As we started a brand new WISP, we wanted to make sure we were using the best equipment we could find. We spent 4 months field testing APs and CPEs (3 months longer than we thought we would). This turned into an antenna test as well.
I won't post the full details here but you can read the test and results at:
Antenna selection is critical. With your installations where customers are quite far apart you need high gain sectors and high gain subscriber units. The RF-Element Simper are low gain as they have an antenna pattern which is different from normal sector antennas. They are a good selection for short range where you need a wide elevation pattern.
RF-Elements sectors are good (the cambium sectors are custom built RF-Elements). The UBNT-AC Sectors are good too. They are difficult as they have a very sharp pattern. They are -6db at the edges. You might realign them with customers online. Dont trust the written degrees.
With lower signals 11ac does not shine compared to 11n. NBE-19 is a short range device. You've should take dishes.
UBNT does beta testing at customer sites. Their 11ac implementation is still beta. Esp. the integration with 11n devices is very unstable.
With lower signals 11ac does not shine compared to 11n. NBE-19 is a short range device. You've should take dishes.
UBNT does beta testing at customer sites. Their 11ac implementation is still beta. Esp. the integration with 11n devices is very unstable.
We did use dishes. We tested the Powerbeam 300 ISO dish right next to the Nanobeam. The Nanobeam should not have any problems at 1.5 miles - that is well within its limits. However, to be fair, we tested a high gain dish from both vendors right next to a lower gain small form factor CPE. The Force 180 performed much better than the nanobeam and the Force 110 performed better than the powerbeam. I think those were pretty fair comparisons.
AC is not in beta. While UBNT has beta firmware (which we tired many versions of), it is a release product with second generation hardware coming out now. You are right, the 11n / 11ac integration is not there. We went with a 100% 11ac environment for testing purposes. Since we are a new WISP, we were able to start from scratch and build a network without having to support any legacy products. In addition, our packages start at 25 Mbps and top out at 100 Mbps to the customer. We put only a few customers on the high speed APs.
@Cambium_Sri wrote: Great story AU. Kudos for doing your due diligence and thorough testing before deciding on the right product.
Just curious, are you planning to do any frequency reuse using GPS sync?
Thanks, Sriram
We did play with GPS sync but did not have it turned on for the testing. We also do not plan on using it in production either. The reasons are:
1) Our sectors are all pointing in about the same direction. 100 degrees covers the town.
2) We only have a few APs on the tower.
3) We offer packages with very high speeds and the "flexible" throughput allowed us to achieve the high speeds. Testing with the 50/50 or 75/25 did not achieve the results we wanted.
Great post, happy to see someone do tests like this -- I've been thinking about it but am not sure about how it could be funded.
A couple of questions just to clarify. The UBNT gear you were using was all rather recent, and (I follow FCC approvals pretty closely) the PBE was only approved for very low power once you upgrade it to DFS. Not just low power onDFS, but low power above 5.725 too. Its 5.15 approval is very low. The smaller NBE is approved for much higher EIRP. So did you record and compare transmitter power levels between those and the ePMP? The Force 180 doesn't even seem to have FCC approvals yet -- I assume you were testing a beta under Cambium's experimental authorization. So we don't know what power it will be approved for. (It is a crap shoot, almost literally, under current testing procedures. A change to procedures is being discussed with the FCC.) The older ePMPs are approved for decent power levels (old tests).
Also, did you test on any cluttered paths -- trees in the way? The 5 GHz band isn't really good there, but sometimes you have enough signal to blast through a few.
BTW it's "line of sight", not "site", but it's a common error. ;-)
A couple of questions just to clarify. The UBNT gear you were using was all rather recent, and (I follow FCC approvals pretty closely) the PBE was only approved for very low power once you upgrade it to DFS. Not just low power onDFS, but low power above 5.725 too. Its 5.15 approval is very low. The smaller NBE is approved for much higher EIRP. So did you record and compare transmitter power levels between those and the ePMP? The Force 180 doesn't even seem to have FCC approvals yet -- I assume you were testing a beta under Cambium's experimental authorization. So we don't know what power it will be approved for. (It is a crap shoot, almost literally, under current testing procedures. A change to procedures is being discussed with the FCC.) The older ePMPs are approved for decent power levels (old tests).
