Choose timezone
Your profile timezone:
Weekly meeting on Xboxes
→
Europe/Zurich
18/3-008 - CLIC Meeting room (CERN)
Description
https://cern.zoom.us/j/6910405761
Weekly meeting on Xboxes
Zoom Meeting ID
64847033219
Host
Nuria Catalan Lasheras
Alternative host
Marçà Boronat
Useful links
Join via phone
Zoom URL
Attending: Paz Alonso Arias (PAA), Nuria Catalan Las Heras (NCLH), Mareike Wendelmuth (MW), Matteo Volpi (MV), Igor Syratchev (IS), Anisullah Baig (AB), Ping Wang (PW), Lee Millar (LM), Olivier Brunner (OB)
News from Australia. MV comments last news from MEL-BOX. Conditioning of RF window and load at same time, but reflection power cannot be measured for window, just for load.
MV shows upE curves for their Klystron. NClH asks how are voltage and current measured. MV reports a shift between the values showed in the monitors and the ones measured with the clam. Voltage is measured from monitors. Impedance of heater element is constant. MV reports lower value of uPE than that reported by Canon. IS comments on how uPE is geometric parameter, does not change, and how we are not trusting voltage values from modulator, and actually assuming uPE is fixed value (1.55 in CERN HEK). PAA is measuring beam current through oscilloscope and calculating beam voltage from it taking this assumption.
CERN Xboxes news. Xbox2. Last week all lines were finished calibration. However, the modulator could not be restarted, as it showed interlocks of the switch. At some point, the interlock disappeared, and the Xbox could be restarted.
It was restarted with 20MW (all going to structure B) on Monday (21.08.23). It was running over night, but interlocked during the night on switch 6. Sometimes during the day it was giving interlocks on switch 4. The needed spare switches were sent for repair and hopefully back within a month. The switches from Xbox 1 were checked and cannot be used. The interlock was sent to Tom and their team recommended to look at the potentiometer.
The conditioning is running, with stopping during the night due to the interlocks.The question was mentioned to reset the interlock and restart automatically, the risks of automatic restarts are discussed. Hopefully, soon people are back from holidays and the problem can be fixed. NClH decides against an automatic restart.
PAA questions regarding the dip in the klystron power: Is due to the phase change during operation to get the flat top.
CERN Xboxes news. Xbox3 HEK. IS suspected that the CCSC reported by Canon was wrong and instead of 7A, 11A. This hypothesis was confirmed. PAA has performed measurements with a new setup and making a fine tuning of CCSC, to get the best diameter for the beam for high efficiency.
To finalise the measurement, one of the HP directional Coupler needs to be calibrated. For this, vacuum needs to be broken and the coupler uninstalled and reinstalled.
Question by IS: Why are the error bands that large for the measurements? It’s due to the error in the powermeter measurements. The error is not attenuated in line, but it’s just a sum. This will be investigated later again. To improve, it can be measured in MW in the powermeter. For this the exact attenuation is needed to be put in the powermeter (coupler measurement needed).
Calibration of the VNA: The measurement of the calibration of the VNA is obviously wrong. NClH, IS ask if maybe the calibration kit is wrong? -> WW: Should not be the case, seems like the VNA is wrong.
The change in 5MHz results in a too high change in output power. The calibration looks like there is a reflection. The jitter in S21 should not be there. PAA confirms that redoing the measurement with more points gave the same results.
Proposed solution: Check the calibration kit with another VNA to check the kit; Use another VNA for the measurements; Use a Coaxial calibration kit to recheck.
WW comments that back in the days you had to put 1 Ohm instead of 50 Ohm for the calibration at waveguides. MV confirms that this was only for saving the file, not for doing the calibration. This was still done with 50 Ohm.
PAA comments that she also checked if there was interception, no interception was found.
Plan was to go back to Tube 1 measurements. But there is an interlock on Modulator A. This stops operation, as the Software interlocks all triggering. To change in Labview gave more interlocks and more errors. One workaround will be to use a signal generator for operation instead of LabView.
MV suggests to put a jumper behind the interlock crate to remove the overage the interlocks. Also to some changes in the Software. He is doing that right now to operate. More adjustments are needed. MV will give the details to PAA and NClH.
Powerheads seem to be incorrect calibrated: Different powerhead give different results for the measurements. Calibration with the signal generator gave no difference. This means that the difference is due to non-linearities. The dynamic range changes and would need to beknown. There should be a calibration table on the powerhead -> PAA to check this.
Pulse by pulse analysis. MW comments on BD investigation of individual pulses. WW says to look pulse before and after with phase. The phase of previous pulse to be sure of transmitted phase and be confident of reflected phase. Log of reflected signal will give us accurate positioning of beginning and end of structure.
There are minutes attached to this event.
Show them.
