First of all I want to thank all the community. I have configured a link that reaches 25km and crosses a 7km sea channel. on both sides I'm using mikrotik 433 with R52hn cards and 30db dishes. One dish is a ARC DA5830SD1 5GHz 30dBi and the second is a Gibertini OP90.

though the theoretical signal should be at -50db I cant make an alignment that reaches less that -73db. I attribute this due to the relative low altitude that the signal travels over the sea. Furthermore the link sometimes degrades bellow -88db signal strength.

In order to do what I can to stabilize the link I am using the following configuration on both ends:

band=5ghz-a/n basic-rates-a/g=6Mbps,9Mbps,12Mbps,18Mbps,24Mbps disabled=no frequency=\
5240 frequency-mode=superchannel ht-ampdu-priorities=0,1,2,3,4,5,6,7 ht-supported-mcs=\
mcs-0,mcs-1,mcs-2,mcs-3,mcs-4,mcs-5,mcs-6,mcs-7,mcs-8,mcs-9,mcs-10,mcs-11,mcs-12,mcs-13,mcs-14,mcs-15 l2mtu=2290 mode=station-bridge \
name=R52Hn rate-set=configured scan-list=5240 ssid=BB.xhaos-tom supported-rates-a/g=6Mbps,9Mbps,12Mbps,18Mbps,24Mbps wireless-protocol=\
nstreme wmm-support=enabled

Has anyone tried to make a long distance link over water to give me a few pointers? For example:
Should I use vertical or horizontal polarization? (I have tried both but don't notice any difference)
Should I reduce the channel bandwidth to stabilize the link? (In fact I have tried 5, 10, 20 and 40MHz bandwidth but I don't see any changes in signal strength or in CCQ)


WIND (program equivalent to RadioMobile) shows clean LoS with 1st and 2nd order fresnel zones clean.

O really would appreciate any pointers as I am new to this kind of thing.
Thank you all for your time.

Hello, i shuld work with two the same antenna devices and have a preference for the dish number one.

On this distance use 20 Mhz bandwidth, i use most of all vertical polarisation and this is the best polarisation over sea.

What protocol use u for the link? 802.11 ... or Nv2?

I am using nstream. 802is out of the question due to long range and ack timeout.
NV2 introduces lag in the link and I am not so sure that I will gain something from NV2 since its a p2p link and not a p2mp

...........................................

WIND (program equivalent to RadioMobile) shows clean LoS with 1st and 2nd order fresnel zones clean.

...................................................................

I agree it's best to use two of the same antennas for PTP link, also can you post a picture of the actual fresnel zone clearance, my thoughts are if there is signal reflections off the sea - how is it best to reduce - increase or decrease height of one side or re-locate the link

You do not say how high the antennas are. At 25km you have to take into account the curveature of the earth.

You are lucky you even have a connection at a low antenna even with 30db antenna.

At 25 km, you need your antenna at the minuam 400 feet high to allow for the center of the signal path (fresnzel zone) to have a clear line of sight.

You are to low to the water, ground and getting bounce.

You do not say how high the antennas are. At 25km you have to take into account the curveature of the earth.

You are lucky you even have a connection at a low antenna even with 30db antenna.

At 25 km, you need your antenna at the minuam 400 feet high to allow for the center of the signal path (fresnzel zone) to have a clear line of sight.

You are to low to the water, ground and getting bounce.

I think your calculations are incorrect for fresnel zone

I have used many different tools to calculate the fresnel lobes including ubiquiti's link calculator. And all show that the zones are clear. One side of the link is at an altitude of 250meters above sea level and the other side is around 70m from sea.

furthermore, trying for better alignment I managed to reach -67db on both sides, but only temporary. with time I notice that the signal fluctuates from -68 to -90.

I think that the problem is sea humidity.

You can try use circulair polarisation antennas.

How are the antennas now polarized?

I have used many different tools to calculate the fresnel lobes including ubiquiti's link calculator. And all show that the zones are clear. One side of the link is at an altitude of 250meters above sea level and the other side is around 70m from sea.

furthermore, trying for better alignment I managed to reach -67db on both sides, but only temporary. with time I notice that the signal fluctuates from -68 to -90.

I think that the problem is sea humidity.

I would in Dude setup log client signal strength
http://wiki.mikrotik.com/wiki/Graphing_ ... l_Strength
and monitor if the signal variation is due to sea tides, if it's signal reflection or not

Hi,

We have 38km link working over the sea.We are using 33dbi antennas with ligo.Signal levels around -40 ,-55,-65 .But sea affecting this link capacity and signal level's because of salty water vapolarization.This should made your problem..

Hello, i shuld work with two the same antenna devices and have a preference for the dish number one.

On this distance use 20 Mhz bandwidth, i use most of all vertical polarisation and this is the best polarisation over sea.

What protocol use u for the link? 802.11 ... or Nv2?

Are you sure about use the Vertical polarization is better over the sea?
I think the vertical signals after collision other the sea with the 180 degree phase difference are back to the other antenna.