Blog

Due to the current Covid situation and the associated restrictions, we have made it easier to access EuXFEL remotely. The operation restart next week as well as some studies during the next weeks will be carried out respectively will be supported by the run coordinators and system experts from their offices or from home.

We have developed a new procedure to explain in detail how to enable remote access and what has to be taken care off.

The procedure can be found on the XFEL Operation Confluence page as usual.

Please let me know if you have questions or if there is something missing in the procedure.



The XFEL restart next week is getting closer and for this reason there will be some blog posts with news around the XFEL this week.

Let's start with the new computer hardware that that was installed in the control room.

The first console on the right hand side of the EuXFEL desk is now a new Mac Pro. Please use this computer and compare it to the Linux machine, which you find on the exact opposite side of the table.

I would be very happy if you could send me some pros and cons about both systems in the next 2-3 months. We will then incorporate this information into the decision for the future hardware.  

Ad-hoc magnet groups

The magnet middle layer server has a new functionality. We can now add ad-hoc groups with a few keystrokes and a mouse click. This can be very handy for setting all magnets in a certain section to zero current:

  • Type C.*SA3 as a regular expression (this means any name that starts with C and ends with SA3)
  • Click the "→ Ad-Hoc Group" button
  • Press "Set Current to Zero"

Regular expressions can be tricky. See https://regexone.com/ for a tutorial. Here are some of the most useful constructs:

.Any character
.*Zero or more occurrences of any character
.+One or more occurrences of any character
\.An actual dot (it has to be escaped with a backslash)
[XY]The letter X or the letter Y
[0123456789]Any digit

Camera servers

Please switch off the camera servers after your measurements or investigations. This can be done with the 'server off' button on the top left corner of the camera expert panels (the position is shown in the left picture). 

In case you want to switch off more than one camera server (or data acquisition and/or light) it is convenient to use the XFEL camera status panel. That ca be found on the main JDDD panel → Diagnostics → Camera status (in the bottom left corner of the panel). A picture of the camera status panel is also shown on the right hand side. 

 

The console on the far left in the BKR is now for test purposes a Linux machine. It runs of course JDDD as well as all other programs needed for  XFEL operations. 

Please consider using this computer during your shifts and give us feedback if you liked it or if you preferred to work on the Apple computers. 

I am also collecting information about which console still has problems and which one is running perfectly fine. The goal is to get all computers in perfect conditions for the next runs. 

SASE2 has been in operation with largely different photon energies this week ranging from 8.9 keV to 18 keV. 

Please consider that the SASE2 XGM's ADC amplifier has to be adapted in each case in order to get the correct readings. The lower photon energies need a preamp setting of 17 dB while the high photon energies (17.8 keV and 18 keV) have to be measured with a preamp setting of 37 dB! 

The picture on the right had side show you how to find the correct panel and which buttons have to be pressed. 

Please find Theo's lates operator training slices for the XGMs here

In most cases trips of RF stations are reset by the FSMs and hardly leave time to make a logbook entry. So far so good. 

But sometime the problem is more severe and needs attention by the shift crew. In these cases please make sure that you evaluate the error messages carefully, call the right expert on call (especially during nights) and attribute the downtime to the right group. 

RF station downtimes can essentially be caused by problems with: LLRF, high power RF (MHFp), couplers (MSL), vacuum (MVS), cryo (MKS) and eventually magnets (quads off → RF off, MKK). Especially the problems concerning the first three categories can possibly be assigned incorrectly. 

In case of doubts please ask your run coordinator or discuss with the other colleagues being on shift. 

Logbook entries

Dear XFEL operators, 

first of all, thanks for the fact you almost always write shift summaries in the logbook! That is great and helps us a lot to get a quick overview of what happened during you shifts. 

Please keep also in mind that it is important to provide the logbook entries with your name, a title and a severity. 

Name and title are especially important when you send the logbook entry to an expert. The title will be the subject of the mail and your name is important in case the expert has questions to the author of the entry. 

Please feel free to add also the keyword and location to you entry. Remember that it is always possible to edit the entries afterwards and to add all required information. 

In many cases it makes also sense to describe e.g. the screenshot with a few words in the text field. Do not hesitate to write the text in German if you prefer that. 

