Hi, Had to wake up and check this. Your math is correct. There are some subtleties regarding the shape of the noise relative to spectrum...these numbers assume an even noise distribution across the band. If you added 2 dB to the received signal (or aimed your antennas to pull the noise floor down to -77), you would bump to 16QAM. At -75 dBm noise floor, you have 22 dB of interference/noise inside the path of the directional antennas.
Having said that, if you choose an Rx and Tx frequency that are very close to each other and you are running in full duplex mode, the noise floor will rise as a result of the transmitter being very close to the receiver. This noise floor change is modeled in the link calculator for AF5 radios - the calculations rely on large internal look-up tables of actual measured data and can not be easily calculated on paper. There is currently no calculator for the AF24 radios as any such effects are already built into the spec sheets (they only have 2 frequency choices). We recommend using that. I will note that we have calculated this using only the data on the spec sheet and a knowledge of the natural RF noise floor of various channel bandwidths. Chuck tl;dr...the natural noise floor for the band minus the sensitivity of that band for a given modulation equals the signal needed over the noise floor for a given modulation. On Mon, Jan 12, 2015 at 10:28 PM, Jeremy <jeremysmi...@gmail.com> wrote: > So, am I understanding correctly... If you are on a 50MHz Channel BW with > a signal of -58 you need a noise floor of -97 to maintain 1024QAM. So, if > the noise floor is -75 we are 22db off so you can add that to -58 and > figure -80, or QPSK. Is this the correct math? > > On Mon, Jan 12, 2015 at 8:44 PM, Chuck Macenski <ch...@macenski.com> > wrote: > >> Hi, >> >> I believe the number you are looking for is the sensitivity per >> modulation rate minus the base noise floor for a given bandwidth (see >> Matt's earlier post). This gives you the required signal over the noise >> floor. If the noise floor goes up, the effective sensitivity goes up >> accordingly. >> >> What am I missing? >> >> Chuck >> >> On Mon, Jan 12, 2015 at 5:23 PM, Josh Reynolds <j...@spitwspots.com> >> wrote: >> >>> This. >>> >>> josh reynolds :: chief information officer >>> spitwspots :: www.spitwspots.com >>> >>> On 01/12/2015 11:00 AM, Mike Hammett wrote: >>> >>> Matt, that doesn't really help. >>> >>> In many areas, my noise/interference/whatever is -75. If I knew how much >>> margin was required per modulation, I could figure out how much signal I >>> need for that speed. >>> >>> I can't do that with your published numbers. >>> >>> >>> >>> ----- >>> Mike Hammett >>> Intelligent Computing Solutionshttp://www.ics-il.com >>> >>> ----- Original Message ----- >>> From: Matt Hardy <m...@ubnt.com> <m...@ubnt.com> >>> To: af@afmug.com >>> Sent: Mon, 12 Jan 2015 12:42:52 -0600 (CST) >>> Subject: Re: [AFMUG] Af5 >>> >>> Hi Mike, >>> The numbers in our datasheet do convey the SNR (which assumes AWGN for >>> noise value); this is pretty standard among all our datasheets. >>> The noise floor value varies based on channel BW, so a rough estimate for >>> noise floor would be (50MHz=-97dBm, 40=-98dBm, 30=-99dBm, 20=-101dBm, >>> 10=-104dBm). >>> >>> As far as transmit power / modulation, all modulations up to 64QAM can run >>> at full transmit power. 256QAM is ~2dB lower, and 1024QAM is something like >>> 4dB lower (it depends on other external factors). >>> >>> All of this and more is already provided with the integrated link >>> configuration tool, including channel bandwidth, duplex frequency spacing, >>> TX power and a bunch of other stuff. It's recommended to use this tool to >>> configure links because of all of the additional factors. >>> >>> But if you need to run a simple link calc to estimate a link's performance >>> at certain distance, you should have enough info. >>> >>> Hope that helps... >>> >>> -Matt >>> >>> >>> On Sat, Jan 10, 2015 at 10:06 AM, Mike Hammett <af...@ics-il.net> >>> <af...@ics-il.net> wrote: >>> >>> >>> Not at all, Matt. Those are receive sensitivities. >>> >>> >>> The left side of this picture is needed for the airFibers. You publish the >>> right side only. >>> >>> >>> This information is also not available for the airFibers. >>> >>> >>> >>> ----- >>> Mike Hammett >>> Intelligent Computing Solutionshttp://www.ics-il.com >>> >>> ------------------------------ >>> *From: *"Matt Hardy" <m...@ubnt.com> <m...@ubnt.com> >>> *To: *af@afmug.com >>> *Sent: *Friday, January 9, 2015 5:34:43 PM >>> *Subject: *Re: [AFMUG] Af5 >>> >>> This? >>> [image: Inline image 1] >>> >>> On Thu, Jan 8, 2015 at 5:18 PM, Mike Hammett <af...@ics-il.net> >>> <af...@ics-il.net> wrote: >>> >>> >>> Hard to tell because they won't publish their Tx and SNR requirements. >>> I'd just make educated guesses and plug it into a path calc. >>> >>> BTW: Ben, Matt... where are those numbers? >>> >>> >>> >>> ----- >>> Mike Hammett >>> Intelligent Computing Solutionshttp://www.ics-il.com >>> >>> ------------------------------ >>> *From: *"Craig House" <cr...@totalhighspeed.net> <cr...@totalhighspeed.net> >>> *To: *af@afmug.com >>> *Sent: *Thursday, January 8, 2015 3:24:58 PM >>> *Subject: *[AFMUG] Af5 >>> >>> I have used some AF 24 hours but no AF5 for backhaul yet what is the max >>> distance people are seeing for the AF 5 at full modulation >>> >>> Sent from my iPhone >>> >>> >>> >>> >>> >> >
