2 sectors back to back on A5c


#1

I found this 180 degrees 5Ghz MIMO sector; http://www.l-com.com/multimedia/datasheets/DS_HG5813DP-180.PDF

I am thinking of using 2 of these back to back on a A5c in interop mode to make a high capacity 360 sector from one A5c where each sector than can tilted downwards to my wish.

Opinions?


#2

This is intriguing if it could effectively replace an A5 with greater range. What are the downsides? Which situations would this kind of A5c setup be superior to just an A5?


#3

The 180º sector is 13dBi where the A5 is said to be 14dBi, per panel.
Effectively this means that for longer range, the CPE will receive or send to the sector on 2 chains. MIMO 2x2 so you get an extra 3dB on top of the 13dBi which makes there for 16dBi.

Using the A5 on short to very short distance will probably be a better idea since here you still will have multi-path signals thus the 14dBi can still result in also 17dBi combined signal.

But on longer to medium longer distances, where remote CPE will have a highly direction 2X2 CPE the comparison probably will fall to the dual 180º sectors since 16dBi will be better then only 14dBi of the A5.

In very specific circumstances where an AP will be located on a steep slope of a hill with an A5 you are very limited. One side the signal burns the ground, the other side it will shoot in the sky.
Only with tow adjustable sectors on each side you can adjust for that.

Off course, 4X4 MIMO will never be achieved this way, but that is also not what I am after. An 250-300% increase in capacity in my network is already a game changer…


#4

Each of the 4 panel antennas in the A5 is 5dbi. Combined they are 14dbi.

See https://apps.fcc.gov/oetcf/eas/reports/ViewExhibitReport.cfm?mode=Exhibits&RequestTimeout=500&calledFromFrame=N&application_id=hVtF%2BB8QTxyCbdz8W9r%2Fmg%3D%3D&fcc_id=2ABZJ-100-00007 document “Test Report Part 90” section 3.4 on page 14.

You’ll also note in the Mimosa design tool, if you configure an A5 unit the reported EIRP is 5db more than the TX power.


#5

Ok, so the two 13dBi 180º sectors on a A5c will give a massive increase in signal compared to the A5.


#6

I don’t think it will work out well. It seems as all beams from the A5c should be as similarly directed and overlapped as possible, hence the 4x4 sectors from KP (and maybe others).
I have not used the A and C series and therefore assumed it worked similar to the B series, but it doesn’t.
Read Jon and Chris comments:


#7

It will work. The situation you’d refer to is when you are looking to get 4x4:2 Mimo with beam forming and Mu-MIMO. In that case it’s indeed designed to use a 4X4 Mimo antenna system. This is the way to reach very high throughput in ranges of 200Mb and beyond.

But like in the old days when a 2x2 AP could be fit with 2 x single chain sector antennas to cover a wider range by one AP than would be possible with only one sector you can now fit a A5c with two sectors. 2 x 2X2 Mimo sector to cover a wider range than you would get with one. Actually examples are given in this blog already. The sit side by side and the horizontal angle depicts the total range. So 2 x 30 degrees would give a high gain 60 degree system. 2 x 60 degrees would give a high gain 120 degrees system and 2 x 180 would give a full 360 degree system with more gain then ever possible from one A5-14.
Off course the achieved high throughput per CPE is much lower. Probably less then half of a full 4x4:2 system with MU-MIMO and beam-forming. But if you are not looking for that, but in stead want a high power AP system that is able to reach further away clients this is a nice working solution.

Presently we work with tdma networks from Mikrotik and like ubnt maximum amount of clients are sort of in the range 30-40 with top total throughput of <100Mbps in crowded spectrum.

We have now one A5-14 working with 60-63 associated clients (all Mikrotik) and the AP reaches tops total aggregate speeds 125Mbps and higher and we can run 3-4 clients simultaneously with 30-40Mbps.
The only downside is that the range of an A5 is insufficient for CPE’s longer then 1km away (which is already a lot). Now we are looking to reach further clients that will benefit from the high capacity Mimosa AP all we need is more gain. So higher gain CPE’s but also higher gain antennas from the AP. One A5c with 2 x 180 degrees sectors has more gain to trow into the game then a A5-14. For instance, with two 18dBi 30 degree sector you can throw a massive EIRP in a 60degree sector with a powerful Mimosa to provide service levels towards clients at 5 - 15 km’s in the world of UBNT, eCambium and Mikrotik you can only dream off…That’s the philosophy behind it.


#8

Yes, I do understand what your point is. I had the same thought before but gave it up because of a comment from some employee at Mimosa concerning using several 2:2 sectors on the A5c. But I can’t really find that comment/thread now and I may be wrong here…

The fact that the C-series is 2:2 hints that it should be possible to have 2 A5c 2:2 sectors with different angle, but the A5c is not just an A5 with connectors and it’s designed for 4:4. Using a client/station of another brand than Mimosa may also perhaps be a positive thing in such a scenario, since it’s working in a less advanced mode and is less likely to try anything stupid :slight_smile: .

If your concept works, you may also be interested in the horn antenna. It’s maximum 90 deg but it has other features that may make it a good choice. The next generation of UBNT AC seem to have copied this design, more or less.
https://www.rfelements.com/products/antennas/symmetrical-horn-carrier-class/symmetrical-horn-carrier-class/


#9

Actually I have seen in the blog guys using two 2x2 mimo antenna with one A5c and I mean to have read somewhere in this forum that some technician of Mimosa said it should make no problem.
My specific use is in fact for an AP that is to be located on the edge of a hill. Like houses on one end are at the same or slightly higher level, whereas the houses on the other end are all lower. Just a plain A5 probably would overshoot these low houses. With two adjustable sectors I could point these towards the clients.
In this specific case I still need omnidirectional coverage hence the two 180 degrees I am looking at.
Off course I can use two A5c’s or even more but that investment wouldn’t be justifiable for only a total of 30-35 clients.

I know the horns, we are going to use these in some other setup. In fact I already use some horns with the Mikrotik simper adapter and they do fine. But they are not that good as many say. The sidelobes are 100% suppressed is the claim. But on a Mikrotik OmniTik (8dBi antennas) that sits 1,5 meters above and half a meter closer to the tower (so behind the horn) I still measure -14dB from a 30dBm Netmetal on this 18dBi Horn…
(Would that make 30+18=48dB EIRP plus the 8dB of the OmniTik makes 56dB link budget and if we measure -14db that means we have a s/n ratio of 70dB? That is still impressive…)


A5-14 in urban/city site
#10

@Rory’s doing it:

I also had the feeling @Kent was also doing this on some sites.

And I’m sure I’ve seen a post reply from a Mimosan that said this will work fine… but I can’t find that threat right now.


#11

I’ve got 2 locations where we replaced the A5-360 with the A5c and dual RF Element antennas and they work great. The A5-360’s were picking up too much noise from behind and out of the area we needed to cover. One A5c has 62 users. The other one was at 60+ users but to test TDMA, we had to pull the Ubiquiti radios off of it. I think it has about 42 users on it now running TDMA. I’ve got about 10 of the RF Element antennas going up next week. We have pictures on the Triad Wireless Facebook page.


#12

As luck would have it, Jamie refers to this exact design in this WispAmerica interview recorded the other day: https://www.youtube.com/watch?v=tyCNwS_P4Ao