As much as I hate omnis, we have a few sites that just don’t justify sectors (unless we used something like the $100 Ubiquiti Lite AP, but that’s another discussion). The thing is, our definition of a small cell is more like 3 miles, with every link having excellent signal and 6x or 8x modulation. We can’t afford to shrink the cells and still have poor performing links because of poor antenna gain.
So my question is whether anyone has converted a site with a connectorized 450 AP and a dual pol omni to the Micropop AP with the integrated omni, and compared the signal level and modulation before and after? Or some other method to compare real world antenna performance?
With the exception of 3.5 GHz where we have used the Alpha Wireless omni which is excellent, we have used the 13 dBi omnis from various sources like L-Com and KPP. I think they are all basically the same Chinese made antenna no matter who you buy it from. And I know that in the antenna world specs are often aspirational more than real world numbers.
If I remember correctly, the Cambium integrated antenna is supposed to be 9 dBi. In theory that would be 4 dB less than our current setup. The antenna pattern suggests much nicer null fill which would help with subscribers very close to the AP. But for the subscribers at 1-3 miles range, will we really lose 4 dB of antenna gain? I don’t think we can afford that.
I also know that some omnis have “cloverleaf” patterns, where the gain will vary 3-5 dB over the 360 degrees of azimuth, and that the two polarizations may not track each other. So if the Cambium antenna really does have 4 dB less gain but keeps that gain very constant with azimuth and polarization, maybe it’s not as bad as it sounds.
The decision would be easier if the gain spec was a little higher, like maybe 11 dBi, with better null fill and more constant performance vs. azimuth than the Chinese antennas.
We have a few kpp omni’s and a couple other brands too. Th kpp omni is by far the more stable when it comes to gain pattern gain shift. Some antenna manufacturers report the -3db pattern gain, others report the -6db pattern gain(kpp) The cheap knockoffs typically use the highest power recorded and dont bother to state that.
Above aside, we have links 5 and 6 miles off a kpp omni that would never link with the cheaper antennas we tried. We typically use a 10mile cell size and monitor each links quality during install to ensure a trouble free connection. Its amazing what a 8ft pole can help with!
To the casual observer, the KPP and L-Com/Hyperlink omnis certainly look identical.
Every few years it seems we have to replace them. Usually we suspect storm damage. We’ve seen the tops popped off and the antenna filled with water or one time full of bees. Right now I’ve got one where every SM has 8-10 dB V-H ratio and I suspect another water or bee event.
We put Polyphaser inline coax surge protectors on both antenna ports so at least we’ve never lost an AP due to lightning. Using the capacitive coupling type that don’t pass DC, not the gas tube type.
I have a fairly good communication line with KPP, never had a problem with them when expressing concerns of poor design issues.
If the cap is coming off, let them know!
I did have one of their early omnis that got full of water due to the vent/drain becoming clogged with what looked like mud (suspect insects), next antenna I ordered had silicone sealant where there was none before.
Bees are another matter, they will eat their way in if they want to! Since its ABS plastic, there are options to make these less attractive to bees, but also being way up in the air and away from trees helps too.
For our go to lightning protection of omni’s, we simply use a bonded aerial about 10ft long and for 5ghz, 24" off to the side of the omni. We havent had a direct strike yet with this method but make sure the rod is bonded not just to the tower but to the runner to the ground. This is more important than bonding the radio and antenna.
We have lost radios to lightning, but none had burn marks on the antenna or radio but the radio was dead. Thinking either a near field strike (where its close enough to dump the emp into the antenna and just fried the receiver) or it was a ground strike surge which is very hard to proove especially if the cisco switch its attached to is fine.
We are looking at Mimosa’s 4port omni for use with epmp3000. Heard some good news from some wisps about them. Also trying to get kpp to make something similar at a better price point.
We have a few KP Omni’s, 2.4 and 900 in the field, and their performance far exceeded our expectations, and we’ve not had any issues with them. Perhaps your installers were a little too brutal with them when rigging the tower, or perhaps KP really improved their durability/design over the years.
Either way, if you can’t justify sectors, I would have no problem recommending them.
Well, KPP and L-Com share the same parent company these days so that’s probably by design. That said, they’ve borrowed designs a few times over the years too.
Well, we did the swap today (already had the hardware, had to wait a couple days for an unlock key because we had a couple subs at 2.5 miles).
All subs were as good or better signal and modulation with the Micropop, both AP->SM and SM->AP.
This might not have been a totally apples-to-apples comparison. The old setup was a 450 AP with an L-Com 13 dBi omni, 16 inch coax jumpers, and coax surge protectors. We had a connectorized 450i AP and a KPP 13 dBi omni on hand in case Micropop didn’t work out, but we didn’t need them.
