What are your thoughts on for a small tower using a 4600 4x4 AP connected to the Mimosa M5-360 4x4 Beamforming antenna?
The antenna is 4 overlapping 180 degree sectors and has an operating range up to 6400mhz. For Mimosa it works and gives a 3db beamforming gain.
If this is a viable option what setting should the AP Antenna options be put on? Sector or Omni/Split Sector? The description says “This mode enables operation of split sector antennas with non-overlapping patterns”, but the Mimosa Omni does have overlapping patterns 4x180deg so would Sector be the better setting? Will this Omni give beamforming gains?
I don’t know if this is helpful info or not, but for what it’s worth, we have a 3000 with a 5Ghz Mimosa 4x4 Omni, with the 3000 AP set to Omni/SplitSector mode, and it’s in a small sub-division with only 9 clients. About 1/2 of them report being groupable and benefit from MU-MIMO. It’ll do close to 300 Mbit aggregate when testing to multiple ‘groupable’ clients, and the AP is reporting better than 1/2 of that download bandwidth as ‘green’ MU-MIMO (145Mbit ‘green’ MUMIMO of 235Mbit ‘blue’ download in this test – so about 60% mu-mimo data)
@ninedd What is the proper cabling order to use when connecting the omni to the AP? The 4x4 Onni has 1(V), 2(H), 3(V), and 4(H). On the 4600 AP it just has Ch0, Ch1, Ch2, Ch3
I actually don’t remember. I’d have to climb to see, and it’s blizzarding here. In pretty sure there was a thread on here somewhere that described the correct way for the 3000.
Cool. Let me know the results if you get your installed before me. I am going to test it at a smaller tower site where, based off of the N5-360 gain, I plan to convert any existing AirmaxAC clients that are 1.5 miles or less from the tower.
This site just does not justify yet putting up four 4600L or 4600 sectors.
If it works well, I should have around 21 clients on it.
By vertical spread you are talking about the Elevation Beamwidth correct? If so that is about the same Elevation Beamwidth as the Ubiquiti 13db onmis I am used to.
Am I remembering correctly that the horizontal and vertical elements are rotated 90 degrees from eachother? I tried one on an ePMP 3000 when they first came out, but I could never get anything clients to show up as groupable, so I assumed that the ePMP didn’t know how to deal with that antenna layout and ended up replacing it with a KPP 4x4 omni when those came out, but maybe I just didn’t have the cables connected correctly (or they updated the ePMPs to work better since then).
So, I don’t remember what cable config we used - seems to me it was 0, 2, 1, 4 or maybe it was 1 & 4, 2 & 3 - or some such thing.. but I don’t remember and the Snow and Ice tell me that I’m not going to go climb to find out right now.
BUT, I figured I’d post this to show that we do get MuMimo groupings on about 1/2 of the clients. Sometimes there are 6 groupable, sometimes 5. And sometimes the groups have 4-3 members, sometimes it’s 3-2 members… (plus there is an old goofy Force 180 in the mix, so that one is obviously ungroupable anyway, so it’s really 5 or 6 out of 8 possible) but I don’t stress about this. It was put up as an experiment to see - it’s been there for a couple years and it’s really doing great as a MicroPop for these 9 clients.
So if the Mimosa clients are X (slant) polarized and the ePMP are + (cross) polarized does that mean I would see a 2-3db reduction in gain connecting a + polarized SM to a X polarized antenna?
AFAIK, the Cambium 2.4ghz Sectors were xPol and the F200s weren’t… Cambium at the time said that was beneficial because the noise was reduced more than the signals were… so a SNR benefit.
Yeah, most radios don’t care if you mix X-pol and H/V antennas… they figure it out, and there’s generally no reduction in signal levels. The 2.4Ghz ePMPs are the only other ones I can think of that were intentionally designed that way, but I tested it with lots of different radios after Cambium explained to me how it works, and pretty much all of them handled it fine.
Yes - and on the 2.4Ghz gear, I bought a number of 1000SM’s with external dishes and mounted the feedhorns 45 degrees off to be x-Pol and tested back and forth.. no signal or performance difference at all that I could see. (just remove the 4 screws holding the feedhorn and easily mount 45 cross polarity)
Not sure if you found and answer yet, but just looking at the port layout between the N5-360 and the ePMP 4X4 Sector this is what I’m seeing.
N5-360 - 1(V), 2(H), 3(V), 4(H)
ePMP sector - 0(V), 1(H), 2(V), 3(H)
I tried it out at one site. Did not really work that well IMHO. SM signal was around 5-6db lower than expected. UL signal to AP though was within 1db of expected. 60db DL and 65db UL. Performance on link tests were good 600+DL and 115 UL on 80mhz at 1 mile.
The issue though is the UL rates of this single SM are unstable and cnMaestro reports a retransmission rate of 10-15% on the upload. Not sure why, but I am thinking that the AP cannot really figure out the antenna and is causing some type of problems that affect UL as this is a very rural area so pretty much zero 6ghz noise.
It was worth a shot, but I am going to redo the setup and probably do four 90deg horns with 4600L APs instead. Was trying to save a few bucks on a low client count per AP when doing sectors, but I do not feel comfortable with this omni.
I’ll just repurpose the 4600 AP at another site where I am doing 4x4 sectors and put the omni on the shelf.
What port config did you do for cables if I may ask, and did you have AP set for Omni/Split sector? Going to be trying this next week as we just got cables in and I just finished programming 4600.
