dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
17937
share rss forum feed


shoe1

join:2007-09-28
Colfax, CA

1 edit

10-15 mile Wifi possible cheap? legal?

So I live out in the sticks and it's either Wild blue or dial-up. 10 minutes away though, there is the land of High speed DSL, and I know people in this area. Would it be possible for under $300 total to connect to a connection that could hold 6 Mbps of Bandwidth? The sight of sight is pretty bad, tons of trees but maybe I could install on the trees? The connection can be extremely directional. Should I use the 900 MHz spectrum? since lower frequency go through things better. I was thinking getting DD-WRT routers upping the power on it like crazy, maybe install fans in them, and high gain antennas with surrounding dishes. And I want to stay 100% legal too. What are the FCC regulations in California? Any advice would be very much appreciated.


LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1

You're dreaming.

Without easy perfect line of sight, you will never do it for anywhere near $300. Even with perfect LOS, it'll run you more to go that distance.
900MHz equipment is expensive and cannot go that distance through trees.
Running DD-WRT on routers nullifies their FCC approval.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey


reply to shoe1

said by shoe1:

And I want to stay 100% legal too.
Does that include not asking your friends to violate their TOS with their DSL ISP?

Assuming you are buying DSL and just need it extended, your first cost will be to build masts that get you LOS with a fresnel zone above the tree tops.
»en.wikipedia.org/wiki/Fresnel_zone
said by shoe1:

The connection can be extremely directional.
Two of these would get you started, although I would be amazed if you really only needed 2 feet of cable at each end if you need a mast to get above trees.
»cgi.ebay.com/Home-Networking-Sol···ViewItem
said by shoe1:

What are the FCC regulations in California?
Same as the rest of the US, so far as I know they have not seceded or been thrown out of the US.

LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1

Using that yagi would violate FCC rules as it is not tested and approved with that router.

Any length of coax would negate the gain of the antenna.

Rain on the antenna elements would reduce gain considerably as well. At 2.4 GHz, the elements need to be protected by a radome. Even at 900MHz, yagis are affected by rain.

For the distance needed, outdoor radios should be used and there are many avialable with approved high gain antennas but they will break the $300 budget. Just getting up above the trees in itself, will break the budget.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey


Ledebuhr1

join:2007-12-27
Port Huron, MI

Should antennas then be placed inside a buildings attic as to protect them from rain? is rain hard on the antenna or just the signal?



CylonRed
Premium,MVM
join:2000-07-06
Bloom County
reply to InGreenwood

quote:
Does that include not asking your friends to violate their TOS with their DSL ISP?
Can be legal as far as the ISP goes - several do allow folks to share the connections even if they are not part of the household/house. Most do not allow it - but a couple do.

LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1
reply to Ledebuhr1

Yagi antennas have finely tuned elements that get detuned when water drops change their resonant length. This is a very different phenomenon than airborne rain affecting the signal.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey


stevech0

join:2006-09-17
San Diego, CA

2 edits

Here's the laws-of-physics / engineering on 10-15 mi with "wifi"...

1) Most ill-advised at 2.4GHz due to high user density = lots of interference, though highly directional antennas such as wire-mesh parabolics (like these: »www.hyperlinktech.com/web/hg2424g.php and »www.hyperlinktech.com/web/58ghz_···2dbi.php) will help a lot. Still, users in the near vicinity of either end will be a problem if those users are active.

2) 802.11b/g is a 20MHz bandwidth (wide) signal. The narrower the bandwidth, the less noise. Less noise = more range, in general. Less bandwidth means lower bit rates and speed, usually, but new technologies like 802.16e with "real" MIMO and adaptive rate modulation and beam-steered antennas can eek out lots more speed for the same bandwidth. Kinda like in automobile engines: horsepower per cu. inch has gone way up due to better designs.


LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1

1 edit

Click for full size
The local school board had a 13 mile link on Cisco 2.4GHz gear for years before switching it over to Tsunami on 5.8GHz. Of course they had tall towers at both ends to clear the fresnel zone.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey

stevech0

join:2006-09-17
San Diego, CA

1 edit

yes, long paths over hills or urban clutter = tall towers to get Fresnel zone clearance as shown in the graphic, above.

