MMA Memo Abstract Template MMA Memo Abstract Template # abstract.template -- MMA Memo Series Electronic Abstract Template # # Comments are preceded by the "#" symbol. Anything following a "#" is # ignored. # # The following entry is for internal use only. Do not fill it in. #Number: # # Enter the MMA memo title in the following section # #Title: Fringe Tracking, Sideband Separation, and Phase Switching In the ALMA Telescope # # List the MMA memo authors in the order that they appear on the paper. # Use a comma to separate the names of multiple authors. List each author's # affiliation in parenthesis following the name. # #Authors: Larry R. D'Addario (NRAO) # # Enter the date that the memo was submitted in the format mm/dd/yy. # #Date: 02/18/00 # # Enter a list of keywords which might be used to find this memo in an # electronic search. Multiple keywords should be separated by commas. # #Keywords: phase, switching, fringe tracking, fringe rotation, sideband, Walsh functions, shifted-m sequences, correlator, local oscillator, LO, DDS, direct digitial synthesizer # # Enter a plain ASCII abstract between here and the #END symbol # #Abstract: Methods of implementing fringe tracking, sideband separation, and phase switching in the ALMA telescope are described and analyzed. All required features make use of direct digital synthesizers for precise phase control of the first and second local oscillators. Sideband suppression is achieved via frequency offsets. Phase modulation of 180d is cancelled by bit inversion in the digitizers; it requires no action by the correlator, and a complete period of 64 mutually-orthogonal waveforms can be made to occur within one correlator dump time. Post-correlation sideband separation is not supported; it could be added by 90d phase switching of the first LO, but 64 correlator dumps would be needed for a complete period. Synchronization requirements among antennas and between each antenna and the correlator are also considered. A novel choice of orthogonal switching waveforms (linear-frequency square waves, rather than Walsh or Rademaker functions) may make synchronization unnecessary. Finally, some detailed design considerations for DDS control are discussed in an appendix. #END