1. Studies Leader Report: 9mA webex meeting, 15th March 2011
John Carwardine, Nick Walker

2. Some key analyses needed of the Feb study data
Parallel session on 9mA studies at ALCPG
Long bunch-trains Workshop
Proposed Journal article submission

3. Analysis of studies data

4. Top-level analyses: key information to be extracted
Important top-level analyses
Beam current scans
RF flattop scans
3-hr run
Name of person doing the analysis
Analyze achieved cavity-by-cavity gradient flatness and gradient tilts
Evaluate gradient operating margins
Pk/Ql optimization: bounding sources of errors at the nominal matched conditions
Analysis of detuning measurement algorithms
Impact of detuning errors on usable gradients and on gradient flatness
16-cavity model accuracy
Future 9mA studies: extrapolation to achieving match at 6mA

5. Examples of important detail-level analyses
Examples of detail-level analyses
Beam current scans
RF flattop scans
3-hr run
Name of person doing the analysis
Lorentz-force detuning compensation: piezo impulse response for future LMS optimization
Stability of forward power ratios
Cavity to cavity rf power crosstalk
Correlation of detuning with sense piezo signals
Piezo optimization of static & dynamic detuning and detuning curvature over flat-top
Analyze of Lorentz-force detuning vs gradient
In any given measurement, can we separate errors in detuning from errors in desired Qls?

6. Analysis of the beam current scans of gradient tilts
Beam Current (mA) 1 2 3 4 5
Gradient change over 400us (%) -5 +5
+10
Gradient Tilts vs Beam Current (ACC7)
-10
-15
Intended working point
Plots of tilt vs beam current shown so far use measured tilts from a single pulse at each current
We need a more complete analysis that use all pulses in the scans to get more precise fits and to attach error bars
Compute cavity-by-cavity tilts vs beam current for all pulses in the scans
Make linear fit to the tilts vs current to get per-cavity measures of:
Slope of the tilts vs current
Beam current that gives zero tilt

7. Evaluate gradient margins (usable gradient)
For each cavity, take the maximum gradient over the flat top and add the maximums together for all the cavities to get a maximum available vector sum.
The fraction of usable gradient would the ratio of actual vector sum and the sum of the maximum gradients
Factor in pulse-to-pulse jitter & drift to get an assessment of the needed gradient margin

8. Bounding sources of errors from beam current scans
Beam Current (mA) 1 2 3 4 5
Gradient change over 400us (%) -5 +5
+10
Gradient Tilts vs Beam Current (ACC7)
-10
-15
Intended working point
In reality, there were discrepencies in the currents at which individual cavity gradients had zero tilt
Matched at 3mA
4.5mA
1.8mA
As the beam current was scanned, there was a current for each cavity at which the gradient tilt was zero The goal was for all cavities to go through the zero tilt point at exactly the same (desired) beam current
Using data from the scans, attempt to quantify or bound the sources of these discrepancies, eg Loaded Q, detuning, Pfwd, beam current,…

9. What else can we learn from the results of the beam current scans?
Slopes of the relationship between gradient tilts and beam current
A linear relationship is expected analytically.
The slope is a function of the loaded Q
Calibration errors in the beam current measurement would show up self-consistently in the slopes of all the cavities
What role does detuning play?
Self-consistency check of measured parameters
Back-calculate forward power from gradient, loaded Q, beam current scans. Compare the result with the measured Pfwd.
Repeat this exercise by computing each parameter in turn from all others and comparing results with measurements

10. Analyze and compare detuning measurements algorithms, attempt to bound the errors
The detuning was computed using two methods:
Online real-time computation using Pfwd & Gradient and the cavity equations
From rf decay during scans of the rf flat-top duration
There are some apparent discrepancies between the two methods for the same rf pulses, notably detuning offsets (static detuning?)
Both methods in principle should give the same results, but both are susceptible to systematic errors
How do we make sense of this in terms of controlling ‘absolute’ detuning to the ~10Hz level using the piezos

