On 7/7/2014 6:53 PM, Adam Greene wrote: > Fred, > > I think one aspect of the new 15.407 (U-NII) rules that UBNT may not yet > meet is the 40MHz filter requirement on both ends of the 5725MHz-5850MHz > spectrum, which as I understand it, will effectively limit the usable range > to 5765MHz - 5810MHz. Or maybe they already have the filter? In any case, > the range reduction will still mean replacing some existing deployments with > different frequency gear, if I'm interpreting the new rules correctly. :( > > Thanks, > Adam >
I did a little research on that today, just enough to be dangerous. The requirement in 15.407 that drives everyone nuts is that out-of-band emissions must be at -17 dBm/MHz EIRP at the band edge and -27 dBm/MHz at 10 MHz. That translates to 40 dB at the edge if you use the +53 dB limit of the 5150-5250 band. Compare to 15.247 which is 30 dB relative to the desired signal, not to a fixed EIRP. It's trivial to meet 30 dB. It's harder to meet 40 dB. So either you need more distance from the band edge or a filter, but the filter is frankly impractical, almost a straw horse argument. So I pulled the FCC's type approval report for UBNT's NBM5, which in fact has 15.407 approval. The lab report includes spectrum analyzer plots. As I read it, that radio actually does seem to meet 40 dB suppression at around 10 MHz from the edge of a 20 MHz 11n signal, so (with 5850 the limit) it could be centered at 5830. Perhaps I'm misinterpreting it, but it looked mighty clean, and since the antenna is integral, it doesn't have to worry about EIRP. However, the U-NII plots were being done down in the 5250 band, where the tested power output was only around 0 dBm, since EIRP is capped at +30 (less if <20 MHz wide) and there are two chains feeding a 25 dB dish. It might not be quite so clean at full power, what's now allowed on 5150 and what's proposed for 5725. Heck, at 0 dBm the final amp an run in full Class A and not get warm. (Mimosa's petition notes that a Class A amp is usually around 5% efficient, the more common Class AB about 10%. I'm sort of surprised that microwave GaAs is not more efficient; I'm used to the much higher efficiency of HF ham and broadcast-band transmitters.) Anyway, there are two petitions, Mimosa's and WISPAs. Mimosa just wants the interference cap modified so that if antenna gain is >6 dB, the unwanted signal can go up by 6 dB. That results in a 40 dB ratio from a +30 transmitter, or 30 dB from a +20 transmitter. And the ability of a professional installer to mix'n'match antennas is saved. So I'm working on a Comment in favor of both petitions. > -----Original Message----- > From: wireless-boun...@wispa.org [mailto:wireless-boun...@wispa.org] On > Behalf Of Fred Goldstein > Sent: Thursday, July 03, 2014 1:03 PM > To: wireless@wispa.org > Subject: Re: [WISPA] UBNT RocketAC spotted on FCC site > > On 7/3/2014 9:33 AM, Ben Moore wrote: >> $135 MSRP for rocket-lite. > That's excellent. One of the contractors working with us recently replaced > a pair of old Motorola PTPs with NanoStation Ms. It's just a camera, so it > doesn't need much speed, so when I found its wireless side converging at 270 > Mbps (the bottleneck is the Ethernet), I turned it down to a 20 MHz channel > so it's merely 130 Mbps. And I moved it down to DFSland, where the AP side > properly moved the slider all the way to the right at +14 (since the antenna > gain is 16 dB). But lessee... the old Motorola charged extra for allowing > speeds above 25 Mbps, extra for encryption, and cost about 50 times as much > as the UBNT to begin with. > > Oh, but the PTP had a metal body, unlike the nano. But the new Rockets are > metal too. So really, it's embarrassing -- if you're the one still trying > to sell at the old Moto price points! > > Not to rain on the sunshine here -- but I did see one issue when I actually > read the FCC test report. It was only being tested for the > 15.247 band (5725-5850), not U-NII. At least the old PTPs had DFS (with > separate SKUs needed to use the DFS and non-DFS channels!), and the plain > NanoStation does. So will the Rocket-lite have U-NII support? > That could include either or both of the DFS bands and the new UNII-1 band > at 5150. > > I also notice that WISPA is petitioning to have the 15.407 (U-NII) rules > changed to be easier to meet. But the NanoStation, Rocket M, and NanoBridge > already do, at low cost, so does UBNT know more than its competitors do, or > are the new rules harder? > -- Fred R. Goldstein k1io fred "at" interisle.net Interisle Consulting Group +1 617 795 2701 _______________________________________________ Wireless mailing list Wireless@wispa.org http://lists.wispa.org/mailman/listinfo/wireless
