• Excel spreadsheet for the FEL settings is here.
  • Common trajectory latest values for the X-ray mirrors are here.


Is there an issue? Escalation Path

  • Issue concerns X-ray beam delivery: see here
  • Issue concerns ECS: see here

Program

  • align the beamline with no CRL from source to Zyla
  • set and log all references in the preset file here
  • Timing
    • At TCC
      • Coarse timing
        • X-rays
          • insert prefocusing lenses only and go to the 45° Ti foil
          • send full X-rays and confirm a trace of the X-rays on the scope channel 2 while triggering off channel 1
          • record positions with a better than 20 ps accuracy on M2
        • SPL
          • close the iris to obtain a say 50 mic spot on target
          • with the GAIA timed out, but with full amplified SPL, remove the energy limiter
          • send the beam on the same 45° Ti foil
          • observe the optical trace from the SPL: if it cannot be seen, either trigger from the channel 2 and find how much we are off, or open a bit the iris
          • move the SPL timing to bring the SPL trace coincident with M2 on the rising edge
          • record positions for the VITARA and scope
      • Fine timing
        • close the iris to fully illuminate the imaging system for the spot size while the GAIA is timed out and the EL is OUT
        • move to the 45° Yag crystal extending from the pillar
        • send the optical laser then send the X-rays at full beam
        • timing should be within 20 ps, so changing the VITARA timing by 1 ns should allow us to see the image turn black. Closing the slits to 400 mic for the X-rays help see the localized change of indesx of reflaction
        • once you see a darkening, follow the procedure below to get timed
          • go back half the last step you just did (let's assume you are going negative on the vitara)
          • if you see no darkening, it means you passed t0 and you can go back positive again half the step size you did just before
          • if you see drkening, you didn't go far enough, so keep going negative until the darkening disappear
        • once the timing is done within 100 fs, close the slits to 50 mic and confirm spatial overlap with the pin at TCC
        • record positions of the VITARA for t0 and positions of the varioous targets to be able to go back quickly to this configuration and check timing
      • with the TSO imaging and the 45° YAG extending from the pillar,
    • At the time tool
      • Coarse timing
        • once timing has been done at TCC, timing at the time tool can start
        • insert the Ti foil of the time tool in the beam
        • send full X-rays on the foil and look at scope channel 2
        • if you see a signal, save it on M3, if not, trigger on this channel and see how far you are from the trigger on channel 1
        • observe the position of the timetool signal vs the X-ray signal taken on M2 at TCC: the difference will help you iterating for timing below
      • Fine timing
        • insert the yag crystal and send the SPL time tool beam onto the camera
        • align the cracks in the crystal to be in focus
        • send the X-rays full beam and look at any darkening of the image
        • if you see no darkening, it means you are still too early with the timetool
          • to change the timing of the time tool, you can first estimate how far you are by keeping the delay stages of the timetool where it is and move the VITARA only. The value you started from is the one needed to be timed at TCC, so this value will need to be set back once timing at the time tool has been done.
          • change the VITARA until you see a darkening, using the bracketing of the value found out above (when comparing M2 and M3)
          • once you have found darkening, do the usual bracketing techniques to narrow down the window to 100 fs and see the edge moving in the frame of the timetool
          • go a bit later with the SPL and close the slits to 50 mic with the FEL to center the time window in the image: t0 is shown at this cross
          • record the VITARA value and compare to the original value found for t0 at TCC: this difference is the amount which needs to be either added or subtracted in the delay arms of the time tool
        • get in the hutch and add/subtract the relevant beampath in the time tool: realign the beam path if need be
        • confirm VITARA is set at TCC t0 and send the SPL and X-rays
        • if you are close in timing (can be checked with the VITARA), use the delay line of the timetool to get to t0 at the cross set before on the timetool camera
        • record the delay line values and positions on the screen for timing
        • check signal strength with the other samples (e.g. thick SiN membranes)
      • Temporal calibration
        • open the slits to 2 mm in the temporal direction and 400 mic in the spatial direction
        • confirm you see the edge moving in the vue
        • move the VITARA to the bottom of the screen and record 1000 events at this position: note the VITARA value and the expected change in timing
        • save a reference of the timetool picture with no xrays for background sutraction: 1000 events is likely necessary
        • repeat the step above with 3-4 steps until you reached the top of the screen
        • plot the histogram of the edge position for each run on the same graph (X is pixel, Y is hist)
        • extract the centroid of the distribution at each time steps: this is the actual center of the jitter to be used for the calibration at the time set in the VITARA
        • extract the width of each histogram: this is the actual jitter of the X-rays vs the SPL and confirm it is the same at each time steps
        • plot the centroid (pixel) vs time (VITARA value) and interpolate with a linear function: the slope gives us the fs/px correspondance (this is the time calibration)
        • set the delay on the VITARA to a known amount and confirm by counting the pixel from the centroid to the cross on the screen and applying the fs/px value that it corresponds
        • if not, repeat the calibration process starting from the top of the screen and finishing at the bottom (problem might be caused by the VITARA backlash)
  • MXI alignment at 11 keV
    • confirm the alignment of the pins at about 0 mm position on X of the MXI hexapod
    • start alignment with x2 on the Zyla objective

