Back about 40 years ago I saw a 6 Ghz microwave shot installed over a maybe 70 mile path, with both ends on mountain tops. At one end they had installed (at *huge* expense on top of an 8500 foot high mountain), a 50 foot free standing tower.
Because of the long path, they experimented to see where the dish got the best signal, starting at the top of the tower. It ended up located 12 feet off the ground!
If you have possible path obstructions, do try moving the antenna up and down a bit. It isn't necessarily just a case of getting as high as you can...
One other point... 200 feet of LMR600 coax has a loss of nearly
9 dB at 2400Ghz. It would probably be far less expensive *and* a significant technical improvement to put the radio in a weather proof housing and mount it on the tower. Even LDF5-50A heliax would be
4 dB loss, which might acceptable but the cost isn't worth 5 dB.
4 miles should be achievable with decent antennas. My guess(tm) is that you've skimped on the antennas. Could I trouble you to kindly disclose what hardware you currently have to work with? Let's play with the numbers. My guess(tm) is that you probalby have some kind of dedicated point to point bridge radio such as a DLink DWL-900AP+. I'll also guess(tm) that you've run a bit too much coax cable. Instead of trying to demonstrate why my guess(tm) of your existing system doesn't work, I'll see what it takes to make things work.
At 5.5Mbits/sec, the receiver sensitivity is about -85dBm. Tx power is about +15dBm. My guess is a few feet of coax and connectors will eat about -4dB. A big dish at 24dBi gain is probably required at both ends. I'm looking for a minimum of 20dB fade margin.
Running the numbers at:
Distance = 4 miles Tx power = 15dBm Tx ant gain = 24dBi Rx and gain = 24dBi Tx cable loss = 4dB (including connectors) RX cable loss = 4dB RX sens = -85dBm (for 5.5mbits/sec) Fade margin = ??
That yields a fade margin of 23.8dB which should be workable. So, if you take your existing unspecified 802.11b wireless bridges, buy a pair of 24dBi dish antennas for about $60/ea, repackage your unspecified wireless bridge in an outdoor box with a minumum of coax cable to the antenna, and install some kind of power over ethernet system, it will work.
It's difficult to calculate the effects of the tree that is in the Fresnel zone. I would guess about 6 dB additional loss which may be a problem. You can compensate by slowing down the connection to
2Mbits/sec. However, that's a bad idea. Instead I suggest you purchase a pair of 802.11g wireless bridges and run them at the slowest OFDM data rate of 6Mbit/sec. The will give a receiver sensitivity of -88dBm, which will buy you 3dB additional fade margin and additional tolerance to multipath.
You really do not want to know the price of a 4 mile FSO (free space optical) data link. Try:
hold onto your wallet. Plaintree is the lowest cost of the various FSO vendors.
900 mhz radios would probably work but they will set you back a grand. I maintain and install 802.11b cisco 350's for my employer. What kind of radios and antennas are you running now? Sometimes a bit more gain can help with tree interference. It's not the "right" way but when you are trying to save bucks and only need about a meg of bandwidth, it will work. I just finished installing a 40 foot tower at my residence and will be shooting 2.5 miles to a 400 foot microwave tower that has a ds3 passing through via nec digital microwave bridged together by a cisco router. I will be coming out of the router using a cisco 340's shooting to my house via a
24 db gain antenna mounted at 200 feet using lmr 600 coax. I will also have the same antenna at my house mounted on 40 foot of rohn 25g. I know that I have a fresnel zone issue about a mile from my house but if I can get over a meg of bandwith I will be happy. I will be turning off my dsl connection and my dial tone as I will use vonage for phone service. If it becomes a serious issue for me I will stack enough tower to resolve the tree issue but I will have to guy the tower which I really don't want to deal with. I have imstalled links varying from 100 yards to 20 miles and trees have givin me fits in the past but usually there is a serious rsl issue also. But the quality of radios can make a big difference when you are dealing with a "borderline" issue.
I'm looking for alternatives to 802.11b. I have a 4 mile stretch that is giving me fits. I have clear line of site except that a rather large tree sqeezing in on the fresnel zone about dead center. Does anyone have any recomendations on Laser/Optical/Etc. devices? I can't relocate either antenna more than about 10 feet side-to-side. I feel like a midnight chain-saw visit is in order.
Oh, and BTW... it has to be inexpensive. ( < $750 ).