11. Impact of detuning errors on effective usable gradient
Forward power is fixed by the power distribution system, but the fraction of Pfwd that goes into the cavity depends on the detuning
Our model for maximizing the usable cavity gradients relies on all cavities reaching their maximum gradients together at the end of the fill
But relative differences in detuning from cavity to cavity will impact the relative cavity voltages at the end of the fill, spoiling relative maximum available vector sum
There is potential for errors of 10’s of hertz in the absolute measurements of detuning (enough to impact the gradients by several percent)
Analyze…

12. Analysis of monitoring piezo ‘vibration’ signals
How do the signals on the monitoring piezos correlate with detuning during the rf pulse?
How do they compare from cavity to cavity?
How do the mechanical resonants as picked up by the monitoring piezos vary from cavity to cavity? Does this tell us anything?
How do the amplitudes of the signals compare from cavity to cavity? Do they correlate with relative forward rf power?
Can we make use of the monitoring piezo signals for detuning and/or vibration control?

13. Power distribution power ratios
Compute the forward power ratios, compare with the power ratios based on the measured attenuator values in Katalev’s spreadsheet
Look at how these vary over time (drift) and as a function of gradient, detuning, beam current
Couple this with analysis of cavity-to-cavity rf power crosstalk..?
Power ratio (dB)
Time of coupler scans
Loaded Qs changed here
…and here

14. Parallel session at ALCPG

15. Combined meeting of the Physics/Detector and GDE communities
ALCPG – Linear Collider Workshop of the Americas in Eugene, Oregon, Sat-Wed, March 19th-23rd
Combined meeting of the Physics/Detector and GDE communities
Joint PD/GDE plenary
PD and GDE individual plenary
Parallel Working Group sessions
GDE parallel sessions: Working Groups
Main Linac / SCRF, Damping Ring, Conventional Facilities, BDS/MDI, Sources, Beam Dynamics

16. SCRF WG topics (see ALCPG website for details)
SCRF Working Group
SCRF WG topics (see ALCPG website for details)
Cavities, HLRF, 1TeV upgrade, Industrialization, Cryomodule Test, FLASH 9mA studies
Session on FLASH 9mA studies:
Tuesday March 22nd, 13:30-15:30 Pacific Time
Session will be Webex’d
Agenda
Summary of the February 9mA studies (JC)
Detuning and detuning compensation (Mariusz via webex)
Pk/Ql gradient flattening studies (Julien via webex)
Outlook on future studies (Brian via webex, JC)

17. Summary talk (JC)
Broad outline of the studies goals and accomplishments
Some results
Open questions, observations
Analyses needed (basically what I covered earlier)
Provide a lead-in to the talks by Mariusz and Julien

18. Long bunch-trains Workshop

19. Long bunch-trains workshop: Expect a more detailed announcement soon
Nominally 2-1/2 days, June 6th to 8th at DESY
Monday afternoon to Wednesday afternoon would be the best compromise for remote participation
This time, no parallel sessions (just plenary), but we still want it to be a workshop and not a mini conference
How to best make that happen?
Comments / suggestions on organization and topics would be welcomed

20. Themes
The primary driver for the workshops has been the 9mA experiment
Last year, a dominant theme was the difficult machine operations during the Sept 2009 studies
This time, our February 9mA shifts yielded a lot of valuable experimental data, so this should bode well for good material at the workshop (but not enough for a full workshop)
FEL operation with long bunch trains has had a lot of attention at FLASH in the past year
There is much common ground with the 9mA experiment

21. Some suggestions for session topics (Five half-days means ten ~90 minute sessions)
Topics specific to the 9mA experiment
Pk/Ql (gradient flatness studies with beam loading)
Strategies for reaching high beam loading and maximum gradients without quenching
Future 9mA studies
Module tests at Fermilab (NML) and KEK (S1-Global)
Topics generally of interest to LBT ops and studies
LLRF Control with long bunch trains
Exception Handling
Detuning control / Lorentz-force detuning compensation
DAQ and analysis tools
Topics focused on FEL user ops ?

22. Proposed journal article submission