A suggested beamline alignment procedure

The thing I like to keep in mind is that the center of the undulator beam is best defined at photon energies corresponding to odd harmonics (1, 3, $\ldots$) and it is very important to work at such energies; see the movie. The other thing is that the shutter photodiode has a broad collection area so it is a great, steady, microscope-and-mirror-independent alignment detector.

So here's how I like to do it:

1. Close the shutter in the SM program to get the shutter photodiode into the beam. Hook up a picoammeter to its signal, and run the analog output from the back into an ADC channel. Pick the appropriate picoammeter fixed scale (not auto). Open up the vertical exit slit with motor_move exsy_o to set it to a position of 0.

2. Take a spectrum with a command like

   % sgm_scan adc0 250 600 1 -outboard
   

3. Now go to the peak (let's say at 294 eV) by using the program sgm_move.

4. Now take a entrance slit scan with (for example)

   % motor_scan ens_o adc0 24000 25000 10
   \begin{verbatim}
   Find the center of the peak and go there using the
   \texttt{motor\_move} command.
   If you were significantly off, you may need
   to iterate the steps of going to the peak energy and recentering
   the entrance silt.  Note that 24500 for \texttt{steps} will
   of course need to be modified for where you actually think
   the slit is at.
   
    \item Still using the shutter photodiode, center the vertical
      exit slit with
   \begin{verbatim}
   % motor_scan exsy_o adc0 23000 24000 10
   

   where again the actual position range of the scan will vary.

5. Now you have the entrance and vertical exit slits well centered on beam at the undulator peak. You can now open the shutter photodiode in the SM program. Move the quadrant photodiode down by about 0.15'' so that one segment is fully in the beam (the signal does not vary much as you move it up or down) and then move the order sorting mirrors up and down. Record the positions at which you have half the maximum signal, and go to the midpoint between them. At the end, move the quadrant photodiode until the top and bottom signals are balanced. Now you have the mirrors well centered on the beam, and the hole in the quadrant photodiode well centered as well.

6. Next you should close down the exit slit to something really small like 5 or 10 $\mu$m, turn off helium, and remove the zone plate and 5 $\mu$m pinhole. You now have to iterate moving the snout inboard/outboard and up/down to find the half maximum points and pick the center, then move the counter up/down inboard/outboard to find the center, and iterate between snout and counter a few times.

7. Now put in the 5 $\mu$m pinhole, open up the slits, and scan the exit window to check for evenness of illumination. You may have to tweak the mirrors around by 50-200 $\mu$m, or do some other sort of fun and games.

Now you're ready to go on to STXM alignment.