Fast feedback for linear colliders Page: 3 of 6
This article is part of the collection entitled: Office of Scientific & Technical Information Technical Reports and was provided to UNT Digital Library by the UNT Libraries Government Documents Department.
Extracted Text
The following text was automatically extracted from the image on this page using optical character recognition software:
optimization packages, the actuator is scanned through a
range of values and a parabolic fit is performed; but this is
an invasive procedure. An alternate method implemented
in the feedback system is called "dithering" (51. This
involves perturbing the actuator by a tiny amount above
and below its nominal setting while taking synchronous
beam measurements. After many pulses. an average slope
is calculated for the measurement versus the actuator
change. The slope of the parabola is linear with the
actuator, so the linear feedback calculation is formulated
with the calculated slope as a measurement. Minimization
is accomplished by keeping the slope set at zero, bu the
feedback system is generalized so that the slope can be kept
to any requested setpoint. In the SLC. generalized
capability for ditherirg and minimization has been
developed and demonstrated. The system has been able to
perform dithering and control the heam to a stable point on
the parabola. Unfortunately, use of the dithering system has
proven to he rvasive; the smallest dithering bit sizes cause
unacceptable beam disturhances.
Hoeur. a noninvasive minimization feedback has
been implemented and commissioned. The kicker timing
feedback loop reles on the natural filter of the ring
extraction kick time (measured with a TDC) to produce
slope calculations of BPM measurements versus kick time;
dithering is not needed. In order to provide a reasonable
slope estimate. 10.000 pulses of data are averaged; at a
BPM measurement rate of 60 Herl, it takes almost 3
minutes to produce a single calculation. Commissioned
only a few days before the end of the last SLC run. the soop
has already been shown to improve machine performance
156
0 0 -
h-15
4
v 0
~ -
-4 ' . ,.
0.20
0.12k
0 4 a 12 16 20
Time (hours)
FIGURE 2 Minmiation kicker timing feedback improves
the RhS beam jitterFigure 2 shows the feedback response to changing ring
extraction conditions; the RMS beam jitter (calculated by
a downstream feedback loop) increases until the feedback
corrects the kicker timing, which reduces the jitter to its
normal value. Note that the feedback response is slow due
to an intentionally conservative initial design.
PERFORMANCE ISSUES
In the past year, progress has been made in identify mg
and analyzing SLC feedback performance issues[4] Of
particular interest is the response in the inac There was
concern that imperfections in the feedback modeling and
the large number of loops may result in amplification of
beam noise for some frequencies. However, with the finite
sampling rate used in the feedback system, even ideal
conditions would result in noise amplification for some
frequencies. Techniques were developed to analyze the
feedback response for both single loops and for the lina
system as a whole. Several sources of feedback
imperfection were identified and studied
The matrices for a launch loop incorporate a transport
model for that area of the accelerator. including transport
elements between BPM readings. heam positions and
angles, and corrector settings. In some cases, the online
model does not accurately reflect the accelerator response.
so feedback calibration is needed. This is accomplished by
moving each corrector one at a time through a range of
values and measuring the fitted beam positions and angles
for each setting. The slopes of these states versus the
corrector settings are incorporated into new feedback
matrices. Recent software improvements have made the
calibration system easier to use. hut it remains an mv sive
procedure for which it is difficult to get sufficient beam
time. In a few areas of the machine, the model is so poor
that the feedback cannot be used without calibration. In
marginal cases. imperfect modeling simply degrades the
feedback performance.
Another performance consideration is the time
response of the correctors. The design for linac loop,
assumes that corrector changes are implemented with a
delay of three feedback iterations; one iteration is allowed
for calculations and communication. with two additional
iterations for the magnetic field to change Recent
measurements indicate that typical hnac corrector can
move from 10 to 907 of a requested change within about 9
120-Hertz pulses. For most of the Inac loops. which run at
20 Hertz, the response is close enough to that used in the
feedback design However, for the last loop in me inac.
which was upgraded is run at 60 Hertz. the corrector
response is relatively slow and not adequalciy modeled
Simulations for the standard feedback design. shown so
Upcoming Pages
Here’s what’s next.
Search Inside
This article can be searched. Note: Results may vary based on the legibility of text within the document.
Tools / Downloads
Get a copy of this page or view the extracted text.
Citing and Sharing
Basic information for referencing this web page. We also provide extended guidance on usage rights, references, copying or embedding.
Reference the current page of this Article.
Hendrickson, L.; Adolphsen, C.; Allison, S.; Gromme, T.; Grossberg, P.; Himel, T. et al. Fast feedback for linear colliders, article, May 1, 1995; Menlo Park, California. (https://digital.library.unt.edu/ark:/67531/metadc691680/m1/3/?rotate=270: accessed March 19, 2024), University of North Texas Libraries, UNT Digital Library, https://digital.library.unt.edu; crediting UNT Libraries Government Documents Department.