I am trying to branch out wirelessly from my house in Karori, Wellington, but being cursed by the hills I am considering maybe building a repeater to stick somewhere high that me (and others) can use to bounce off to get out of the valley.
Does anyone have any experience with building a repeater station?
I have some design considerations like:[list] What would be better out of just an access point (or maybe 2) or a PC-based solution?
[*] Should I use a waveguide antenna and a parabolic dish to beam the signal away or just a single waveguide to allow a distant node to link us out?
[*] What are the chances I could get away with hiding something like this up on a hill somewhere and not get in trouble/have it smashed to bits? Or anyone know what the application protocol is to put something like this somewhere?[list]
Now another consideration is that I'm planning on making it solar-powered. I can get BP solar panels through work (DSE) in various wattages, and coupled with a car battery I'm hoping I can run this thing 24/7. I was thinking maybe a 50-watt-peak (@12v) panel would be sufficient if I was running just an access point (which draw ~1A peak @ 5v from some light research I've done). The other solution is an Epia 500ish system with PCI wlan cards (could use Epia double-PCI risers if more cards are needed) and run it off a DC-DC power supply which I'm more than familiar with from my Epia in my car... I think that'd need quite a bit more power, so I'd be getting a bigger panel....
At this stage, just some advice would be nice. All comments welcome
Speak to Murray Pearson from the University of Waikato.
Their network CRC-Net (Connecting Remote Communities) uses a solar powered node on Pirongia mountain/hill/bump (Hamilton).
I believe 50watts will not be enough, you need to be able to go through a dark winter week without losing power.
I know they used a lot of deep cycle batteries and ower powered them with solar panels but as far as I know, it has been up almost permanently.
for more info, go to www.crc.net.nz
After having a look there myself it looks like they have more than just the one solar site in operation. I spent a day with them all last year but due to weather we never got to go to the solar powered site.
I'm looking at doing a similar thing where I live, I can't offer much advice to specific situations but you can read what I'm planning on doing if it helps make some sense reading over someone elses case study.
A battery more suited like those sealed lead acid panasonics DSE sell are better. They are deep cycle batteries not surface charge batterys like hevy bulky car batteries. You'd need something to provide 2-3 times more charge than draw. IE: draw 2-3Amps, panel 6-9amps, simply incase it's cloudy, plus you will want to recharge the battery during the day for the increasing night haul now that we're out of daylight savings.
As far as finding somewhere to put a repeater, all my sites in commercial radio broadcasting I've used are leased on someone elses mast somewhere, or on top of a residence I knew at the time or local garage etc... You could always talk to local council authorities enquiring about a peice of land but that's to expensive.
I would suggest for best performance to add a power amplifier to the mix, like the 500mW 2.4GHz boosters you guys at DSE are now selling for $297, at trade that's cheaper than Borg Wi-Fi are currently doing them (staff discount should give you more?).
This is the two cases I've planned I'm looking at adopting for my own sitiation with wi-fi where I live wiht two access point sites:
1>
Main access site (located on personal residence)
x2 access points. *1st - public access point channel 01 boosted to 500mW (17dBm transmit gain/10dBm receive gain) in to a 3-4dBi omni-directional antenna for public (ERIP 1000mW, that would cover 3,000 people here if they used a booster aswell). *2nd - private access point standard 30mW output channel 07 in to 20dBi high-gain parabolic directional narrow beam antenna for link to secondary private access point (remote site) on seperate channel (2000mW ERIP). Mains power available and mounted away from my house as I don't like radiation full stop, on the property boundries up a pole to the limit the local laws allow, here it's 7m's long if attached to the house, 13m long if it's standing alone before resource consent is required.
Secondary access site (located on business building in town for a few beers and self powered). x2 access points *1st - private access point standard 30mW output in to 23dBi high-gain parabolic directional antenna for link to primary access point site as a client on channel 07. *2nd - public access point channel 02 boosted to 500mW (17dBm transmit gain/10dBm receive gain) in to a 3-4dBi omni-directional antenna for public (ERIP 1000mW, that would cover 3,000 people here if they used a booster aswell). 30Amp deep cycle panasonic etc... capable of 6-9Amp charge, seald lead acid battery with voltage regulator and 6-9Amp panel array. Draw design of around 2-3Amps per hour. (effectively providing 2-3amp charge per hour over 2-3 amp draw).
This case shows 2 public access points on their own channel 01 and 02, not causing bandwidth limitations between each other, and linked using channel 07 with a 3rd and 4th access point used for the purpose only to link the two public accessable access points together.
2>
Main access site (located on personal residence)
x1 access point. *public access point channel 01 boosted to 500mW (17dBm transmit gain/10dBm receive gain) in to a 3-4dBi omni-directional antenna for public (ERIP 1000mW, that would cover 3,000 people here if they used a booster aswell). Mains power available and mounted away from my house as I don't like radiation full stop, on the property boundries up a pole to the limit the local laws allow, here it's 7m's long if attached to the house, 13m long if it's standing alone before resource consent is required.
