RW> DF> NA>CB>SSB does away with one sideband and the carrier....this leaves
RW> DF> 1/3 the width
RW> DF> I believe this to be incorrect. The correct value would be 1/4 not
RW> DF> 1/3.
RW> DF> Half the power would be in the carrier, and the other half divided
RW> DF> equally between the sidebands, or 1/4 of the total power each.
RW>It is not *power* that requires the bandwidth. It is the type of
RW>emission.
  Understood perfectly, mistook the 1/3 width for 1/3 power. sorry
RW>CW requires less than 1 khz of bandwidth for transfering
RW>*intelligence*.
  Much less, bandwidth for cw would be dit rate dependant, the faster you
go, the wider the bandwidth required to prevent ringing in the recovery
phase. 50 words a minute will fit well in a 300hz passband filter.
RW>"modulation product".  For lower sideband (LSB) it is below the center
RW>freq. For upper sideband (USB) it is above the center freq.
Yeah, done my time with collins mechanical filters, and even done the
phasing method as well.  Anyone remember the B&W 2Q4 module ? 
RW>Most SSB transmitters and microphones "shape" the audio somewhat to
RW>*limit* the passing of frequencies from 400 to a maximum of about 2700
  Ain't digital processing and rf clipping just great.  Shame many ops won't
take advice when they are told they are splattering all over the band, but
better design with end users in mind is relieving this to a degree.
RW>CW may be operated within 1 khz of a band edge.
  Advancements in counter tech. and synthesizer usage  have made
operation at band edges a mere formality these days. When I started,
within 2 kc near a band edge marker, or 5 khz if a marker was not
available was the best you could hope for, unless you had lots of
money to spend !
Regards,  Dave
Internet: dave@svcity.usbbs.com
 
... Break out the BC 221's .....
--- PCBoard (R) v15.3/2
---------------