Size of zone varies with frequency - higher = easier.
Path loss per Km increases with frequency too!
But antenna size reduces as freq. increases. Bigger = more gain, as a general rule.

If you have enough antenna gain, you can work out a link budget that can tolerate some blockage of the Fresnel zones.


LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1

said by stevech0:

If you have enough antenna gain, you can work out a link budget that can tolerate some blockage of the Fresnel zones.
Yes, in the plot I posted there is not a clear fresnel zone because of the trees. I have the option in RM of showing all the fresnel but I chose only to show .6 which many consider the minimum for long shots. The tree height (only estimated) leaves a possible .9 fresnel.

Higher frequencies have smaller fresnel zones but also have greater free space loss.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey


shoe1

join:2007-09-28
Colfax, CA
reply to shoe1

Wow. Awesome advice guys, thank you.

well this is my geographical situation:
»maps.google.com/maps?f=q&hl=en&g···aveplace
pull the map up a little and you will see Weimar. I can set up access points from Weimar(place with dsl) to Colfax. I'll break $300, I know, I just aim low and see where I get, and I'll probably put Wireless Bridgers on the way to me too. Wiemar is at 2000ft elevation and Colfax is at 2500ft.

I know about the dsl terms of use and I'm working on that as-well. Will the FCC fine you if you use a dd-wrt router with a untested antenna(not tested for that router). I'm not worried by the FCC regulations on hardware as much as FCC regulations with signal distribution.

I'm still unsure if I should use 900mhz or 2.4ghz or 5.8ghz, anymore advice? It needs to be weather protected and signal needs to be just as good in heavy rain(not really really heavy) as in good weather.


stevech0

join:2006-09-17
San Diego, CA

I'd recommend 5.8GHz with the highest gain antennas you can afford - wire grid parabolics are good.

Less costly: same thing, 2.4GHz. Don't do it in an urban area due to interference.

Bridge-to-bridge in both cases.

900MHz is likely to cost much more. The US band is 902-928 MHz and therefore most gear is NOT 20MHz bandwidth as is 802.11 in 2.4 and 4.8GHz where the bands are much larger.

5.3GHz is licensed (easy to get, on line), more costly.

FCC type certification is voided using higher gain antennas. However, the regulations permit higher radiated power as the beamwidth narrows, and there's an FCC formula for this. At 2.4GHz, it's max 4W radiated as I recall, for quite narrow beam. Note that there are some 12dBi and higher antennas that have 7 degree beamwidth ON THE VERTICAL but are omni on the horizontal. These would be used by WISPS, but I say they're illegal. The intent of the FCC regs is to minimize interference- so narrow horizontal beamwidth (as in yagis and dishes), meet this intent.


LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1

Click for full size
I don't know what SteveO is talking about. The FCC does not allow EIRP based on beamwidth. It is based on frequency band and sometimes on whether the link is PtP or PtMP.

Here is a shot between the two towns mentioned. The tower heights are 24 metres. If you want a more accurate shot, give me GPS coords.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey


shoe1

join:2007-09-28
Colfax, CA

1 edit
reply to shoe1

Amazing. Thanks for that graph. Is that a free program I could download? So I do have a pretty good line of sight? I could probably install on a 100ft tree on both ins. But is there risk for lighting? What do I need to ground it? Would it fry my system if it's hit even if grounded?

I don't know the GPS coords but the Points are right off the exits on both ends.

4W radiated power? Wow. Doesn't the DD-WRT firmware only do 271mW? What could do 4-6W?

So what frequencies are guarded by the FCC? I only need the signal to be 6 mbps of bandwidth that doesn't require 20 mhz of bandwidth does it?


stevech0

join:2006-09-17
San Diego, CA

1 edit

That's 4W >>> RADIATED

That means the amount of power launched into the air. So the antenna's gain makes the difference.
Consider:
30mW at ant. jack of wireless access point/router (this is typical for 11g consumer products).
30mW is about +15dBm (dBm is the usual log scale for RF power)
Add, say, 14dB of antenna gain at one end
= 14+15=29dBm radiated (ignoring coax cable attenuations)
+29dBm = 800mW radiated power.

