--- In [email protected], Matthew Kaufman <matt...@...> wrote:
" Reducing the modulation index and simultaneously reducing the receiver bandwidth from 5 to 2.5 kHz results in a situation which requires ~6 db more signal level for the same demodulated quality (ex. 12db SINAD)" -------------------------------------------------------------------- I believe that this is a point which requires further evaluation. As Jacobsmeyer points out in the article "CUT YOUR LOSSES", TSB-88 documents the need for a 6 dB increase in C/N required for 12 dBS at narrowband versus wideband but that does not equate to "6 db more signal level for the same demodulated quality (ex. 12db SINAD)" Signal level [carrier] is an absolute whereas noise is not. Any stated noise level value MUST be considered in the context of a reference of some sort - for example - one might state "The noise measures -155 dBm/Hz." This value is meaningful because it is referenced to a specific bandwidth - in this case normalized to a one Hz ENBW [Equivalent Noise BandWidth] An examination of the 6 dB difference in C/N stated in TSB-88 should lead one to consider the "N" value for wideband versus the "N" value for narrowband. The "N" in each of these cases is actually the value of noise within the receiver's ENBW. As others have stated, narrowband receivers have narrower IF filters and therefore, the ENBW is less. Assuming a receiver IF bandwidth optimized based on Carson's rule, the result is that the value of "N" is 3dB less. [For a specific radio where the only difference is wideband versus narrowband, in absolute terms of dBm/Hz, the value of "N" would be constant and is based on the receiver's Noise Figure] Thus, the actual impact of changing an analog system from wideband to narrowband is not 6dB but 3dB. In fact, 3dB is the value that has been stated by APCO engineers in their narrowbanding presentations.