Gordon Montgomery Living Scriptures, Inc email@example.com (anti spam - replace lsi with livingscriptures) (801) 627-2000
Well, don't laugh. I'm using SMC2682W bridges on each end. Both ends use 24dBi grid antennas. One end is already mounted at the antenna, maybe
16 inches of coax. The other end has 50 feet of LMR-400. I do have another link running 9 miles using the same setup for equipment, except both radios are mounted at the antenna with about 20 inches of coax going to 24 dBi grids. That ran for years until about the last 8 months I had to add an amplifier on one end. I just didn't want to go through the hassle of adding an amp to the 4 mile setup. But maybe I will have to try it.
Oh, and my outdoor box? It is an inverted plastic garbage can, open at the bottom. I was in a hurry to install the old link, and just grabbed the closest water-tight thing I found just to get it running quickly. I always meant to get a _real_ box later, but I just never got around to it. It has been over
3 years now of 100+ degree summers, -5 degree winters and storms strong enough that I had to re-aim the antenna, but the radio has been running fine. Someday I'll get that box though.....
Thanks for your help and information.
Gordon Montgomery Living Scriptures, Inc firstname.lastname@example.org (anti spam - replace lsi with livingscriptures) (801) 627-2000
Transmit cable type : Times Microwave LMR-600 Transmit cable length : 200.000 feet (60.960 meters) Total transmit cable loss : 8.815 dB (0.044 dB/foot 0.145 dB/meter) Total transmit connector loss : 1.000 dB through 4 connectors Total transmit cable loss : 9.815 dB Transmit miscellaneous losses : 1.000 dB
Transmit antenna peak gain : 24 dBi RF input power to the antenna : 5.185 dBm (3.300 milliwatts) Allowed input power to antenna : 24.000 dBm (251.189 milliwatts) Per FCC Part 15 rules Transmit antenna height : 200.000 feet (60.960 meters) Distance to the radio horizon : 19.406 miles (31.231 kilometers)
Receive cable type : Times Microwave LMR-400 Receive cable length : 50.000 feet (15.240 meters) Total receive cable loss : 3.370 dB (0.067 dB/foot 0.221 dB/meter) Total receive connector loss : 1.000 dB through 4 connectors Total receive cable loss : 4.370 dB Receive miscellaneous losses : 1.000 dB
Receive antenna peak gain : 24 dBi Receive antenna height : 40.000 feet (12.192 meters) Distance to the radio horizon : 8.679 miles (13.967 kilometers)
Received power level : -64.626 dBm (131.270 µV) Receiver's threshold : -82.000 dBm (17.762 µV) Thermal fade margin : 17.374 dB This current link setup should be possible. Probability of outage : 1 % Total free space path loss : 112.441 dBi over a path length of 2.500 miles (4.023 kilometers) Peak transmitted EIRP : 28.185 dBm (658.348 mW)
Midpoint 0.6 Fresnel zone boundary : 27.998 feet (8.534 meters) Maximum space wave communications distance : 28.084 miles (45.198 kilometers)
a.. Radio horizon data is for 4/3 Earth, k = 1.33. b.. All cables, antennas, connectors, and adapters have a 50 ohm impedance. c.. Receive and threshold voltages are for a 50 ohm load impedance and don't take into account the receiver's bandwidth or temperature. d.. Horizontal polarization will generally provide less multipath in urban areas and may provide lower path loss in non line-of-sight situations. e.. Attenuation from trees is approximately 0.35 dB/meter at 2.4 GHz. At lower frequencies, the attenuation is somewhat lower for horizontal polarization than for vertical, but the difference disappears above about 1 GHz. f.. Attenuation of a 2.4 GHz signal in heavy rain (150 mm/hr) is 0.02 dB/km. g.. Assumes 0.25 dB loss for each coax connector or adapter. h.. Use as much antenna gain as possible, and get your antenna as high as possible. i.. Maximum space wave communications distance is the longest distance possible using the choosen antenna heights. It does not mean a link that long is possible. j.. Try to obtain a thermal fade margin of 20 dB, or larger, for a more reliable link. If the fade margin is negative, that link is impossible. Calculated on Mon Aug 1 06:15:27 2005 GMT
17 db is fine for 2.5 miles so this path will work but like I said in an earlier post, the trees could be an issue.
| I use ydi 6ghz 45mb (full duplex) tower mounted radios on a link that we | have and every damn time that tower gets hit with lightning it smokes the | radio. I am presently getting ready to move the radios inside the building | and run 300 foot of waveguide, not cheap, but I am tired of calling a tower | crew out to replace the radio. I told the boss when we installed the link | that they would be a headache but it was a money issue. The many times that | we have sent a tower crew up and replaced the radio it would have been | cheaper to run waveguide in the first place. The grounding system is fine | and the megger measured less than 2 ohms. | Talked to too many in the business and know from previous experience. Tower | mounted radios are not reliable in an area with a lot of electrical storms.
