Wireless Joint

I have 3 working routers (Mikrotik RBM33g + 2 RBM11G's, all running R00ter) with 3 providers.

At my current location, towers are in various directions, with Verizon to the East, AT&T to the East-SE and T-Mobile to the West. My ultimate goal is to bond all the available LTE connections together (with OpenMtcpRouter, of course,) and I want to make the the whole system as portable as possible, which is why I'm currently using a one-pole arrangement with a 'cluster' of antennas on it. (ideally a single PoE connection up the pole, but might need 3.)

I understand for any given provider/connection, the Mimo antennas need to be either cross-polarized or at least ~1m apart, so I have the 4 LPDA's pointing at AT&T all at different angles (vertical, horizontal, +45', -45') and this works well (seeing ~200mpbs down ~50mpbs up on Quectel EM160)

The Verizon and Tmobile SIM/modem/routers are currently each fed by a single cross polarized MIMO panel antenna, pointing in opposite directions (East & West), but they are installed pretty much in the middle of the cluster of LPDAs.

My question is whether the close proximity of other antennas is a problem for any one of the connections. IE, does having 'competing' antennas/connection so close to each other cause interference between the antennas? There is definitely 'overlap' in the frequency/bands used by the carriers (they all seem to use LTE band 2, for example.)

If each carrier needs to have the antennas further apart from the others to really maximize things, I'll have to reconsider the one-pole approach for multi-carrier aggregation/bonding...

TIA for any experienced feedback or input on this. Paul.

PS: I've also considered an approach using antenna signal splitters, with one set connected to the directional LPDA array, and another connected to a set of Omni's. Then I'd swap connections around as needed in whatever location. but concerned about signal loss due to splitter/etc, so thinking best to keep each carrier/modem with it's own antennas...

maybe i just need one of these! arg! https://www.amazon.com/-/dp/B0851Y4CLV

When getting in as deep as you are I would suggest to start with each provider and the exact bands and frequencies that you will need to acquire. Hopefully you are close to all towers and have line of site? Antennas would then be dependent on the frequencies your after for that antenna port. I think you would be okay with single mast if you can keep at least a few feet between each providers antennas. I've also found it best to stay away from antennas that cover a large spectrum and go with high gain antenna for each modem ports frequencies.

Things get much easier when you have strong signals. In my experiences I have found that T-mobiles primary 4G band (66) operating at 2140mhz is very much affected by one of Verizon's main primary 4G bands (66) at 2150mhz. Also I would pay close attention to your SINR values which IMO are most important when looking for 'speed'. I've tried it all in my area and best I can do is an average of 8 out of 15 on my CQI numbers for T-mobile signal (2140mhz) and thats at 100% signal strength using an inline amp and a couple of splitters/combiners that you mention. Ultimately I get about a third of what I should at 100% signal strength due to noise.

Oh and btw, cellmapper is your friend

mtl26637 wrote: Fri Dec 10, 2021 2:57 pm I think you would be okay with single mast if you can keep at least a few feet between each providers antennas.

thats' the sort of input i was looking for. Thanx.

Someone needs to invent a 'tuneable' antenna. Where you can change/dial-in (either mechanically or electronically) both the frequencies it's best at as well as the directionality, from broad-band omni-directional (just to see what you you can get from anywhere at that location) to directional at a specific frequency (once you've determined which frequencies you need and which direction is best to point in today) One might be able to 'combine' an Omni and one or more Yagi's for that effect, i suppose, but that gets complicated/expensive (and NOISY!) fast when you've got 8 or more antenna inputs to your modems...

That is where cellmapper comes in extremely handy. You can find your tower and provider and see exactly what frequencies they are using at your location in most situations. Once you know which frequencies the provider is using you can then focus on finding which antennas are best suited for your location. The antennas that cover a wide range of frequencies are not going to perform anywhere close to those antennas that are dialed in to a specific bandwidth of frequencies. This mostly comes into play when you are not close to the tower and do not have direct line of site however. If you are looking at T-mo 5G I would def. recommend looking for something that will pick up the 600mhz range as most of the LTE stuff has 700mhz as a lower end. They will still work but not optimal. Once you get your setups running post some of your signal stats and we can take a look and see what may be able to use some tweaking.

packlet wrote: Fri Dec 10, 2021 4:56 pm Someone needs to invent a 'tuneable' antenna. Where you can change/dial-in (either mechanically or electronically) both the frequencies it's best at as well as the directionality, from broad-band omni-directional (just to see what you you can get from anywhere at that location) to directional at a specific frequency (once you've determined which frequencies you need and which direction is best to point in today)

If interested and willing to take things a step further look into something like a 'NanoVNA'. They are fairly cheap and work great for measuring return loss and SWR and things of LTE style antennas to find out exactly what frequencies they actually do work at and how well they work across the entire spectrum.