Could be this possible??

Hi all, 

ive just seen that ubnt is selling the AF-MPx4 .... that will you save space in tower connectting up to 8 rockets in only 1 dish ...

i dont want to go back to ubnt gear...i prefer cambium, but the thing is that i need to make a bonding on my main link to gain some extra throughput and not wasting space on a second antenna on each tower. 

my biggest problem is that in one of the towers...i have no more space available

what im thinking is ....


i really need to have a higher throughput and also redundancy on the link. 

i thinks that if this works....could be the best option for high demmand links.


i have tried rocket AC and its like playing russian roullete with a glock, ( link is up for 1 day....then 22 is resting)

i will appreciatte your ideas . 


1 Like

Why not just skip all the UBNT headaches and use PTP820 in a licensed band, like 11GHz... if you need even more bandwidth, get the dual core, want even more, dual-radio XPIC setup... you can bond two high capacity radios to one dish. For that distance at 11GHz, you could get away with a 2' dish, but you might want to go with a 3' dish if you want the best performance. You can use LinkPlanner to figure all this stuff out as well, along with providing you a bill of materials (BOM).

1 Like

keep in mind that these NxN combiners are passive "resistive combiners", or "wilkinson combiners". 

in order to multiplex two inputs into one output, keeping both inputs isolated from each other (so that one radio's transmitter does not fry  the other radio's reciever), the combiner sacrifices 3db (4,1 in the real world) of signal. 

For a 2-radio combiner  the losses ammount to 8,2 db. 

to keep in compliance with EIRP rules, you can add 3db of cable loss at the radios, but that will still rob you 3db at the recieve side. 

bottomline,in the best case scenario you loose 3db. there is only a point doing that until you reach 18dbm of TX power at the radios. that's the maximum TX power they can run at @MCS15. 

in the end, altough it could be tried, there is no guarantee the GPS sync is tight enough to work this way , And you would need to use 30dbi dishes to get comfortable signal levels. 


22km = -135db free space loss @6ghz

-135+18+23+23=-71  >> bare minimum to run atMCS15, using cambium Force dishes

-135+18+28+28=-61 >> comfortably runs MCS15 with "30"dbi Dishes.

-135+18+28+28-4.1-4.1 = -69,2 >> perfect alignment 30dbi dishes with combiners. 

The force is definitelly not with you on this one.

What size channels are you currently running? in TDD, TDD-ptp, or ePTP modes? maybe some tweaking could buy you enough troughput.


better PTP radio will server you much better, look at the ptp650 radio...   you'll get 3x the bandwidth and none of the headaches.  

also many less parts to break.


Thanks Chris, 

licensed  link is not an option by the moment, that link may be we have to remove it in a few months.

650s are 5ghz unlicenced, they have some amazing RF technology in them to deal with noise and are very fast.

They can move dynamically to dodge noise, the master and slave can use different tx/rx channels as well.
And more of cambiums magic sauce.