Opera's AM key-clicks ,

 

VERY  True , there  has to be an associated  bandwidth, however ,  hard  non-conditioned on off  keying  of  a  carrier  will  produce , key-click's  , this was the  subject of a  e-mail , sent to  me some time  back, over my  TX bandwidth  on  477 /op8 .

 

Notably , the  email  enquired as  to  how  my Tx b/w  was  in the  order of  10 MHz or  less and not  the  mode usual  400 / 500 MHz

 

Its quite  simple,  I use  the  audio  drive  , via a  totally  linear  TX path  ma1723 > skanti amp

 

What appears to be a  simple  task,   is often quite  complex , as is the  use of the  R-PI as a  rf carrier  source

 

The  integrity of the  timing loops [ as before] are  compromised  by the  higher order  IRQ requests ,  hence  setting  a  flag high/low , with  a  executed code routine  to  generate a 'stable'  carrier  source  cannot be relied on.

 

In the  development of the  Opera  R-PI tx code ,  the  action  of the  IRQ's could be  observed as random  noise , acting   cycle  by  cycle , hence any  long persistence  in the  spectrum plot , simply  masked the  spurs , however , the  Opera  dsp  returned varying  s/n  levels  for the  same  test  set up , indicating the  channel  background noise varied. [ opera decoder  is a  sampling system running far faster than the  data  rate ]

 

The  project was  shelved , other than for  hard on-off  keying  of a  free  running  carrier ,  in this the  clock  jitter is  of no consequence.  

 

73-G,