Log entry time 10:26:12 on June 9,2004

Entry number 125121

Followups:

• 125226 06/10/04 08:33 lkaufman Follow up to HAPPEx source install

Summary of Source work from June 7, 2004

Our main goal in our source work was to center the Pockels cell geometrically in order to reduce position differences due to the Pockels cell lensing effects.  We also wanted to make sure the optical elements in the beamline were aligned well to maintain good linear or circular polarization.  We also wanted to make measurements of our setup on the laser table in order to better understand what we see on the electron beam.

Here is a list of what we did and learned along the way:

1. Measured DoLP (degree of linear polarization) incident on the Pockels cell with the IHWP in and out.  In order to get high DoCP (degree of circular polarization) from the Pockels cell, it is important to have high DoLP going in.

IHWP out: DoLP ~ 99.97% => DoCP ~ 2.6%
IHWP in : DoLP ~ 99.74% => DoCP ~ 7.2%

In order to get better linear polarization with the IHWP in, Matt Poelker adjusted the pitch and yaw of the IHWP to make it more perpendicular to the beam.  Remeasuring the linear polarization of the light, it didn't get significantly better, DoLP ~ 99.76% => DoCP ~ 6.98%.  We also adjusted the roll of the IHWP so that its axes are aligned to rotate the incoming light's polarization by 90 deg.

2. Align the Pockels cell for best circular polarization.  This alignment is the normal procedure that the source group already does with a few checks along the way.  We obtained the best alignment of pitch and yaw of the Pockels cell (no HV on) by checking the extinction ratio between crossed polarizers.  We then put HV on the cell and used a spinning linear polarizer and power meter as our measuring device, we set the roll angle of the Pockels cell and quarter-wave voltage for each helicity state.  Our measurements of circular polarization seem too good to be true since they give us values better than the input DoLP.  For both IHWP states, the DoCP > 99.986% with measurements at best being around 99.999% which seems unrealistic since our incoming DoLP was only 99.97%.

3. From tests conducted in the Injector Test Stand (ITS) laser room, we have learned that the Pockels cell acts as a lens and steers the beam producing significant position differences as you go through different parts of the cell.  Fortunately, we also found that it is possible to center the cell in order to minimize these position differences.  The procedure of centering the cell to minimize these position differences was the most significant configuration change that we made to the source setup.  We conducted iterative translation scans of the cell in both x and y measuring position differences using a quadrant photodiode (QPD).  Once we had centered the cell, we ended up with position differences of ~ 25 nm in both x and y at a lever arm of 1.2 m.  The lever arm to the cathode is ~ 2 m; therefore we would expect to see position differences due to steering in the Pockels cell of ~ 40 nm at the cathode.  These results are consistent with what we have seen previously with this Pockels cell (Arwen) when we made these measurements in the ITS laser room.

Next, we put a linear polarizer in our beamline as an analyzer in order measure position differences due to birefringence gradients of the Pockels cell.  In the ITS laser room, I made these same measurements for Arwen once she had been centered to minimize steering effects and measured position differences of 5 um in x and 2.5 um in y with a PITA slope of 635 ppm/V.  When we made these measurements on the tunnel laser table, we measured position differences of 25 um in x and 16 um in y with a PITA slope of 660 ppm/V.  This is at least 5 times the size of the position differences that we expected based on our measurements in the laser room (position differences should be the same for similar PITA slopes.).

4. Next Matt aligned the RHWP so that it was perpendicular to the beam, and we rotated the waveplate with a linear polarizer in our beamline to verify at what angles the minimum and maximum transmission were located.  This confirmed that the waveplate is indeed a halfwave plate for our wavelength of 780 nm.

5. The polarized source group has a pickoff line on the laser table in the tunnel in order to simulate the beamline to the cathode.  We used this beamline to put our QPD at the approximate distance of the cathode in order to measure position differences we should expect to see at the cathode.  Our PITA slope was measured to be ~ 5 ppm/V during these measurements (probably due to the bounces of circularly polarized light from 3 mirrors).  We measured the IA slope to be ~ 200 ppm/V with no noticeable position differences induced at the cathode.  The PZT slopes were also measured to be 6-7 um/V with very little charge coupling ( <34 ppm/V, and only in PZT Y).We also performed low resolution (every 20 degrees) RHWP scans to check if we could see the charge asymmetry flip sign with the insertion of the waveplate.  It seems that it did, but it is not obvious.

The most important measurement in this "simulated beam on cathode" setup was the measurement of our position differences with the IHWP in and out.  With the IHWP in, we measured delta x ~ 140 nm and delta y ~ 630 nm.  With the IHWP out, we measured delta x ~ 45 nm and delta y ~ 704 nm.  These position differences in y are much bigger than we expect while x looks ok.  If the position differences are due to Pockels cell birefringence gradients, we would expect them to scale with PITA slope, so we expected to see delta x ~ 190 nm and delta y ~ 121 nm.  Position differences in x seem reasonable or at least expected based on the measurement of 25 um at a PITA slope of 660 ppm/V.  Position differences in y are much worse than expected (a factor of 3.5), so we ask ourselves from where do these position differences originate?  The y position differences are the same sign for both IHWP states, so perhaps if due to steering, they will cancel with IHWP insertion.

That was it for the tunnel.

Later that evening, Kent stopped by MCC to check on the progresso beam recovery.  Due to a hardware problem, the accelerator could only run beam in the 100 keV region, so we had a chance to take a little data.  The PITA slope was measured to be 7 ppm/V.  The position differences were measured to be ~ 600 nm in x and ~ 1.0 um in y.  These measurements are roughly 300-400 nm larger than what we measured in the tunnel "at the cathode."  This could partly be due to the fact that the bpms were not calibrated before the data were taken, but it is still obvious that position differences are bigger than we had hoped.  Perhaps a different angle of the RHWP will help reduce these position differences if due to birefringence gradients; and if these position differences are due to steering, they should cancel with the IHWP insertion.