After setting timing witht he diode, we look at the PAM spectrometer and scan the laser timing a bit.  We found a clear change in the spectrum - at >t0, the spectrum was very attenuated and at t<t0, the spectrum was larger.  This persisted for many 10s of ps (very unexpected behaviour).

In attachment are the LA3 timing scene, the SQS delay stage, the BAM, and the spectrometer signal from Metro at t0. att. 1 - 4.  The averaged ratio shows the cut the best.  Unfortunately we do not have references on that, so we will ask Sergey or Bjoern to do it tomorrow.

We found that the intra-train timing feedbacks of the accelerator are off. We ask BKR to fix it, which took a while. Aparently, we have to communicate the need of the timing feedbacks (and BAM) more clearly.

Standard deviation along train is now 18 fs with feedbacks on, see att 5. Was around 100 fs without the feedback!

We take a run and manually move the PAM delay stage so we have a timing calibration for PAM.

run 32: Rubbish because DA03 got stuck.
run 33: scanning PAM delay stage manually. device: SQS_AQS_LAS/MOTOR/WAVEPL_LMD2_1030. Typically taking 5um steps.

Went to the PAM delay stage position shown in att. 6, and had the PAM signal shown in att. 7