Blog

We have 2-color delivery in SASE2 this week. The SQS team will call every now and then and ask you to switch off either the first or the second color. 
This can be done by changing the linear taper of the respective color to -400. Of course you also have to press  the "Set" button afterwards to apply the change. The trajectory feedbacks SASE3 and SASE32 will always keep the trajectory the same. 

Please ensure that the photon energy selector (that is the info send to the XGM and you can find the check-boxes in the upper right corner of the panel) is always set to the correct color. In case both colors are used, you have to choose the lower color (color 1 in this case). 

This information is not valid any more. The phase shifter is back in operation.



The phase shifter of SASE2 cell 17 is currently not working.

As you know, cell 17 is upstream the second chicane and important for proper pulse intensities. In other words, the unavailability of this phase shifter is currently preventing reasonable operation of the beamline.

The undualtor group is informed and will repair the phase shifter as soon as possible.

Well, it was me...

We have several deviating energy measurements in the I1 section and we have always wondered which of the measurements is correct. In fact, the problem is not solved yet but we made a decision during setup that I want to tell you.

The currently correct electron beam energy measurement in the I1 dogleg is 129 MeV.

This might be counterintuitive because now all measurements are below 130 MeV but there are indications that this might be just fine. The good performance of SASE1 @ 7 keV supports this assumption (The performance of SASE2 however...)

We will keep you updated if we change the default energy again.

Hi all, last Monday, al XFEL tunnels were closed again, and since Tuesday afternoon, the interlock light tests have also been completed and we are allowed to restart XFEL. 

As usual, you could find the restart plan in the logbook following the link "Schedule" or you can klick here

First electrons have been accelerated already in the injector and we hope that we can achieve transmission to TLD, T4D and T5D during the coming weekend. 

Next week (CW3) will be used for final maintenance and repair work (as far as really necessary) for some further tests, beam based alignment of the quadrupole magnets in the undulator beamlines and, finally, SASE tuning. 

Many thanks in advance to all of you for the efforts in the coming days and weeks!

Many thanks to all colleagues who helped to run the European XFEL so successfully this year! We are already looking forward to getting started again at the beginning of January 2023 (smile)


We wish you all a Merry Christmas and a Happy New Year! 



(more about fractal trees and snowflakes can be found here (smile) ) 


On special request of the hard X-ray self-seeding team, the photon beam pointing server can now perform adjustments on smaller portions of the undulator. We do not recommend this for normal operation because it tends to introduce "kinks" into the orbit and its reference. However, if you have no other choice, you can enable "cell selection" from the Expert tab and select the cell numbers on which the correction should be done.

This is a major change to the server, so please test it thoroughly, for example on the Virtual XFEL.

Do you remember all the steps for switching off the SASE2 chicanes correctly? You no longer need to. There are now buttons for two Taskomat sequences on the chicane control panel that will take the necessary steps. The sequences take care of disabling write protection, switching off trim coils, degaussing the main magnets, and so on. Please try them out and let me know if there is any problem.

Gun Ramp up after a trip

After a gun trip we experience that the following ramp up fails early on in the ramp three times after which the FSM gives up.

A few minutes later the next ramp then is successful - so it seems that the gun needs a bit more time to recover. Due to this I encreased the time for a rest of the interlocks in the FSM to 2 minutes.

Please wait for the FSM to do the reset after 2 minutes and do not reset by hand earlier.


In former times a reset by hand was necessary, but this has changed now.

Hi all, we have a special setup this week (calendar week 39, 26.9.-3.10.) and I would like to take this way to inform all operators, especially those who are on shift this week. 

SASE2 will get all bunches every of 100th bunch train (in other words, every 10 seconds), SASE1 and SASE3 go empty on this shot. 

For that reason, we need the same RF settings from 800-1400 us (the full RF pulse). If that is not the case, only a fraction of the full bunch train will lase properly in SASE2. 

We have setup I1, L1 and L3 with only one Beam Region and we would like to ask you to keep that until next Monday. 

All SASE tuning panels are currently using Beam Region 1 as default. 

Delivery week schedule


Probably most of you already know it, but I wanted to point out a table in the schedule where you can find all details about the desired photon energies, pulse lengths etc. during the user runs. 

Please review the table relevant to your shift and get familiar with the requested setups. If you have any questions, please feel free to contact me or the RCs. 