    • set full energy beam without prefocusing lenses in to start with

      op.SiT(1)
    • move +14mm relative to this position to get to the 50 x 50 mic stack for 11 keV
    • if you are lucky, you might see light through the MXI
      • if yes, start aligning the lenses with the 'fish eye' method: intentionally move V to create a symmetric pattern on the image and align tip, then intentionally move U to create a symmetric pattern on the image and align tilt
      • once both axis are symmetric, close the slits to 50 mic and translate X and Y to center the image on the original beam position when the MXI was out
      • keep iterating until the feature is round
    • record positions of the hexapods at this Z value in the preset to be able to come back when necessary
    • save this position on the first preset of the virtual motor
    • align the virtual motor by moving Z 10 cm away and realigning the MXI following the steps above
    • once done, save the position on the second preset of the virtual motor
    • confirm alignment by moving 10 cm away in Z using the virtual motor only
    • save 1000 images of the beam going through the MXI and no target at TCC
      • confirm the XRT spectrometer is in
    • insert a Siemens star at TCC
    • reduce intensity to 1%
      op.SiT(0.1)

      or 10% if you see nothing

    • confirm you image the siemens star pattern correctly
      • play with the focus of the MXI until the image is the sharpest
      • update the virtual motor
      • save runs of 1000 images (could be 10 x 100)
    • you might try to image other samples
      • try the targets produced at Stanford and experience alignment procedure

Be CRL parameters

      • XRT sets
        • Set 1 (1500 mic)
          • @ 9.5 keV, FWHM at MXI lens is 500 mic
          • @ 11 keV, FWHM at MXI lens is 571 mic
        • Set 2 (2x1000 mic = 500 mic??) -not used
          • @ 9.5 keV, 62 m : too short focal length so not used
          • @ 11 keV, 88 m : too small beam so not use
        • Set 3 (1000 mic)
          • @ 9.5 keV, FWHM at MXI lens is 340 mic
          • @ 11 keV, FWHM at MXI lens is 460 mic
      • Hutch sets
        • Set 1 (2x100 + 1x200 mic) + 400mic pinhole
          • @ 17 keV, FWHM at MXI lens is 374 mic
        • Set 2 (6x300 mic) : 800 mic pinhole
          • @ 9.5 keV, FWHM at MXI lens is 286 mic
          • @ 11 keV, FWHM at MXI lens is 416 mic
        • Set 3 (5x300 mic) : 800 mic pinhole
          • @ 9.5 keV, FWHM at MXI lens is 371 mic
          • @ 11 keV, FWHM at MXI lens is 480 mic
      • MXI sets
        • Set 1 (50 x 50 mic)
          • @ 9.5 keV, f = 1.323e-01 m
            • magnification: 4.5/0.132 ~ 34
          • @ 11 keV, f = 1.774e-01 m
            • magnification: 4.5/0.177 ~ 25
        • Set 3 (25 x 50 mic)
          • @ 9.5 keV, f = 2.646e-01 m
            • magnification: 4.5/0.264 ~ 17
      • Viable combinations

Alignment procedure

  • Prepare the hutch for the shift (done 1h before the actual start time)
    • open the grafana dashboards located here (requires UNIX login/pwd) to access beam owner, energies and coatings
    • no gate valve in the trajectory
      • check in the vacuum windows that no gate valves from the NEH to the MEC hutch are IN (red)
      • it is ok that the DG2 STP 1 is IN
      • it is ok that the GL window is IN in MEC

    • open the rolling status and confirm that no devices upstream the hutch obstructs the beam

      op.rs()

