Minutes of the ABP-LCE team meeting on 26.09.03
present: GA, EB, OB, HB, TdA, AK, EM, FR, DS, EV, LV, FZ
excused: WH
------------------------------------------------------------
(1) Actions and Comments on Minutes
-----------------------------------
ACTION -> EM will further check the space charge detuning.
Not yet done due to PS MD.
ACTION: FR will suggest a possible LCE team sharing of the workload.
Done, see point (2) below.
(2) Future Work on Zbase (FR)
-----------------------------
The proposed sharing of the workload is as follows:
EM+LV: Resistive wall impedance of beam-screen (cold+warm),
collimators, TDI absorbers, MQW, MBW, and septa.
DS: Geometric impedance of collimators
(and interface of ZBASE with GDFIDL?).
Bellows and Interconnects.
FZ: Resonator impedance (due to HOM's of RF-cavities and
trapped modes in experimental chambers): narrow-band and
broad-band.
Kickers, BPM's, and cold-warm transitions.
Quadrupolar impedance giving rise to s-dependent incoherent
tune shifts (for a few selected elements such as collimators
and maybe later e-cloud?)
LV: Help and advice on all impedance estimates
TdA: Help and advice on all computing aspects (especially ZBASE)
FR: All remaining items
(e.g. pumping slots, high-frequency resistive impedance of the
beam screen, ... and consistent merging of all data in ZBASE)
The first stage will consist in collecting all the existing data and
estimating the amount of work required to complete it (e.g. if analytic
or numeric estimates need to be done or repeated). In 2 weeks time, on
the basis of the first results, a re-shuffling of the workload may be
decided in case some items turn out to be too easy or too difficult. We
may ask for help by external collaborators (e.g. Bruno Spataro), fellows
and students. But the main responsibility should be with a CERN staff.
ACTION -> Collect all impedance informations within 2 weeks time
(EM,LV,DS,FZ,TdA,FR)
FR will also look at the tensor transformation of the impedance
required for 45-degree collimators.
ACTION -> Tensor Transformation (FR)
Regularly, the LCE teams receives excel spreadsheets
from RA. FR suggests to write an automatic processing
routine, which can quickly re-compute the impedance for
each parameter update.
ACTION -> Automatic Excel spreadshit for collimator
impedance (EM+LV)
OB explains that different users can work simultaneously
by accessing the same afs directory. Old versions of
impedance estimates should be kept to allow tracking
the history. TdA will create a new machine version
called LHC September 2003.
OB points out that Zbase can compute the impedance from
resonator parameters. It can also convert impedances into
wake fields and vice versa. Bunch length information is
needed for this type of transformation. At the moment
impedances can be either broadband or narrowband. Double
counting should be avoided, if a broadband part is added.
We may need to add additional elements if we want to
represent other frequency-parts of the impedance in
another way. Numerical evaluation routines are not (yet)
available for an arbitrary analytical expression.
(3) Coherent vs. Incoherent Tune Shift in Non-Collimation Plane
---------------------------------------------------------------
LV and FR discussed this question. According to LV, one can
distinguish three types of tune shifts: incoherent, coherent,
and collective. EM illustrated the stability diagram with and
without incoherent tune shifts. The result is that the incoherent
tune shift can be ignored. FR concluded that the stability limit
including collimator-impedance in the LHC design report is correct.
(4) Possible Collimator Test in the SPS Next Year
-------------------------------------------------
LCE team should contribute to SPS collimator test planned
by RA next year. 2 collimator jaws should be installed in
the ring (near QD519) and 1 in a transfer line. Impedance,
heating, trapped modes, response matrix are issues to be
looked at. JW will take care of the orbit stability.
LV suggests to stabilize the beam with octupoles and then
make it unstable by closing the collimators or lowering the
octupole strength.
FZ recommends to install collimators in the vertical
plane rather than horizontally, and at a plane with
large dispersion. This way one can also measure the
energy loss (longitudinal impedance), by closing the
collimator jaws symmetrically around the beam.
Moreover, the beam is smaller vertically, and, therefore,
the wake could be made larger.
OB and others agreed that both jaws should be uncoated
to maximize the effect. For the same reason the
measurement should be done at maximum energy since the
wake field effect scales as gamma^(1/2) for a constant
distance in sigma.
The team does not object to the possible installation of
additional coated jaws. These would also permit the study
of damage diagnostics.
Temperature of the jaws will be monitored.
Electron-cloud probes could be installed downstream
of the collimators.
LCE team members should physically attend the MD.
LV mentions that the Schottky monitor should provide
a signal with the best resolution.