Also, did you test on any cluttered paths -- trees in the way? The 5 GHz band isn't really good there, but sometimes you have enough signal to blast through a few.
The UBNT gear (and Cambium gear at this point) was tested in 5.8 I have one clean 20 Mhz channel in 5.8 so all the tests were done up there with the exception of some later tests with the TwistPort antenna on the ePMP. They were tested with both DFS and non-DFS "upgrades" since I have keys for those radios. At the 5.8, I do not recall any changes in EIRP on UBNT gear and it was not on a low power. I want to say 27??
I have 3 Force 180's. The one used in this testing was retail purchased from DoubleRadius. No beta agreement. They are back out of stock but they were shipping.
Everything was pure line of "sight" :-) No trees and no fresnel path problems.
Nice tests I was wondering how whill RF elements Twistport designs antenas preforme, it seams for micro locations where is a small amount of space available whill be amazing!
And amazing how big profiles you are selling up 100Mbit for residential users is wow :)!
i deployed ac gears .... 1 year later ac gear has been replaced by EPMP for PTP, reason? ubnt ac 40 mhz 21 km .... 35 mbps down...10 mbps up ... EPMP 65 down 50 up ... and fully stable not like ac which was everyday disconnecting and sometimes we need to go to tower to reboot it phisically
for PTMP ...we are now migrating all our clients from ubnt (nanobeams and powerbeams) to EPMP force 110, and results makes us crazy, we are getting 80/85 down/up mbps instead of 3 ...1.5 ....2....0.400 ...that ubnt uses to give.
now we are also opening new areas with epmp ....sincerally epmp is allowing us to grow as ISP
i deployed ac gears .... 1 year later ac gear has been replaced by EPMP for PTP, reason? ubnt ac 40 mhz 21 km .... 35 mbps down...10 mbps up ... EPMP 65 down 50 up ... and fully stable not like ac which was everyday disconnecting and sometimes we need to go to tower to reboot it phisically
for PTMP ...we are now migrating all our clients from ubnt (nanobeams and powerbeams) to EPMP force 110, and results makes us crazy, we are getting 80/85 down/up mbps instead of 3 ...1.5 ....2....0.400 ...that ubnt uses to give.
now we are also opening new areas with epmp ....sincerally epmp is allowing us to grow as ISP
regards and happy new year!!
Hi GINEBRA254, It is fantastic to hear that ePMP is helping grow your business! Happy New Year!
This is intersting and funny ... can't help myself :)
All equiment used from Ubiquiti was the wrong selection(NBE-19 goes max up to 1.5 - 2 miles) for your location(no mention of PowerBeam 500) and when you say you compared apple to apples what you are saying that you compared a Cambium N fully mature product vs new Ubiquiti AC products mostly still with beta software which is still a product in the making .... u say apple to apples ... just one apple looks orange for some reason ...
Other key elements on your COMPLETE field test ... After a long conversation with Cambium, we decided to deploy some ePMP TEST GEAR. - did you also talked with Ubiquiti people just for the apples to apples sake.
BTW: your speed offering makes no business sense at all ... offering 50 and 100 Mbps Download speeds plans and with those prices you are transforming this into a full time hobby .... not a product that will make you money.
Of couse people here will agree with u since it is a Cambium forum ... but some business owners here are also smilling but will stay quiet - go post this on dslreportsdotcom for more impartial comments.
This is intersting and funny ... can't help myself :)
All equiment used from Ubiquiti was the wrong selection(NBE-19 goes max up to 1.5 - 2 miles) for your location(no mention of PowerBeam 500) and when you say you compared apple to apples what you are saying that you compared a Cambium N fully mature product vs new Ubiquiti AC products mostly still with beta software which is still a product in the making .... u say apple to apples ... just one apple looks orange for some reason ...
You do realize the Ubiquiti AC gear has been being sold for nearly 2 years and it was announced at nearly the exact same time as epmp gear. Whose fault is it after 2 years that ubiquiti's software is still beta and why are they selling and marketing it if it is beta. Are you defending ubiquiti or criticizing them?
And he compared force 180's(16dbi) and epmp 110's(25dbi) to nanobeam 19's(19dbi) and power beam 300's(22dbi). I'd say they're comparable.
See when you say apples to apples means N gear vs N gear Ok .... AC is different and works different, I'm not going to go into details ... when Cambium comes out with AC gear do a Complete Field test then ...