You must be doing something wrong. I've been using TR up for almost 40 years all along the Gulf Coast and off shore with few problems. BTW there is much more in a ground system than how low the dc resistance might be.
Sigh. I remember paying $450 each for a pair of those bridges. I still have a few SMC2652W boxes on the shelf from those days. What a piece of junk.
So much for my astute guesswork. You did that part right.
OK, lets rerun the numbers:
Running the numbers at:
Distance = 4 miles Tx power = 12dBm Tx ant gain = 24dBi Rx and gain = 24dBi Tx cable loss = 6dB (including connectors) RX cable loss = 3dB RX sens = -85dBm (for 5.5mbits/sec) Whirrr, whirr, grind, kerchunk... Fade margin = 19.8dB Well, that's not too horrible. However, if the tree eats another 6dB, it will probably be rather flaky. Methinks you're on the bitter edge. Picking up another 3dB by shortening the 50ft run will be a big help.
Well, that works out to a 16dB fade margin, which isn't to good. I'm surprised it works.
Did you add the amplifier to overpower co-channel interference, or because you got a bunch of water in the coax cable and were trying to compensate for the additional loss? Water in the coax is fatal at
2.4Ghz. It doesn't take much.
Amplifiers are evil. Detailed rant on request.
Sigh. One of the local wireless sector antenna farms is on a residential rooftop. Someone installed a fake plastic chimney to hide the mess. It looks like it originally arrived with a plastic Santa Claus climbing out. I also have an antenna hidden inside a plastic owl. I've also seen some very creative cell sites with camouflages antennas. However, I have my limits. No way is one of my installs gonna be hidden by an inverted trash can. The aesthetics committee would have me lynched if I did that.
Not with LMR-400 which is foam dielectric and where the water ends up soaked into the foam. I can usually measure some loss from end to end. However, you need bench test equipment to do that accurately. My favorite alternative method is to simply replace the coax. If the replacement coax works better, then I know there's something wrong with the original coax.
Dealing with the pigtail on the 24dBi dish's is more difficult. On some dishes, the coax is an integral part of the feed assembly. I have spares so I can also check these by replacment, but that's not always an option. The pigtails are RG-8/u which really sucks the water up the braid by capillary action. However, the dielectric is much more dense and water incursion has less of an effect. One the PacWireless barbeque grill dishes, the feed assembly has an "N" connector on the side and is water tight. Much better idea.
The old coax isn't always a lost cause. I bake the coax, which drives the water out by evaporation. Takes about 7-14 days depending on temperature.
Incidentally, for waterproofing, I use 1" wide PTFE (teflon) pipe wrap. I embalm the connectors in the usual spiral fashion. That provides the waterproofing. On top of the PTFE, I add a layer of 3M Scotch 66 tape to hold the PTFE in plce. The PTFE prevents capillary action from sucking water in through the knurrelling on the connector. IT also cold flows into a solid jacket. When removed, the connetors are as shinney as new. You should do this to every exposed connector on the rooftop.
| >Incidentally, for waterproofing, I use 1" wide PTFE (teflon) pipe | >wrap. I embalm the connectors in the usual spiral fashion. That | >provides the waterproofing. On top of the PTFE, I add a layer of 3M | >Scotch 66 tape to hold the PTFE in plce. The PTFE prevents capillary | >action from sucking water in through the knurrelling on the connector. | >IT also cold flows into a solid jacket. When removed, the connetors | >are as shinney as new. You should do this to every exposed connector | >on the rooftop. | >
| Excellent suggestion. I used to use PTFE tape all the time when I | was building fire sprinkler systems in buildings. I'll start doing that. | | Might also look into shrink tubing with a shot of RTV/silicone compound then heat shrink the entire connection. to remove score the heat shrink tubing then heat and it will peel back.