I will say that the KPP omni has 2 degrees downtilt so it might have performed better than the L-Com antenna, also it had a nicer mount. (Personally, given the choice between downtilt and null fill, I’ll take null fill.)
It’s possible the old AP or omni might have degraded over time, although the omni had been replaced a couple years ago and looked almost new. But based on this one experience, I wouldn’t be scared off by the 9 dBi spec on the Micropop. Obviously it will be more attractive to you if you don’t need the unlock key, there’s a big difference between a $1000 price and $2500. It would be nice if the restriction was just on number of subs and not the max distance setting.
I’ll second the suggestion that it would be nice if the “lite” restriction was just on number of connected subs and not distance. I have a few of the micropop units that I keep wanting to deploy, but each time there is at least one subscriber just outside of the 2 mile distance limitation. I haven’t been able to justify the extra licensing costs in any of these situations, so install ePMP instead.
We also have a competing WISP that is kicking our butt on price because they use the Ubiquiti LiteAP GPS ($99) and LiteBeam CPE ($65). They can put up 4 sectors for under $400. From a hardware perspective it’s similar to the the sector version of the MicroPOP, but no unlock key needed.
They are sweeping up the new customers and taking away our existing customers because they have “disruptive” pricing. I’m tired of doing the install so they can come and swap a CPE radio and take the customer because of pricing. They are able to convince customers it’s because they are using new, superior technology.
“I’m tired of doing the install so they can come and swap a CPE radio and take the customer because of pricing. They are able to convince customers it’s because they are using new, superior technology.”
Not sure of your area, but here we use a term contract to help prevent this kind of thing from happening. Free install IF they sign a 24month contract and if they cancel within the first 30 days (mandatory cool off period here) then they are responsible to pay for the non-recoverable install which is fairly pricey, if they choose a shorter term, they pay for the portion of the install not covered by the contract, based on a 24 month max contract term. If they cancel after the first 30 days, then they have to pay out the contract at the current applied plan rate. You would be amazed at how quickly your client realizes that they will be out a fair chunk of cash just to switch to a NEW provider.
I would also be asking how they can afford to offer such low pricing and still keep afloat? Is this one time offer rates with them goin up once they have a sizeable stable client base? Better business model that you could adopt parts of? Better upstream costs? (this one is what my competitor tried, using DSL packages to provide shared uplinks. This does work but there is a breaking point!)
The UBNT LiteAP and the LiteBeam CPE are hardly NEW and it is questionable to say superior technology since they use the Atheros MIPS 74Kc, 533 MHz CPU with only 64MB ram. This is a 2x2Mimo radio with a fixed 120degree sector (60degree -6db roll off, -15db @120 degrees!), the MicroPOP is already a better radio! The epmp1000 is a comparable radio except it does the same speeds with an N based chipset! Cambium may not be on the bleeding edge of tech but at the same time they do more with whats stable than the competition does with the bleeding edge.
The only ways I know to beat these disruptive tactics once they are started is to either be distruptive yourself (the attrition game, who can last longest loosing money), show value to your service (the customer general help desk), superior plans that the competition cant carry at good price points or sign ups price breaks (first two months at a significant loss on a full term contract?).
Honestly, just better, more stable service with minor perks like actually providing good customer service (I get more clients from the competition over this issue than I can count!) generally proves to your clients that they made a bad choice. Initially we lost a few to the Starlink craze, all but 3 have come back (these 3 are clients you would give to the competition willingly), we just let those whom are unsatisfied go, but are always ready to welcome them back just takes time.
I do read these, and appreciate all the commentary. I would definitely lower pricing if I could. Undoubtedly, we’d sell more units, but at the cost of continued funding of the business… comments here have started conversations around many aspects of the product and the business though. I really do appreciate the community feedback.
We only dig at price because just about every month someone decides to use cheap hardware of specific branding to try undercutting and generally leave a bad taste in the publics mouth for wireless internet services.
The issue here isn’t the actual facts and truth, it is about what the customer perceives or believes. If the competition is telling the customer that they will have newer less expensive technology; the customer isn’t necessarily going to look into the manufacturers of the equipment to make their own determination. Their decision is going to be based on the sales pitch and how they feel about trust.
One of the challenges is that most manufacturers do just fine being the last hop of an Internet network. From the customer experience perspective, the customers I work with don’t know if their CPE is ePMP 1000, ePMP 3000, PMP450, or even Mimosa. Their services just work unless they are broken or oversubscribed.
[This is something that makes it very hard for me to buy more PMP450. If the customer can’t actually tell the difference, why should I spend significantly more on it? We can talk about it conceptually, but each of these systems still makes money for us and they have been for many years.]