There is also a 6 GHz cavity by FC and maybe a PLL by BDI.
(5) Evaluation of the Kicker-Impedance Measurement (GA, LV)
-----------------------------------------------------------
The idea is to make a local bump at the position of the
kicker and look for bump non-closure at high intensity.
An earlier estimate for a single kicker module indicated
that the effect was marginal, but taking into account the
effect from 5 kicker units and a resonable bump amplitude,
the net kick is about 10 microrad, which can be detected.
(6) Report from SNS ASAC Review (FZ)
------------------------------------
Most interesting beam-dynamics studies were related to
the interference of adjacent magnets (could this be an
effect in the LHC?), self-consistent stationary distributions
with space charge (generalizations of KV distribution to
3 dimensions), electon-cloud simulations at BNL and ORNL
(including scanning of clearing voltage and implementation
into ORBIT for fully self-consistent simulations), and laser
stripping as a robust alternative to foil stripping.
(7) Matlab Simulations of Multi-Bunch Feedback (EV)
---------------------------------------------------
EV explained the concept of a 1-turn tune measurement
using two BPMs. He can simulate the measurement including
the effect of feedback and noise using MATLAB SIMULINK.
The circuit employed corrects for a possible phase error
at the location of the kicker, allowing to locate the
kicker at any position. The single-bunch model
is described by a state space representation. This can
be extended to multiple coupled bunches. This is done
by diagonalizing the coupled problem, transforming
back to the single-oscillator form, and later re-
converting the solution to the original coordinates.
Movies were shown for 1 bunch and 4 bunches.
EV demonstrated that the single-bunch impedance can
also be added quite easily.
FR proposes a presentation to a wider audience.
Future plans include to use the FFT information instead
of the direct pick-up data, expecting that the noise
effect can be reduced, when only unstable modes are damped.
It was suggested by FR and GA to take account of the
finite bandwidth, e.g., by adding a frequency-dependent
filter.
This scheme could be compared with the B factories
(e.g., work by J. Fox et al.).
Aims of the study are to improve the noise behavior
and to measure 1 mode vs. time, while damping all the
other modes, for diagnostics purposes.
AENDERUNG EV):
--------------
>EV explained the concept of a 1-turn tune measurement
>using two BPMs. He can simulate the measurement including
>the effect of feedback and noise using MATLAB SIMULINK.
>The circuit employed corrects for a possible phase error
>at the location of the kicker, allowing to locate the
>kicker at any position. The single-bunch model
>is described by a state space representation. This can
>be extended to multiple coupled bunches. This is done
>by diagonalizing the coupled problem, transforming
>back to the single-oscillator form, and later re-
>converting the solution to the original coordinates.
>Movies were shown for 1 bunch and 4 bunches.
>EV demonstrated that the single-bunch impedance can
>also be added quite easily.
EV explained the concept of a 1-turn tune measurement
using two BPMs. He can simulate the measurement including
the effect of feedback and noise using MATLAB SIMULINK.
The circuit employed corrects for a possible phase error
at the location of the kicker, allowing to locate the
kicker at any position.
A single-bunch model can be described in the discrete state
space representation. This can be extended to multiple coupled
bunches. This is done by diagonalizing the coupled problem and
transforming the input- and output variables back to the single-
oscillator form. Movies were shown for 1 bunch, 4 bunches and
16 bunches. At the 16 bunch model EV demonstrated a one turn
modal analysis and that a narrow-band impedance can also be
added quite easily.
>Future plans include to use the FFT information instead
>of the direct pick-up data, expecting that the noise
>effect can be reduced, when only unstable modes are damped.
Future plans include to use the fast modal analysis (done
with a FFT) information instead of the direct pick-up data,
expecting that the noise effect can be reduced, when only
unstable modes are damped.
>EV's scheme could be compared with the feedback at the
>B factories (e.g., work by J. Fox et al.).
Das was ich an Veröffentlichungen von den B-Factories gesehen
habe sieht eher nach den klassischen Konzepten von Multibunch-
Feedbacks aus und hat wenig mit dem von mir verfolgten Konzept
zu tun. Neu und fortschrittlich ist dort die Diagnose unter
Verwendung der digitalen Datenaufbereitung der Feedbacks.
Ich würde den Ansatz komplett aus den Minutes streichen!
>Aims of the study are to improve the noise behavior
>and to measure 1 mode vs. time, while damping all the
>other modes, for diagnostics purposes.
Aims of the study whether it is possible to improve the noise
behavior and/or to examine whether it is possible to measure
1 mode vs. time, while damping all the other modes, for
diagnostics purposes.