Yes, this. You need sufficient gain and isolation on the CPE end. Don't expect you can do it with 'small' CPE like the old ubnt nanostation M5 size things with fat oval noise-collecting RF patterns. Use of the 40cm and 50cm size parabolic ubnt powerbeam AC (and powerbeam AC ISO with shroud/radome) may be necessary. -56 signal means a happy 256QAM link to a sector.
You have the option to go all the way up to the 62cm size powerbeam AC if necessary for a CPE. On Thu, Jun 2, 2016 at 2:00 PM, Stefan Englhardt <s...@genias.net> wrote: > UBNT AC with RFelement horn antennas or with the 3x30 degree UBNT sector. > > Firmware is stable enough now. Outperforms ePMP for sure. > > Dont believe these „you will not get 256QAM“ prayers from people selling > > 64QAM Equipment. > > > > > > *Von:* Af [mailto:af-boun...@afmug.com] *Im Auftrag von *CBB - Jay Fuller > *Gesendet:* Donnerstag, 2. Juni 2016 22:32 > *An:* af@afmug.com > *Betreff:* Re: [AFMUG] PMP450 vs. ePMP > > > > > > We are looking at an environment of about 3/4ths of a mile, high > concentration, line of sight. > > Hoping to do at least 60 down - prefer 100 + down and at least 4 - prefer > 40-60 up. > > Had considered EPMP. Had also considered ubnt ac. Seriously considering > mimosa. > > Which product will do the best with the most subs? > > > > > > > > ----- Original Message ----- > > *From:* can...@believewireless.net > > *To:* af@afmug.com > > *Sent:* Thursday, June 02, 2016 3:21 PM > > *Subject:* Re: [AFMUG] PMP450 vs. ePMP > > > > ePMP is adding beam forming and should be available in the next few weeks. > 450's beam forming will be a little > > later this summer. > > > > If you get use lots of small antennas, ePMP works well. If you want a > limited number of antennas with high > > subscriber count, 450 is better. > > > > On Thu, Jun 2, 2016 at 4:10 PM, Matt <matt.mailingli...@gmail.com> wrote: > > > This is our most loaded 5Ghz ePMP AP: > > > > Users on this AP have plans from 1.5-10Mbps. > > > > I think it averages 70-80% of downlink frame usage during peak times, > but I don't have a graph available because Cambium won't expose this as a > percentage via SNMP (you have to look at the real-time value in the web UI). > > > > Are these all excellent connections? On average what distance on these > CPE's? > > > >
