Hello,

can I achieve a 70 KM link with Mikrotik 90 cm dish antennas and R52H or XR5 cards ?

Thoroughput on these links should be 12 mbit duplex (real).

I have seen other threads on long distances, but everybody is using 1.2-1.4 meter dishes. I cannot use anything greater than 90 cm, since only towers that exist are not strong enough to support 1.4 meter antennas.

you have to make a longer tower or bigger dish. or an amplifier. but use amplifier very carefully .it can results very poor performance, if you do not use it at the right place and at the right power.

First of all, it is important the quality of your dish (90 cm is enough for that distance), pigtail... I suggest to use only XR5 or SR5! You will get even 25-30Mbps, but is dependent of you configuration of your system (cpu, ram, etc)! Also a good LOS is needed!

You need to think very carefully about fade before making blanket statements about whether this will work or not.

It may work very well at some times of the year, and depending where you are, may work very badly at other times of the year.

If you are planning on putting customers on this link, calculate an uptime forecast first!!!

George

uptime calculation is very imortant. when the season changes at mediterrien zone signals drops up to 12dbi. I have a 60 km link 45mbit download and 45 mbit upload. the anttenas are at mountaitns 800 meter one side 120 mtere other side. with 29 dbi grid both sides and +5tx and +12rx amplifier at the both sides. pigtails are important too. use high quality pigtails.

We're up on the Canadian Shield north of Lake Ontario running a number of 50km+ PtP links.

In the Spring and Fall we can get more than 45 dB of fade, coupled with a 30 dB rise in noise floor due to ducting on links facing Toronto.

The combination of the two effects means that even 3.6M dishes are not enough to hold the link together. Instead, we use spatial diversity with 3.6M dishes (about 30M spacing between antennas) to get enough path variation that we can run at 99.999+% on licensed 6 gig protected diverse wireless.

Very expensive architecture but it works, which is my point about knowing where you are and what your conditions will be like.

I (we) have ased MikrtoTik for diversity for exactly these situations, but so far our requests have fallen on deaf ears.

George

We do not have that much fade here, avg +-5 dB on 30 km link in the spring and fall.

On one side I can put a 1.2 meter dish, but nothing more. On the other side, situation is really critical, the "tower" cannot support anything over 90 cm since it's very weak, and there is no other tower nor other can be built (nearly impossible to acquire building permissions, complicated paperwork).

Here is the config that I have in mind (both sides are the same):

RB532 x 1
XR5 x 1
Pacificic Wireless HDDA5W-32 High performance dish antenna
6ghz MMCX pigtail + 1m LMR400 jumper cable
Mikrotik PoE 48V
Pacific Wireless aluminium enclosure

Anybody used HDDA5W-32 antennas ? How well are they performing in comparison with other 90cm dishes and feed systems ?

we use spatial diversity with 3.6M dishes (about 30M spacing between antennas) to get enough path variation that we can run at 99.999+% on licensed 6 gig protected diverse wireless.

Raises an interesting point: for spatial diversity requiring ~30m antenna spacing, for licensed links the ODU<>IDU interconnect uses <450MHz IF, so it's not a problem to have long cables for such wide diversity spacing;
But with an MT-based radio, If you want a:b diversity switching using a single Atheros card, you end up with long cable runs @ 5GHz to one or both antennas, with the losses that would entail.
Does that imply the best MT solution for such wide diversity would be separate radios/routers each end with the diversity done at networking layer?

Personally I believe MT should implement diversity switching, based on some user-customisable set of variables, for various other applications, including indoor and mobile use, solves some of the fades that occur.
Quite likely it would be easy to implement, but it's probably not on their priority list. Vote for it on the Wiki ...

Regards

Hee Hee! We cheat.

The radios are on the ground and we run waveguide to the antennas. About 4 dB loss for 100M.

Where we are using smaller antennas, say 1 - 1.2M, I would be happy with a shorter distance between antennas for diversity. Using 5/8" hardline we could run 40' up and down the tower with little impact on signal levels.

George