now make the antenna gain be like a wire-mesh parabolic dish:
+15dBm again, from transmitter
23dB antenna gain
= 38dBm radiated
= 6,300mW = 6.3Watts (illegal in the US, for 2.4GHz)

Now take the radiated power for some antenna, in dBm. Subtract the path loss, depending on path length (Km end to end) and terrain/foliage clutter (at 2.4GHz). This will often be 100dB or more. So you have, say, +29dBm radiated - 120dB of path loss, plus more loss if the fresnel zone isn't clear or about -90dBm at the far end. That's a weak signal for WiFi. Now put that -90dBm into a gain antenna at the receiving end, say, another 14dB antenna. So the signal at the antenna output, to the far end's receiver, is -90+14=76dBm which is good.

This is a little "Link Budget 101". It's how things are done, rather than totally trial-and-error. It's done for microwave links, satellites, cell phones, etc. There's much more to it, like fading margins and so on.
===================================

Some outdoor products claim more than 30mW (15dBm) transmitter power. Many are B.S. in 11g/OFDM modes. Going from 30mW to, say, 100mW is expensive, like $100 or more, due to OFDM-compatibility (a.k.a. peak-to-average "Backoff").

DD-WRT's user interface has choices for power above 30 or 30mW but most hardware ignores this, as it is incapable of doing so, or it causes too much distortion of the OFDM (11g) signal, or it violates what the product was FCC certified at, in terms of the "emissions mask" compliance (inteference to an adjacent 20MHz channel, among 1, 6 or 11.



shoe1

join:2007-09-28
Colfax, CA

2 edits
reply to shoe1

So a Linksys WRT54GL(2.4ghz) with 100mW output with a 24dBm yagi antenna would be illegal? When raising the power output on routers with dd-wrt is heat the only issue? I can install extra cooling if need be, but will the router mess up for other reasons? When does too much power start affecting things?(preferably in the WRT54GL)

I can't figure your formulas for finding how much mW translated to dBm, could you explain that a little bit more?

So i need at least 29dBm on both ends? What will the bandwidth on that connection be, due to signal loss over travel?

I'm thinking the 5Ghz spectrum. First off because it wouldn't mess with anyones personal WiFi(at least reduce it) and it can handle objects in the Fresnel zone better(is that correct?) What are the other advantages and disadvantages of using 5 Ghz?


russotto

join:2000-10-05
West Orange, NJ

said by shoe1:

So a Linksys WRT54GL(2.4ghz) with 100mW output with a 24dBm yagi antenna would be illegal? When raising the power output on routers with dd-wrt is heat the only issue? I can install extra cooling if need be, but will the router mess up for other reasons? When does too much power start affecting things?(preferably in the WRT54GL)

I can't figure your formulas for finding how much mW translated to dBm, could you explain that a little bit more?
1 mW = 0 dBm (this is part of the definition of dBm)
10 mW = 10 dBm
100 mW = 20 dBm
1000 mW = 1 W = 30 dBm
10000 mw = 10 W = 40 dBm

Your antenna is not 24dBm. It is 24dBi, where the "i" means that the reference (0dBi) is a theoretical perfect isotropic radiator (an antenna which radiates equally in all directions)

You can add those figures to get effective isotropic radiated power -- 20 dBm + 24 dBi = 44 dBm EIRP. This is well above the legal limit of 4W (about 36 dBm)

stevech0

join:2006-09-17
San Diego, CA
reply to shoe1

said by shoe1:

So a Linksys WRT54GL(2.4ghz) with 100mW output with a 24dBm yagi antenna would be illegal? When raising the power output on routers with dd-wrt is heat the only issue? I can install extra cooling if need be, but will the router mess up for other reasons? When does too much power start affecting things?(preferably in the WRT54GL)

I can't figure your formulas for finding how much mW translated to dBm, could you explain that a little bit more?

So i need at least 29dBm on both ends? What will the bandwidth on that connection be, due to signal loss over travel?

I'm thinking the 5Ghz spectrum. First off because it wouldn't mess with anyones personal WiFi(at least reduce it) and it can handle objects in the Fresnel zone better(is that correct?) What are the other advantages and disadvantages of using 5 Ghz?
Please re-read my message, above: A WRT54G cannot produce 100mW.