You cant stop a greasy sales pitch without the customer having to weight the options against breaking a contract. Not having a service contract these days is about as dumb of a business practice as it gets. This peice of paper protects you, your equipment and you customer base rate of change. The other half of the equation is customer retention tactics, ever call the local cable company or satellite service to cancel? Hell, call Rogers Cellular and try to cancel service! They go out of their way to make sure you dont leave including temporarily lowering the price point to where they cant afford to keep you!
You cant afford to come down to their pricing for everyone, but you can afford to grease the ones that squeak. Just make sure you add a non-disclosure statement in your contract to prevent them from spreading the word about their specific pricing.
You can mitigate the claims of newer technology by basing the same or similar speed results with a more mature technology, which has been the Cambium method since inception.
There is a huge difference between the epmp series and the pmp450 series, subscription rate and spectrum efficiency. The epmp can have up to 128 (e1k) or 64 (e3k) SMs per AP but that doesnt mean that all of them can be serviced at full rate, in fact at about 15 SMs the epmp series starts st slow down. The pmp450 is more capable and can have 100 SMs and at 30SMs at full rate it still has power to burn! The pmp450 also has higher spectrum efficiency so a narrower channel carries more data thus you only need a 10MHz channel with a pmp450 where you need a 20MHz channel with epmp. Now consider humans like trees around their houses and a 10MHz channel goes through trees better than a 20MHz channel at the same frequency.
The way I see your problem, it isnt a technology issue but a people issue. If you customers want to leave your network, find out why. They will tell you, listen to them and offer a solution that works for them.
Quality internet service is not dirt cheap but doesnt have to be expensive either. Since 95% of internet subscribers over buy their needs, offer smaller more agile packages. If you got a family with young kids, a 10/2 package makes disney+ work on three different tvs at once at high def resolutions. Bump to a 15/5 to account for a facebook video call while the kids have their shows playing. Yes it is nice to be able to offer a 100mbps package and deliver it, but those few customers whom will actually use it, only need it for a few hours a month. If you have radius setup and are using eap-ttls you can make a web portal that will allow your customers to provide boosted services when they need it for a fraction of the cost of the monthly plan. They pay upfront for the requested time duration of the boost based on a standard 24hrs of boost and you reap the benefits. A little extra planning on install will qualify an install so that available boost packages are given to that customer based on what is in the database for max package speeds capable.
I can spit-ball ideas all day, but usually what works is the two hook method:
First: get that serivce contract in place at install and dont be afraid to remind the client that they have to pay out the remaining time of the contract to cancle.
Second: offer a solution based on the customer data (ie their usage stats) that is cheaper or offer a fixed term discount. $5 off for 3 months does actually work and the loss is minimal.
If a customer wants to leave after that then prepare the invoice and set a hardware recovery date and wait, they usually come back after a while to reconnect, which you get to charge a reconnect fee and an new contract term.
In the mean time, dig into your competition and learn about how they are doing things. Most likely they are operating at a huge loss to get a solid contracted client base where prices will rise to cover the costs. This is too common of a tactic and once you can confirm this it is an effective deterrent for customers to swtich away.
New relay site, PTP850c in and out, cabled for 4 sectors but this is more of a micropop and was waiting to see if we would get customer installs before deciding 5 GHz or CBRS and what kind of AP to install.
Got a customer, had a couple 450 omni micropops APs on the shelf so we’re going to install one of those. But of course, the customer is 2.3 miles from the tower. Grrrrr.
In my mind, micropop radius is 3 miles. Maybe a little more with CBRS. I understand an urban WISP might consider a much smaller cell radius, but then you’re probably not limited to 20 subs. Or maybe you go with 4 sector micropop APs and up to 80 subs and life is good.
Not sure what to suggest, but it seems I’m doomed to always buying an unlock key for these things because there will always be customers at 2-3 miles. Probably there is an ePMP product that would have been a better choice.
It’s also awkward we have to calculate timing parameters to coexist with our nearby towers using 10 miles distance parameter, because you literally can’t set max distance to >2 miles without the unlock key. Would an SM at 2.3 miles register with max distance set to 2 miles? Maybe. Would it be reliable? Don’t know. Would I rely on that? No.
I would have preferred if the limit was only on number of subs. Or that the distance limit was 3 miles, even 10 subs and 3 miles would have been better than 20 subs and 2 miles, but that’s not a change Cambium could make now, you can only open up the restrictions not tighten them. I guess they could also have gone with 2 keys, one to unlock the sub limit, another to unlock the distance limit. $1000 for the hardware and $750 each for the keys. But we all hate license keys, so I’m not going to suggest more keys.
Thanks for pointing this out, as this is a bug. As a work around, run Frame Calculator on each radio that you are colocating. This is always a good idea anyways as timing changes might happen in future firmware releases.