Triple Polarity Worth It for BackHaul?

I just received my first Netmetal5 pair and realized I accidentally ordered the 3 chain version. Is anyone doing this yet? Does the bandwidth go up with that 3rd chain? Triple Polarity dishes are expensive. I could try 2 dishes with one being a slant polarized but this seems messy. If I decide to use them as a dual polarity setup will it hurt anything to just turn off the 3rd chain?

you didnt write about your backhaul: distance, desired throughput, environment, etc…

I used a 2x2 antenna and a panel.

Well, I’ve got about a hundred of em floating around the network. The first set will probably go up on a short hop. 2.5 miles. We’re using 3’ dishes now with some Motorola PTP backhauls that are capped at 200 megs 1 way. It’s a backup link for some airfibers but the airfibers shutdown in heavy rain. We generally run 450 megs max 1 way through that link and I’m hoping to get close to that.

I can’t even get a true 200Mb out of the new AC products linked at 866/866 data rates, I don’t think that adding a 3rd chain to bring the rates up to 1.3Gb/1.3Gb will help much. Mikrotik still has some work to do on getting proper throughput. How is it I can’t get above 170-200mb on 866Mb data rates? And the rates don’t even crash when I push the speedtest??? Locked at 866 solid, I have 3 PtP links now with Netmetl and Netbox5 and all have the same issues. 5.19 > no improvement. 40Mhz 802.11n was even better in some cases…

That is your problem not an general problem.
Here is my btest between SXT Ac limited by the cpu. In real trafic I’ve seen about 220mbps TCP in 40Mhz.
sxtAC 20Mhz.jpg

That’s what I thought too, until I put my THIRD link, and had the SAME issues. The best throughput is maybe 300 but that is 802.11, not with Nstreme or NV2. Funny I put back up rb800 boards and get 180Mb.

I think were going to try these at dual polarity. Will it harm the boards if I keep that 3rd chain unselected?

It will if you have it selected with no antenna connected. Keep it unchecked and you’ll be fine.