In addition to the crib sheet provided above, Google for "dbm to mw" or some such and you'll find on-line tables and calculators.

Advantages of 5.8 or 5.3GHz: less crowded.
Disadvantages: more costly hardware and a bit more attenuation per Km; not much.

reading and research on your own from here would be beneficial to you.

LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1
reply to russotto

said by russotto:

This is well above the legal limit of 4W (about 36 dBm)
Depends...
In the context of a Linksys, we are talking about 2.4GHz band which for PtP allows heigher EIRP. The 36dBm limit is for PtMP and even then, only for the AP. The clients are allowed the PtP exception.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey

stevech0

join:2006-09-17
San Diego, CA

I should re-check, but as I recall 4W is the US FCC limit for point-to-point with about 10 degree H beamwidth. Point to multipoint was far less. But my memory may be at fault.

And the FCC regs I recall do not distinguish between a client and an access point. From an FCC viewpoint, they're all radios and can interfere.



Killer Max

@rr.com

said by stevech0:

I should re-check...
Don't ! This thread is much more interesting with all the speculation going on.

LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1

said by Killer Max :

said by stevech0:

I should re-check...
Don't ! This thread is much more interesting with all the speculation going on.
LOL
SteveO just loves to challenge everything I say and you want to take his fun away?

Of course I have to go by IC rules despite the fact I can throw a stone across the river onto US soil but I am quite confident in my knowledge of FCC rules even though they don't apply to me. You'd think that Americans would know their own rules better than Canadians.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey

stevech0

join:2006-09-17
San Diego, CA

2 edits

FCC 15.247(b)(3)
for 2.4GHz, the transmitter power must reduce by 1dB for every 3dB of antenna gain in excess of 6dBi.

Also, freq. hopping system have more stringent power limits. At 2.4GHz, this is mostly Bluetooth these days.

Interesting that this is NOT required in the 5.8GHz band.

------------
What I do know about the US FCC is that they have taken in many BILLIONS of DOLLARS in the spectrum auctions. That's where they sell the ether that nature and God gave us.


LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1

Ain't that the truth.

Here is the IC prayer book that I follow.
»strategis.ic.gc.ca/epic/site/smt···210e.pdf

For a quick lookup of FCC stuff however, I often refer to the Telex FAQ.
»www.telexwireless.com/FAQ/Defaul···rketID=1
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey


LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1
reply to stevech0

Click for full size
BTW, here is a view through one of my IP cameras aimed South. The lights on the other side of the river is US soil. That thing in the foreground is a hydraulic truck dumper. It picks up giant B trains and stands them on end, dumping out the wood chips. The construction equipment is a second dumper being built.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey

stevech0

join:2006-09-17
San Diego, CA

I was on a project in NYC where we ran wire mesh parabolic dishes, 802.11a 5.8GHz, and had 10 or so near line of sight links, some of which were 5-10 miles. Two used high gain omni's that provided service on the across the river between Manhattan and NJ, and ashore in NYC for a half mile or so to rooftops.



shoe1

join:2007-09-28
Colfax, CA

1 edit
reply to shoe1

ok so now i need to get up 100ft. SMA is the connection for antennas right? Well I was thinking getting this:
»valley-ent.com/catalog/female-ma···244.html
and then another cox male to female adapter
and getting this:
»www.monoprice.com/products/produ···fication
A regular SMA cable is expensive as heck but coaxial is dirt cheap.(would this plan work?)
The coaxial cable says 3 GHz sweep tested, What does that mean? I will be running a 5 GHz wireless signal through it.


LLigetfa

join:2006-05-15
Fort Frances, ON
kudos:1

You've lost me there... what radios? What antennas?

Don't use long coax. Either use outdoor radios with integrated antennas or external antennas with short coax.
--
Strange as it seems, no amount of learning can cure stupidity, and formal education positively fortifies it. -- Stephen Vizinczey



shoe1

join:2007-09-28
Colfax, CA

1 edit
reply to shoe1

ok. well how do I do a 100 ft install then for cheap then. A outdoor radio is too expensive for me. With longer cable the signal degrades, but is that the same even with coax? Doesn't cable internet have no distance limits? As long as the cable is strung out there.

If long coax is bad could I buy a coax repeater?