The L-IBFB reduces the arrival time jitter within a bunch train. Like the transverse IBFB it needs several bunches at the start of a bunch train to measure and to apply the correction. It acts on the amplitude of the RF-station A5. The feedback does currently only run on the second flattop!

Please consider that the run coordinator has to be informed before you switch the L-IBFB on.

There is now a procedure available that explains how to switch the L-IBFB safely ON and OFF again. 

You can find the procedure at the same place as all other procedures in Confluence or by clicking on this link: L-IBFB procedure.





KL Kicker Adaption

When running the kicker adaption tool please apply 'Reset All to Zero' beforehand. The other selections only apply to the bunches that are currently in the machine and lead to unexpected behaviour when the number of bunches is changed frequently.



Phase shifter scans

Hi all,

as it is well known, phase shifter scans can help to increase the pulse energies drastically. However, sometimes scanning the phase shifters one by one and applying the results from the scans leads to strange gap offsets and, in the whole, eventually to less pulse energy. 

For that reason, please check the results of the single scans before you apply them to the undulator! Maybe the following rules can help to increase the success rate of these scans. 

  • You can find two values describing the phase shifter setup: The gap offset as well as the corresponding period. Both are shown in the upper picture, marked with the blue rectangle. The individual period is shown in the lower field. It describes the phase shifting in units of 2 Pi. This number must be close to an integer (in best case to the global phase shifter period!) and should never be something like x,5! If that is the case, set the offset back to zero and start again! 
  • Please consider that the first and the last closed undulator cell's phase shifter may not increase the pulse intensity during a scan. Thus, when you see only little amplitude changes during the scan, do not apply the result.
  • You can save a lot of time by only scanning the important cells. Those can be identified by changing the phase shifter gaps in a few cells in the center of the undulator such that the period changed by about 0.1 (times 2 Pi). This happens typically with gap offsets of 0.5 mm or 1 mm. Check if the phase shifter's change has an impact on the pulse energy. If yes, include it to your later scan. Go ahead and test another cell quickly and make the same decision. 
  • If you are not sure at the end if your scans helped, at all save a sequencer file, set manually all phase shifters back to zero, compare the SASE signals before an after. It was better with the modifications? Reload the phase shifter offsets form the sequencer file. 

BTW: The phase shifter setting shown in the picture on the right hand side was found lately with scans. SASE1 delivered around 700 uJ with this setup. We sett all offsets to zero and ended up with almost 2 mJ! 

One more thing: If the period of at least one of the phase shifters is ±∞, check you global phase shifter period setting (next to the SET- Button). It is most likely set to a too small period! 



We have a new procedure on reset options for the dump vacuum control available.

Please follow this procedure when you have problems with the dump pumps in the future.

In case you have questions concerning this or other procedures, do not hesitate to contact me!

The Interleaved Pattern Wizard on the bunch pattern server has received an update – it now analyzes the old pattern and automatically tells you the right frequencies and dividers. No more need to remember that SASE1 wanted 282 kHz while SASE3 is frantically switching between 1128, 564, and 376 kHz.

This feature was actually requested by at least three people independently, so thanks to everyone for the input!

Accelerator Snap Tool

The accelerator snap tool is available again!

It can be opened from the Main JDDD Panel → Status → Machine Status → Accelerator Snap 

You can take a snap of the current machine setup and you wind then find in the list below deviations of BPM offsets, changes in magnet strengths or beam energy etc compared to your snapped setup. 

My suggestion is that you press the 'SNAP' button in the top left corner when you are happy with the machine status. In case the machine machine performance becomes worse, check the lists for large deviations. Maybe you can find a single source of your problem. 

The configurations of the single beamline sections should be ok. But you can also change it in case that is necessary. It would be great if you could inform me in such a case. 

Forgotten Tools

Do you remember when we always saw the strong correlation between the gun phase and lasing in SASE1 and SASE2? 

The Correlation plot tool was a big help during that time to optimize the signals and it was used a lot. We have now a much more stable gun phase and this is why the plot does not look as impressive as before any more. And with that, the tool was not used regularly any more. But, it is still useful and can help us improving pulse intensity! 

We halve also the correlation plot server available! Both, the server as well as the Python tool are accessible via the Main Panel → Tools → Data Analysis. 

Are there other tools that we used month and years ago that we do not use any more? Could you give me a hint what we should use again more often? Thanks in advance!