrofoam with spray-on contact cement (choose a kind that does not dissolve Styrofoam ) This gluing step takes practice and patience to get a smooth durable bonding. Make sure that the ends also have a smooth bonding.

Now mark the aluminum with a thin marker pen and cut the slots with an Exacto-knife.

For the signal receiver I use 1/4 inch copper tubing soldered to LMR-400 cable center wire. The length of the copper tubing is about 26-27 mm (just less than the 30 mm guide thickness). A clean hole for the tubing is cut into the waveguide with a flush ended drill bit (you don't want to puncture the aluminum foil on the back side). The tubing is inserted into the drilled hole and the LMR-400 braided shielding is flattened out into a disk; soldered to give it rigidity; and attached to the aluminum foil around the drilled hole. The best method of attaching that I have found is Crazy-Glue.

The antenna is ready for testing but is too fragile for normal use. To protect the antenna from abuse and weather, I use a 4 inch plastic drain pipe (the 30 mm by 90 mm cross section fits nicely); a plastic electrical box connector to fasten the cable; and end caps for the plastic pipe.

I have made and sold uni- and omni-directional slotted waveguide antennae from 4 and 4+4 slot to 20 and 20+20 slot using this construction method ( at the moment, I have a 4+4 omni listed on eBay http://cgi.ebay.ca/ws/eBayISAPI.dll?ViewIt...gory=45003&rd=1)

Larry

Larry,

Great stuff. Did you gix the connector using a peice of copper?

I ran some calcs using that application but found them to be slightly off. im not sure if i was missing something. you can get an adhesive aluminum tape that may be a little better suited. also there is a stainless foil avalible that could also be interesting to play with.

I spoted on one of the news groups that someone had been using a similar design to feed a parabaloid.

Andrew

ShadowX, I'm not sure if this what you were asking but ....... :roll:

The connector (a.k.a. receiver) is a 26 mm long piece of copper tubing attached to the end of a LMR-400 cable. The end of the cable is peeled back by 30 mm so that the braiding/grounding is out of the way. A stub of about 5 mm of insulation around the copper core is left on and the rest is stripped away. Now the copper tubing is fitted over the copper wire and fitted about 2 away from the braiding. The copper core is soldered to the inside of the copper tubing. This is the receiver. The braiding is spread out into a flat disk about 2 mm from the inner end of the copper tubing. I apply some solder to this braiding to stiffen it. This is the receiver grounding.

If you go to the eBay web site given at the end of my posting, you'll find a table taken directly from the program. Those are the parameter's I use. I found they give very good results. To the best of my ability and equipment, the observed gains and beamwidth are as predicted. What parameters were you using?

The reason I use aluminum foil is that it is CHEAP ($2 for 25') and aluminum is a better conductor than stainless.

The crazy glue works very well for me and it is cheap (and I get it free ) and I have experienced no problems with it so I haven't needed to search for other solutions.

The application with the unidirectional slotted waveguide with the parabola was posted by me to the http://www.everettwireless.net forum. My logon there is larrynavery
Larry