Nope. The coax ends have to be exposed so that there's some place for the evaporated water to go. Just heating the coax won't do much. I use two part connectors (Andros) that can be disassembled which works well for getting the water out of the coax. However, the one part (RF Industries) crimp takes much longer to drain. If I'm desperate, I punch a hole through the outer jacket and shield, and use a basketball inflation needle to pressurize the cable. It barely works with LMR400, where the foam is bonded to the outer shield, but is better than waiting. For air dielectric coax (9913, Heliax) I can pressurize the coax and flow dry the cable fairly easily.
I have one rooftop installation that always seems to leak during the winter rains. This is the first year in about 5 years where I didn't have to drop the 30ft pole and dry out the connectors. The basic improvement was to use 1" PTFE tape instead of the more common 1/2" PTFE with a 50% overlap. Anyway, when the rains stop, it previously took about a week for it to dry out. The signal level differences between wet and dry is about 10dB.
I think I invented the method although it's fairly obvious to anyone that's done any plumbing. I June, I removed the embalming wrap from some connectors I've had on my roof for about 5 years. No sign of any corrosion or water incursion. The connectors were just as shiny as if they were new.
Incidentally, some of the manufacturers will recommend shoving silicon grease into the connector to keep the water out. That resulted instead in intermittent connections. Not a great idea. I do put a little grease into the threads so that the connector will be easy to disassemble, but not inside the connector.
On Tue, 2 Aug 2005 23:30:24 -0400, "NotMe" wrote: |
The RTV will permanently glue itself to the connector. That's fine if you want everything to be permanent. The acetic acid in the RTV will also attack silver plating.
3M makes cold shrink tubing with a sticky interior. They're used by power utilities to insulate and waterproof cable splices. I've used them for coax connectors but they have some nasty limitations. The sticky goo is well... sticky and gooey. No fun to remove. The cold shrink is very thick and therefore does not deal with the radical transition in diameter at the end of the coax connectors. The tubing has to be much longer than the connector in order to make a good seal. The thickness also makes is rather difficult to cut and remove. The large air gap at the coax connector to coax cable junction often gets wrinkled if the cable assembly is bent. The resultant gap at the coax end lets in considerable moisture. Also, they're kind expensive.
| >>The old coax isn't always a lost cause. I bake the coax, which drives | >>the water out by evaporation. Takes about 7-14 days depending on | >>temperature. | | >So does 45 days of no rain with temps ranging from 90 to 103 | >count? | | Nope. The coax ends have to be exposed so that there's some place for | the evaporated water to go. Just heating the coax won't do much. I | use two part connectors (Andros) that can be disassembled which works | well for getting the water out of the coax. However, the one part (RF | Industries) crimp takes much longer to drain. If I'm desperate, I | punch a hole through the outer jacket and shield, and use a basketball | inflation needle to pressurize the cable. It barely works with | LMR400, where the foam is bonded to the outer shield, but is better | than waiting. For air dielectric coax (9913, Heliax) I can pressurize | the coax and flow dry the cable fairly easily. | | I have one rooftop installation that always seems to leak during the | winter rains. This is the first year in about 5 years where I didn't | have to drop the 30ft pole and dry out the connectors. The basic | improvement was to use 1" PTFE tape instead of the more common 1/2" | PTFE with a 50% overlap. Anyway, when the rains stop, it previously | took about a week for it to dry out. The signal level differences | between wet and dry is about 10dB. | | >Excellent suggestion. I used to use PTFE tape all the time when I | >was building fire sprinkler systems in buildings. I'll start doing that. | | I think I invented the method although it's fairly obvious to anyone | that's done any plumbing. I June, I removed the embalming wrap from | some connectors I've had on my roof for about 5 years. No sign of any | corrosion or water incursion. The connectors were just as shiny as if | they were new. | | Incidentally, some of the manufacturers will recommend shoving silicon | grease into the connector to keep the water out. That resulted | instead in intermittent connections. Not a great idea. I do put a | little grease into the threads so that the connector will be easy to | disassemble, but not inside the connector.
If you ever have access to lab grade equipment might do loss sweep on connectors with silicon grease vs. no silicon grease.
Ditto on coax (wave guide too for that matter) that has had water inside. The effects of corrosion on the skin effects are bad.
Also if you're desperate use dry nitrogen vs. air to dry the moisture from the line.