You can use the 'Schedule'-Link in the logbook to find the parameter table. The author is typically XO (X-ray operations) und der title of the entry looks like 'X-Ray Delivery table cw39/2022'. As you can imagine, cw39 stands here for calendar week 39. An example of such a table is shown below. 

There is a second table available for each user delivery week. Also these tables are available via the logbook and the entry titles look like: 'Operator Table (shift-by-shift) cw39/2022'. As the title suggests, this table is a bit more elaborated with more details for each shift. 


On the left you can see the an exemplary table for the delivery week 39. 

You can find there the involved experiments of the week (typically SASE2 and SASE3 instruments do not change, while the SASE1 instruments changes between day and night shifts). 

There is also the name and phone number of a contact person listet. 

Also important for operation is the information on requested photon energies and number of pulses at a given repetition rates. 

Last but not least you can check whether the experiments will be carried out in interleaved mode plus you get some additional comments in the last column. 

These tables are developed and checked for consistency in the weeks leading up to their validity in collaboration between the instrument-scientist, users, XO and MXL. 

All setups in these tables should be prepared at the beginning of the week. This applies especially to the different photon energies, but also e.g. to special pulse patterns.

Of course it can happen that one of the setups listet in the tables could not be prepared before. In such a case, the setup has to be prepared at the moment when the colleagues from the instrument ask of it for the first time. A detailed description on how to prepare a new, not prepared photon energy setup is described in the corresponding procedure

In case you get a request that is not listed in the table, please contact the RC in order to get information on how to proceed. 





Concerning the puls length I would like to add the following: We often setup the bunch pattern on Mondays with a reduced number of bunches on the last beam region (typically SASE1/3 in interleaved mode) and run the accelerator without the dump cleanup in order to have a constant number of bunches in the accelerator. 

This setting is then often present until one of the instruments asks for more bunches than possible at this time. In such a case, please adjust the number of bunches on the last beam region in the buch patter builder accordingly. The limit that you have to consider is the total pulse length of the gun. 

In case you are not sure how to proceed, please contact the RCs of you week. 


Hi all, please note that only one feedback loop should be closed for linked beam reactions. Otherwise both loops would write to the same parameters.


In the picture on the right side you can see an example. The beam reagions 1 and 2 are linked in the injector and therefore the feedbacks for the injector can/must stay off for SASE1 and SASE3 (i.e. beam reagion 2).


Currently this setup has to be considered by the operators. In the foreseeable future the correct setting will be taken over by the feedback server and you don't have to care about it anymore.

In case you were not able to attend the Beam Regions operation training on Tuesday, or if you want to have another look at the slides, you can find them here now: → Operator training Beam Regions


Each LLRF station can now be operated either in table mode or in scalar mode. Here's the difference between the two:

  • In scalar mode, the station has just a single setpoint for voltage and phase. It produces a single, flat output curve. This is somewhat similar to what we had before the shutdown, except there are no more flat-tops or slopes. Just one amplitude and one phase, that's it.
  • In table mode, the station accepts full set-point tables from other servers or tools. This mode is much more flexible than everything we had before, but it requires an external piece of software: The voltage and phase controls on the LLRF panel cannot be used anymore and are grayed out.

The LLRF panel has a few new controls to switch between the modes in the top left corner (see image).

Where and when do we use scalar mode?

For now, scalar mode is the only mode used for the gun and for all L3 stations. Scalar mode is also used by the FSM for ramping stations up and down. Finally, LLRF experts use it when working on the station.

Where and when do we use table mode?

Table mode is used by default for the injector (A1+AH1), L1 (A2), and L2 (A3–A5). For these stations, the tables are written by the sum voltage control server. More on this in an upcoming post.

Beam regions are coming

With the restart of the XFEL, beam regions are replacing the trusty old RF flat-tops. Very little will change for the initial operation of the machine. But later on, this will open up many more capabilities – in a nutshell:


Flat-TopsBeam Regions
Maximum number of regions3≫10
Table shapeConstant + slope

Free, limited only by klystrons, couplers & cavities

The image on the right shows the LLRF sum voltage control panel on the Virtual XFEL running three beam regions with funny tables.

Keep watching this space for more posts on how to operate the new LLRF and sum voltage panels!

Here comes a short post to inform you how to act in case of a "Safety locked" signal appears in the MPS. 

Actually, that is easy: Call the PRC. They have a procedure how to continue.