Chuck is right. The system noise figure is a combination of amplifier gains, amplifier noise figures, and filter losses, so best just to refer to the manufacturer specs for comparison. The sensitivity specs will include the effect of noise figure, in case the noise figure is not given.
On Thu, Sep 21, 2017 at 11:15 AM, Chuck McCown <ch...@wbmfg.com> wrote: > I would add front end filter loss adds directly to noise figure which > affects sensitivity. > > *From:* Chris Gustaf > *Sent:* Thursday, September 21, 2017 12:11 PM > *To:* af@afmug.com > *Subject:* Re: [AFMUG] radio specifications... HOW? > > You are correct that wider front end filters give up masking and rejection > of out of band interferers, but the intermediate frequency (IF) and > baseband filters are what determines RX sensitivity. > > In the case you outlined, if the channel width was say 56 MHz for both > QAMs and the tuning range was 200 MHz or 1000 MHz, there would be no real > difference in sensitivity for the same modulation/channel width. The 1000 > MHz tuning unit range may be more susceptible to interference outside 200 > MHz filter range is all- this depends on other radio design elements. > > You can try asking the supplier for a Threshold to Interference (T/I) > curve, which will show how strong a like signal at various frequencies will > cause BERs on the radio. Most licensed radio manufacturers should have > this data. > > > Chris > > > > > On Thu, Sep 21, 2017 at 10:22 AM, Josh Reynolds <j...@kyneticwifi.com> > wrote: > >> It is my understanding that the larger the operational range of the >> radio, it is using vastly different filters to accommodate that range. >> These filters give up masking performance outside of the channel mask for >> an increase in range. >> >> Those wider range filters end up costing the radio *selectivity*. >> >> If this is incorrect, please say so. >> >> On Sep 21, 2017 12:07 PM, "Chris Gustaf" <ch...@trangosys.com> wrote: >> >>> In general for each increase in modulation level it requires about 3 dB >>> more SNR (see below), and for each doubling of channel bandwidth you lose 3 >>> dB sensitivity. >>> >>> SNR Requirements for Trango and others: >>> >>> QAM/SNR (dB) >>> 256QAM/26.1 >>> 512QAM/29 >>> 1024QAM/32.5 >>> 2048QAM/35.8 >>> 4096QAM/40 >>> >>> The frequency tuning range of the radio has nothing to do with the >>> sensitivity. >>> >>> So 4096QAM requires about 14 dB More SNR than 256 QAM for the same >>> channel size, but since in this case it uses a smaller channel bandwidth it >>> has a lower noise floor by 7 dB as Forrest stated. The 4096 QAM radio will >>> require a 7 dB stronger signal to hold its modulation.. >>> >>> Hope this helps >>> Chris >>> >>> >>> >>> >>> >>> >>> >>> On Thu, Sep 21, 2017 at 5:24 AM, Josh Reynolds <j...@kyneticwifi.com> >>> wrote: >>> >>>> I would rather not name the vendors at this time. >>>> >>>> On Sep 21, 2017 2:30 AM, "Mitch Koep" <af...@abwisp.com> wrote: >>>> >>>>> Josh >>>>> >>>>> Will you shall the radio info? >>>>> >>>>> Mitch >>>>> >>>>> On 9/21/2017 12:57 AM, Josh Reynolds wrote: >>>>> >>>>> The frequency agility between the radios is 800 mhz. Vastly different >>>>> radio filters. >>>>> Both are capable of similar channel bandwidths. >>>>> This is all done at MCS0 / QPSK >>>>> >>>>> I'm comparing, basically, their sensitivity and selectivity. The radio >>>>> that is frequency capable of 200mhz is also capable of hitting 4096QAM, >>>>> while the radio capable of operating in a 1GHz range of frequencies is >>>>> 256QAM capable. >>>>> >>>>> Basically, it's (the 256QAM radio) sensitivity at the same modulation >>>>> and the same power is wayyyyyy closer than I would expect it should be - >>>>> vastly different, and far cheaper components. TL;DR: I'm calling BS on the >>>>> data sheet :) >>>>> >>>>> On Thu, Sep 21, 2017 at 12:23 AM, Forrest Christian (List Account) < >>>>> li...@packetflux.com> wrote: >>>>> >>>>>> Can you give specific examples? Having a hard time understanding for >>>>>> sure the exact specs you're comparing. >>>>>> >>>>>> In relation to the thermal noise floor: just reducing from 1000mz to >>>>>> 200mhz will gain you ~7db of noise floor. But usually that's in a >>>>>> channel, not in the entire 'frequency agility' area. Maybe they aren't >>>>>> all >>>>>> that selective within the 1Ghz bandwidth. >>>>>> >>>>>> I've never been able to find a chart of theoretical required s/n >>>>>> ratio for each of the QAM's so I can't comment on how much difference >>>>>> there >>>>>> is supposed to be - after all, with everything else being the same >>>>>> (channel, modulation, power, etc), 256QAM should definitely require a >>>>>> lower >>>>>> signal strength than a 4096QAM radio. They definitely shouldn't be the >>>>>> same with the same channel width, unless one radio is noisier or more >>>>>> susceptible to noise. >>>>>> >>>>>> And sensitivity should just be about the receiver, not the >>>>>> transmitter. >>>>>> >>>>>> On Wed, Sep 20, 2017 at 9:20 PM, Josh Reynolds <j...@kyneticwifi.com> >>>>>> wrote: >>>>>> >>>>>>> Can someone smarter than I fill me in on something? I'm comparing >>>>>>> some >>>>>>> radios here (no names...) >>>>>>> >>>>>>> One radio is 256 QAM, with a 1000mhz operating range >>>>>>> >>>>>>> Another one is 4096 QAM, with a 200mhz operating range >>>>>>> >>>>>>> Can you explain to me how the sensitivity on the 256QAM radio, at the >>>>>>> same modulation rate, same (scaled) power level, claims to be with a >>>>>>> single dB or two as sensitive as the 4096QAM radio with an 800mhz >>>>>>> smaller operating range? >>>>>>> >>>>>>> Anyone? >>>>>>> >>>>>>> Thanks :) >>>>>>> >>>>>>> --- >>>>>>> Josh Reynolds >>>>>>> >>>>>> >>>>>> >>>>>> >>>>>> -- >>>>>> *Forrest Christian* *CEO**, PacketFlux Technologies, Inc.* >>>>>> Tel: 406-449-3345 | Address: 3577 Countryside Road, Helena, MT 59602 >>>>>> <https://maps.google.com/?q=3577+Countryside+Road,+Helena,+MT+59602&entry=gmail&source=g> >>>>>> forre...@imach.com | http://www.packetflux.com >>>>>> <http://www.linkedin.com/in/fwchristian> >>>>>> <http://facebook.com/packetflux> <http://twitter.com/@packetflux> >>>>>> >>>>>> >>>>> >>>>> >>> >