Secondary access site (located on business building in town for a few beers and self powered). 1 access point *public access point channel 01 running in 'repeater mode' boosted to 500mW (17dBm transmit gain/10dBm receive gain) in to a 3-4dBi omni-directional antenna for public (ERIP 1000mW, that would cover 3,000 people here if they used a booster aswell). 30Amp deep cycle panasonic etc... capable of 6-9Amp charge, seald lead acid battery with voltage regulator and 6-9Amp panel array. Draw design of around 2-3Amps per hour. (effectively providing 2-3amp charge per hour over 2-3 amp draw).
This case shows only 2 access points, one the 'access point' the other the repeater point. However, bandwidth limitations are included as both points are serving all nodes on the single wireless channel with no individual backbone between the two. If the repeater is to far away from the access point, the repeater will be redundant as signal can not be acheived between the two. In this case, a power divider splits the 500mW signal from the booster to two outputs for two antenna's. 3dB loss must come in to effect for the divider leaving only 250mW to be split, = 125mW output to each antenna. Using a higher gain omni-directional antenna to serve the public with around 200-500mW ERIP left over as omni's don't get to high in gain before becoming directional, and a higher gain directional antenna such as a parabolic or grid to fire the seconnd output through a more narrow beam between the access point and repeater site. This dividing idea would have to happen at each end to create an antenna array, one for local and one for directional linking.
If those give you any ideas cool, that's kinda the same as what I'm doing.
Those DSE sealed lead acid batteries (7.5A/H) work out to like $10 per Amp hour.
Get deep cycle marine batteries, these are around 70A/H and cost $250-$300 (for the better brands).
Solar panels are rated for direct sunlight, with no cloud - unless you are automatically moving the panels via a tracking device, expect them to be running at only about 75% capacity beyond the 2 hours each side of noon.
There are special formulas for working out how to mount the panels, something to do with pointing them due north (in the southern hemisphere) and pitch them at an angle relative to the sites latitude (North/South of equator).
Ideally, have enough battery capacity to run for an entire 7 days with no charge, and supplying a charge (when charging) of 6 times the average power requirement - Remember that a panel putting out 3 times the requirement is only enough to run a site with 8 hours of direct sunlight EVERY day - 6 times is not an overkill, it is a bare minimum. - This will mean that for any good day of sunlight, the gear can be run, and another day of charge can be supplied.
hehe. It's nice to know you guys don't mind paying retail for these bits
I get the boosters below half-retail and the batteries about 1/4... So that makes it a bit easier on the pocket. The solar panels however will remain expensive.
I took a drive up the hill yesterday to take photos of the kinda coverage I could get from someone's house up that road, if I asked nicely.
If I can get a roofspace, I can probably get power too for a few $ a month... But I'm still considering solar hidden up in the bushes somewhere. There's lots and lots of transmitter towers up there, and all the poles look pretty full. I'm a bit worried some ISM-band transmitters might be up there already, but none of them looked like home-made access points.
If you wanna have a look at the shots, they're here: http://www.fishy.net.nz/gallery/wireless
The first page is from my roof, and the Wrights Hill gallery is up the hill a bit.
I wouldn't go near license holders transmission towers. On them is much better lol, but usually beside them tweaks the interest from RSM on unlicensed gear. Some unlicensed transmission gear can be somewhat unclean, like boosters that create to much hash (wide spectrum noise during amplification).
Funny thing is, if you're a few feet away it can cause lots of interference even with good gear, but being on the same pole makes it much less likely to cause interference. If you ever want to setup two antenna's or more always use the same pole if possible, and same earth for ligtning strickes, as the potential between the two can also cause problems in rf signals being warped in pattern from the site. As far as stacking antenna's on the same pole, it's called vertical isolation. This setup is less likely to cause distortion (intermodulation problems) between each antenna/receiver/transmit system, as they transmit and radiate outwards from their vertical alignment, not directly through each other at full force.
Either that or have the polorization of each antenna opposite to each other. One vertical the other horizontal etc... put them on the same pole about 1 and a half wave lengths appart (so you can fit another antenna and a bit in between).
I don't know if any of this makes sense or is of any use to anyone building a wi-fi site, but it's an excilent rule of thumb especially with flea power gear like wi-fi, where you want as much clean signal as possble. Also 1M height on any size pole from small to large makes a huge difference. Try to get it off any roof as far as possible, the more room to raidate and the less reflective surface for multipath problems (signals arriving at the same place at different times) the better the reception and speed.
It puts out close to 40A and is mounted not far above a house roof in the city. You can build things using the motor from a Gentle Annie washing Machine and install prop's or mount a motor on and old shop sign that spins around, but they do tend to blow apart if not strengthend.
This can be a cheaper, but longer and frustrating. But very much intertaining and a learning curve.
Matt - Yobbo Wireless
PS: He also built something simliar for someone with a water wheel running a Gentle Annie motor 110V over 200M of cable to a transformer to 12V to charge a batt in the middle of the Southern Alps. Worked very well and it was very cheap.
Solar-powered repeater - advice! 
Linear Mode