    • insert yag3 to provide a photon terminator before TCC prior to send any beam in the hutch

      op.yag3.insert()
    • confirm that the chamber is loaded with targets and being pumped down to be ready at the start of the shift
    • search the hutch but leave SH6 IN (you are not the beam owner yet)
  • 15 min before the shift begins, call ACR at x2151 to check
    • look at current ACR beam operator on shift here
    • confirm photon energy
    • confirm pulse energy
    • confirm pulse duration
    • confirm beam mode
    • confirm multiplexing mode
    • ask for an e-loss scan
  • transition to beam ownership
    • ACR calls you to confirm they are tweaking the beam for you
    • confirm with ACR that MR1L0 and MR2L0 have the right coating following this page
      • if moving the mirror yourself, then open MR1L0 and MR2L0 HOMS GUI located in the mechome > LCLS tab > HOMS overview button
      • set the coatings to the appropriate material as a function fo the photon energy as per the page from above
  •  getting ready to accept the beam:
    • close DG2 STP 1
    • make sure the reference laser is out (check in the rolling status)
    • confirm the target chamber is pumped down
    • insert the Be window IN the beamline
    • search H6 if not already searched
    • remove SH6 OUT of the beamline
  • beamline alignment
    • ACR calls you to confirm beam is ready for alignment

    • force close the pulse picker to make sure it does not let the beam propagate to the hutch yet

      op.pp.close()
    • check MR1L3 mirror OUT (-6000) position (XCS mirror)
    • check the XPP slits (slit1) are open (20mm, 20 mm) in the rolling status
    • open the mirror settings located here
    • confirm bending values for MR1L0 and MR2L0 (advanced)

    • insert YAG0

      op.yag0.insert()
    • confirm FEL beam is on the cross for the OUT position (undeflected beam)
    • confirm the shape of the beam is round (advanced)
    • set the MR1L4 coating as per this page
    • insert MR1L4
    • confirm FEL beam is on the cross for the IN position (deflected beam)

    • insert YAG1

      op.yag1.insert()

    • remove YAG0

      op.yag0.remove()

    • remove all Si attenuators (send full energy beam)

      op.SiT(1)

    • open the pulse picker

      op.pp.open()
    • fine tune the pitch of MR1L4 to center the beam on YAG1

    • insert YAG3

      op.yag3.insert()
    • remove hutch Be CRL (Mechome > Beamline > Beamline CRL(hutch))

    • open slit 2

      op.slit2.move(5)

    • remove YAG1

      op.yag1.remove()
    • fine tune the pitch of MR1L4 to center the beam on YAG3
    • to adjust the height using the YAG3 red cross centered at (265, 282), call ACR, and for a 200 mic motion on the yag, ask them to move 200/4 mic up or down.
      • you could set SiT(0.2) to not saturate the image
      • you could set 10 images averaged to get a cleaner picture
    • log the mirror settings in our table here
    • confirm photon energy and lens stack to use and log the change of stack in the paper authorization document for Be CRL log

---------- not for SPL ---------

  • once they are done tuning
    • timing check
      • turn off laser triggers

      • go to the titanium foil:

        op.ti()
      • move hutch CRL out

      • set slit4 to 400 mic: 

        op.slit4.move(0.4)

      • send full beam on titanium: 

        op.SiT(1)

      • set the EVR to 10 Hz settings: 

        op.lpl_check_timing(rate='10Hz')
      • set vertical division to 10 mV/div on oscilloscope Lecroy 1
      • once you see the signal of the X-rays, set 100 sweep to average the signal
      • save it on memory 2
      • set back sweeps to 1 on channel 2
      • move target about 0.8-1 mm negative to target the Al frame
      • turn on the LPL trigger on, event code 43 for 10Hz
      • once you see the signal of the LPL, set 100 sweep to average the signal

      • move timing to overlap the LPL with the memory trace 2 using the python command

        op.nstiming.mvr(2e-9)

        to move the LPL 2 ns later than the FEL, but this is just an example! Move it (or not) by the necessary amount to overlap best the rising edges of the beams.

      • save the current value as our t0, using hte python command 

        op.nstiming.save_t0()
      • once timed, set 100 sweep to average the signal and save the trace on memory 3
      • take a screenshot and put it in the elog
      • move back the hutch Be CRL
      • set the slits back to data-taking move size

      • set the EVR to single shot settings: 

        op.lpl_check_timing(rate='single')
      • set the scope Lecroy 1 channel 2 voltage/div to the maximum (1V/div) to be able to observe the shot on the diode and monitor the timing

1 Comment

  1. Eric M Galtier

    Did the timing at TCC without the prefocusing lenses.