To properly configure the timing of the traces on the oscilloscopes:

  • first have the instrument scientists co-time the four legs using the streak camera/etc.
    • the east trombones are still motorized but the west trombone is not (sacrificed for Newport motor card slot); everything gets timed to the one leg without a trombone, with the west stage manually tuned (i.e. b/c it has no connection) and the two east stages tuned electronically
  • once the co-timing at TCC is done, use either the external trigger delay offset setting of the scope (or move the DG645-supplied timing of the external trigger if preferred, but this has to be done on-shot, which is less convenient than changing the delay offset setting of the oscilloscope, as this can be done post-shot with old buffered data) to move the traces such that the fiducial pulses are relatively close to the final division on the scope
    • (side note the scope MUST be set to 5ns/DIV time base!)
  • once all the fiducial pulses are relatively close to the final division on the scope, use the "deskew" setting on each individual channel to move each trace such that the fiducial pulse lines up exactly with the final division on the scope
    • it may not be necessary, but sometimes I like to read out the raw data of the scope using LOSC('2').pchall() (for example) so I can get very precise with exactly how much deskew I need to exactly match up the fiducial pulse of a given trace with the final division (which is at the position of nine-tenths of length of the waveform record)

A few other notes:

  • This procedure is necessary for ALL laser scopes! ALL scopes used for pulse shaping MUST be set to 5ns/div and overlap the fiducial pulse with the final division! 
  • After executing the procedure, the rising edge of the pulse should appear to be close to the first division of the scope. 
  • The fiducial pulse can be turned off and on using the HAWG().FidOff() and HAWG().FidOn() functions in the meclas Python package.
  • This entire process could be automated with just a little bit of time... It would not be hard to read in a waveform from the scope (using LOSC.rch), calculate the position of the fiducial peak in time, calculate the offset required to deskew the fiducial pulse to the desired location, and then push this time offset to the deskew field (using LOSC._ctrl)... 


Wrong:


Right:

  • No labels