Imagine a typical 2dbi or so antenna for 2.4Ghz. My understanding is that these antennas are a monopole and rely on the ground plane to form a ‘virtual’ dipole.
So what happens if I have a point to mount an antenna, which happens to be 5m high, but for various reasons, it’s easier to mount the antenna so it is pointing downward, ie hanging from underneath a surface. (Its actually a cherry picker like machine).
How does this affect the antenna’s gain/pattern?
I’ve seen this done a few times in commercial gear. For instance you’ll see GPS gear with the gps antenna facing upwards, and then hanging dowards is the radio/dgps antenna.
dipole antennas have nulls off each end. Like a dimpled sphere. Those dimples change as you reorient the antenna. Of course, an antenna’s pattern given by its design.
5/8 wave verticals have about that gain and don’t need a ground plane. Most WiFi w-routers use this.
you can go to the manufacturers’ websites and look at the antenna patterns for their various antennas.
At 2.4GHz, there are many popular antennas. the patch antennas offer high gain, like 8-10dBi, but are directional (Horiz) with a beamwidth of say 30 to 60 degrees. There are 12dBi gain omnis that are 4 ft long (big! for 2.4GHz). But their vertical beamwidth is just a few degrees - like a flattened doughnut. Good for wider area coverage.
For the teeny 802.15.4 (ZigBee optional) modules, ceramic “chip” antennas are popular. These are like 1/4 inch square and work better than PCB etched antennas.
OK, steve got the theory. Application now: There might be a slight bias “up” or “down” on the antenna, but in general you’ll have a fairly symmetrical radiation pattern in the vertical slice for such a low gain omni antenna. Your structure will probably interfere with it from communicating in the upward direction efficiently, so hopefully most of your other nodes are either below or about even with your antenna mount.
yep. If you have a wireless router with the the little vertical antennas, on the first floor, immediately above on the second floor the signal might be a little weaker than off-center. But here we’re penetrating a floor and that has a far more profound affect. In some commercial buildings, the floors are poured concrete with a steel “pan”. very lossy. Not so in most residences. Lath and plaster is more lossy than drywall or plywood.
I use a NIST database of losses by material type (dry wood, wet wood, bricks, hollow construction blocks, etc) in some large-building designs.
The general rule is that if the rod antenna is vertical then the mating end(s) so also need to be vertical. In a line of sight link, opposite orientations (horiz vs. vert) costs about 6dB of loss. In a heavily non-line of sight path, this doesn’t apply.
Patch antennas have vertical polarization too - should be marked.
This isn’t anywhere near a building. And the pattern I need is omnidirectional with most gain towards the horizon.
Still not sure though whether I lose gain from flipping the antenna vertically. As I understood it, the non-balanced antennas require a ground plane. Or does that not apply for 5/8 wave?
Has anyone published any figures for the db loss you get when a transmitter is confined inside a metal box? It isn’t infinite. But is it more like 50db, or 150db ?
Yes, the NIST (US National Institutue of Standards) published some years ago (I have a copy) of the attenuation of many common building materials versus frequency. I’ve heard of less rigorous work done by some universities. NIST spend a lot of $ documenting materials like wet wood, dry wood, cinder block, brick, drywall, poured cement with various degrees of rebar, and so on. They didn’t do a metal box because a steel box, ignoring leaks due to doors, is so large of an attenuator that we have a “who cares” situation.
RF test “Screen Rooms” are used to test equipment in a room devoid of external interference and/or to prevent the emissions of equipment under test from escaping. Also to isolate government-classified RF activities.
You can Google for companies that make screen rooms and read their attenuation specs. Some are indeed copper screen wire, while the really overkill ones are solid sheets of ferrous metal with RF gaskets on the doors. Like a microwave oven has.
An intermodal shipping container is like 80+ dB of attenuation per wall at all but very low frequencies. Leaks from poor door seals would be significant only at very high frequencies, like 20GHz and up.
Speaking of shipping containers. That reminds me. The brochure for a certain GPS/GPRS module starts with a large display picture of a pile of shipping containers.
Makes you wonder if there’s an external antenna made to survive the kind of abuse shipping containers get
Quite seriously though, one of my problems is how to get GPS to work on a mining truck (the 300 ton variety). Given that it has to face the sky, can’t be behind metal, and is likely to get the occasional 1 ton rock land on it… :?
Russel:
Given that it has to face the sky, can’t be behind metal, and is likely to get the occasional 1 ton rock land on it… :?
I know it’s OT for the thread, but have you thought of having the antenna “disposible”? Given that the most expensive one on SFE is less than $20, it wouldn’t be much to replace it if something happened. If uptime is an issue, maybe you could have two antennas on opposite ends of the truck feeding two GPS’s? Maybe you could have the connector on the bottom of something to protect it from falling rocks?
A project I’m on is looking at using ceramic chip PCB mount antennas for 802.15.4 radios that go on the outside of intermodal containers. These are the same antennas as you find in WiFi and Bluetooth dongles.
1/2" thick polycarbonate will easily withstand the having rock dumped on top of it. Anyway, the area of the rock is small. If you antenna is 4x4" there is no way that you are going to have that much weight on it. Some clever sheet metal around it with also help.
