This is a point to multi point bridge environment. The root has an Omni and the others have directional ants. There are 3 non-root locations. 2 of the three are the following distances: 400 ft and 1.2Miles, they are connecting, but with lower signals than expected. The third site is 1.9 Miles and occasionally drops in and out because of the low signal. This one is the main problem. It has a Dish, Signal strength is around 20% and Quality IN is 40% and Out is 0-20%. The dish has been changed the cable checked and the radio changed. A carrier test shows no significant freqs interferring. The utils on the bridges do not indicate that I should have any problems. Any suggestions???
How long is the coax cable to the antenna at each remote site?
Did you confirm that you have not only line of sight, but a clear Fresnel zone for each path?
Yes. The line-of-site is no question and there are no obsticles in the fresnel zone. The main site has a 100ft LMR400 cable (low loss), two of the non-root sites have under 75ft and the 1.9 mile site has a 200ft cable. The calculations for the long cable with the low loss show that I should be well within the range for connectivity.
To determine the theoretical distance you can operate this setup, we require the following information (for both ends of the link):
-- Bridge types,
-- Antenna type, Antenna gain,
-- Cable loss/100ft, Cable length
When I make the following assumptions:
2xBR340s,11mbps,12dBi Omni to 21dBi Parabolic, 100ft at 6.7 to 200ft at 6.7
I get the following (theoretical) results:
max distance w/10dB margin = 1.54 Miles
Required Antenna Height above obsticles (in feet) = 22ft
Fresnel Zone = 17ft
When I change the datarate to 5.5Mbps, I get a max distance of 2.45miles, so that might be a simple test that may stabalize the link.
Ultimatly, you may need to try and eliminate some of that 200ft by bringing the bridge closer to the antenna. The BR350 is good for this as it has inline power support (you can't have a mixed br340/br350 link though - you need the same bridge type at each end of the link)
Your assumptions are all correct except for the cabling. The lengths are o.k. but they are using low loss 3dBi/100ft for the entire run except for a 2ft piece of the standard Cisco cable between the low loss and the bridge device. It looks something like this (if you can visualize it). The  is the bridge and the (- is the dish. How much loss per connector? ----===========----(- . One of the non-root sites has the standard Cisco cable and the signal is just O.K. but acceptable. I swapped the cable at the main site with the Standard Cisco cable just to rule out the Low Loss cable and the signal was actually worse, so it is making an improvement. Thanks for your help!
well.... Im a bit out of my league here, but Ill put my money on those 2ft interconnects being the problem. can you terminate the low-loss cable directly on the bridge/antenna?
How about my suggestion of lowering the data rate (1,2,5.5Mbps) to see if that make the link stable at least?
I have suspected the 2ft interconnects myself, but the cable guy checked them out o.k. and when I by-passed that cable with a one piece Cisco cable the signal was actually worse, so it disproved my theory. Lowering the data rate does increase the signals, but I suppose the the issue is that this should work fine at 11 Mbps for this short distance.
hmmm, I'm just guessing here, but here is what we think we know:
200ft of low loss cable (3dBi) should work
200ft of cisco cable (6.7dBi?) will not work
Here is what we dont know:
2ft of high loss cable connected (loss) to 200ft of low loss cable conneccted (loss) to 2ft of high loss cable.
If those 2ft cables are the cisco interconnect cables, they have a high loss and are for connecting a cisco bridge to an antenna over a very short distance where the effect of the high loss is minimized. I don't know what additional loss the connections between the cables introduce, but it could well be significant.
Not sure where you got your numbers.
100 foot of RG-8/9913 has a loss of 8 to 9 dB. (Almost 90% of signal strentgh)
100 foot of RG-58 is more like 19dB (99% loss).
So 200' of anything without an amp is a problem.
I got the numbs from the cable guy. Obviously I lack where the cable starts out to the ant. I have a pic of the cable markings and can e-mail it to you if you send me your email [ email@example.com ]. The markings on it say "84147 Andrew LDF5-50A Heliax Coaxial".
I don't know if you seen my post about the amp.,,,,Can I legally use one?
excuse my ignorance, but this has not come up before. How do you check the polarity of the antennas? Cisco says the ants are vertically polarized. How would you go about verifying polarization? I really appreciate your help and new ideas. Thanks
If two vertically polarized antennas are pointed at each other, but one is rotated 90 (or even 20) degrees they won't see each other. (This is sometimes done on purpose.)
Just make sure they are all oriented the same. If they are all the same type, it's as easy as checking that the pigtails or connectors all stick out the bottom/top or left/right.
If they are not all the same, check the docs so you know which way is 'up' on each.
I feel like I have exhusted the possibilities on the cable, although the connectors could be causing the issue. Does anyone know if I can use an RF amplifier to boost the signal to a stable level? I am getting mixed opinions on this. The company that sells the amps. say I can legally boost it as long as I do not exceed 36dBi , but a rep from Cisco said it cannot be done legally at all. The company that sells the amps hosted a cisco wireless event one day and advertised this product at the event. This has me confused, can I or can't I? Thanks again for the input.
There are multiple interpretations of the FCCs proclamations.
If you use an amplifier, at the very least test the final output power (rather than relying on just the dBm math to tell you).
CISCO cannot sell amplifiers because they would be responsible for such testing, so it's easier for them to say it can't be done.
BTW - Have you considered moving the bridge closer to the Antenna?
I have found that in the past the signal strength is not such a good indicator.I would use a unicast test but set it to 999 packets and see what response you get.
having installed many installations without any problems the only other thing I would check, is for local Microwave , line of sight, and length of cable between Omni/Dish etc:.
What antenna are you using for the Root bridge? Is that location line of site? Can you see the other antenna with the naked eye or binoculars? What is the environment between the root location and the client location?
12dBi Omni at the Root. I have RF Line of Site to all the locations. Other antennas can be seen with the naked eye. Root is in a small city environment, there are alot of other ants in the area but the carrier test is not showing very much noise or channel use in our bandwidth, the remote site is in a less populated area and there is residential in between. I changed the channel to a different one and the signal improved slightly but throughout the entire band of channels there is not much difference.
Frank, are you STILL having problems at that site?
I'm in the office for a couple of days if you want to give me a call.
If you have a extra patch antenna in the 9dBi range, you may want to create a point to point configuration to that location to see what your readings are. If they are better, you can assume that the omni-directional antenna at the root may be the problem.
Also, I have found that sometimes an omni directional antenna may need to be rotated slightly to get a better signal to the client. If you look at a graph of the antenna's spectrum, there can be low spots in certain directions.
Just some ideas.