62km Link - advice required

Hello Folks.

I will be setting a 62km link this morning to link one of my clients remote site. I intend to use the following equipment and parameters.

2 RB435UAH boards (one at each end)
2 DBii F50N Pro mini pci wireless cards
2 32dbi Parabolic Mimo dishes (L-com 4900 - 5800)

I have the following frequencies available - 4980MHz, 5460MHz or 5800MHz (assuming a channel width of 10MHz)
Questions

  1. Assuming the setup like the scrrenshot of radiomobile plan attached - which frequency would be good.?
  2. What level should I tune the tx-power of the radio cards.?
  3. What is the expected throughput of this link?

Due to limitaton on site B (right side) we are not able to go above tower height of 48m.

Advice and guidance will be very much appreciated
test_link.jpg

this link should not be good, since you are blocking the fresnell sone.

  1. try every channel and see, 5460MHz is more usable for most antennas;
  2. leave them default, best solution;
  3. I calculate the signal to around -60 to -65, so for 10Mhz I don’t think the speed will be over 30mbps, maybe less; you could use 20Mhz.
    Inform us about the link when is done.

is this link setup???

Hi Folks

The link is up now. I will post details with pictures tomorrow. Signal is now -69 but we need to do some alignment at one side.

I noticed the ethernet cable running from the parabole on the 73m tower is loosing power: that is when we ping with a packet size of 1500 the routerboard - 435UAH - reboots running at processor speed of 400MHz. its supplied with 24V ubiquity POE. I will run direct power to the top of the tower tomorrow and test again.

will surely await for the pictures :slight_smile:

why are you using Ubiquiti POE on a Routerboard??

Why not?
Check the specs to one from ubnt and then check one from mikrotik.
You will answer yourself.

waiting for the link details for that distance :slight_smile:

Ubiquity POE is 0.5A and Mikrotik recommends 1.5A with its 24V adapters (atleast 1A) hence u might loose power and going for a tower 73mtr is too high to run a ethernet cable.

You have to try different channels and freq. to best test the radio sets.It varies a lot on your landscape and environmental conditions.
I have 2 links 50 km each …1 runs over a river and the other over land(fields). The first link ha signal -70/75 during day and goes to -77/85 during night. The second link has -70/72 signal in day and goes to -63/62 during night. I do not know why this happens. Tower height is 40 mtr at one end and 24 mtr at the other for both links. Using 27dbi grids with 411ah & r5shpn.

Ok as promised find attached pictures. as you can see from the capacity test link is not that very good since the variance in the results is large.

I noticed the is a correlation between the master / slave roles and the download and upload results. That is the side which is AP has least download etc.

The radios are currently running at 22dbm tx-power. at some point the signal was -56.

regards
pilone_02.jpg
capacity_test.JPG
62km_link_slave_stats.jpg

Use nv2 for more stable results.

what tool have you used for capacity test??

yes use NV2 and i think shift the frequency from 4980 MHz to above 5200 MHz.

How much did you invest to erect the mast?

I use the visualware “myconnection server” together with their access series devices. Basically i put the access series at each end of the link and do a host of tests (udp capacity, ping, video, voip etc). One device as slave tests to the master and all results sent to the server for analysis. Very good application.

http://www.myconnectionserver.com/learnmore/quality.html

yes use NV2 and i think shift the frequency from 4980 MHz to above 5200 MHz.

why do you want me to use 5200MHz - any reasons.

I have changed to link to NV2. Will send the results in the evening. My only dislike for NV2 is the latency. But at this point i have no option since the variance in results from nstreme is too much.

ping is lower once traffic goes up with nv2. With nstreme is reverse.