Transcript LER BBA Quad Shunts

LER BBA: QUAD Shunt History
• In 1995, switchable shunt resistor circuits were installed on
each quadrupole in the SPEAR ring to facilitate BBA. They
were used with great success.
• These shunts were removed as part of the SPEAR3 upgrade,
and the hardware was sent to Salvage.
• The shunts were moved to B15 and (expensive) proposals
were made to upgrade them for use in PEP2.
• The shunts were taken as is to NLCTA, cleaned up, and
tested on a LER Quad in Magnetic Measurements. 42/45
were found to be fully functional.
LER BBA: Quad shunt hardware
• The shunts consist of two high-power resistors in parallel on
a 1” copper heat sink, controlled by a basic IDOM-driven
FET switch.
• Water cooling unnecessary for our application (resistors operating at
~15% rated power on a large thermal mass).
– We have tested at 150% expected operating power for ~1 hour. Max
temperature reached was 55ºC without water cooling.
– Each shunt will be tested at operating current/voltage before installation.
• Design is simple yet robust, with a low mortality rate even after months
outside in Salvage.
• Minor control board modifications (snubber circuit rework) are being
made to protect the FET switches at our operating values.
• No-frills controls (+24V, on/off, and as-yet-uncalibrated analog readback
of shunt current) make for straightforward cable runs.
LER BBA: Program goals
• To find the magnetic center of quadrupoles at optically
sensitive locations in the ring. Shunts are used to vary the
field in one magnet which otherwise lacks individual
control.
• In IR2, there are four strings powering a total of 14 quads:
QFCY1 (4), QDCY2 (4), QFCX3 (2), QDCX2 (4)
• Throughout the ring there are 37 individual QD and QF
quads which have been identified as prime candidates for
BBA.
– These are in the arcs, immediately adjacent to sextupoles
– There are 2-5 quads per half-arc
LER BBA: Quad shunt inventory
• We currently have 18 shunts suitable for use on individual
QD and QF magnets.
• High power resistor values 0.45 W and 0.55 W will shunt 3-4% of a
QD/QF magnet’s current.
• We are expecting delivery of resistors for 8 of the IR2
shunts this week.
• Available resistors (9 W) will shunt 1-2% magnet current.
• Resistors for the remaining 6 boards are on order, we
expect delivery in December.
• These will also be 9 W.
Quad shunt installation, QD/QF
• We are installing 1-2 moveable shunts per half-arc (2-4
shunts per IR/micro), based on quad(/sext) population.
– The shunts hook directly across the busses of a quad with leads
that bear an uncanny resemblance to automotive jumper cables.
• Design to be approved.
– Control cables have been pulled everywhere but IR2.
• After a round of BBA, these shunts will be moved (during a
ROD, presumably) to a neighboring candidate QD or QF.
• Lather, rinse, repeat. (Two RODs to complete the QD/QF
magnets)
Quad shunt installation, IR2
• We are planning to install one non-moveable shunt on each
of the 14 QF/DCX/Yn quads in IR2.
– The shunts also hook directly across the busses of the quads, but
with a more captive ring-clamp fixture.
– Modifications to the shunt covers will have to be made to keep
fingers out (quads are not PPS interlocked).
• We hope to install eight of these before turn-on.
• The remaining six will go in this December (or over the
break).
– We are planning to have all the cables pulled this downtime, but
some may carry over into December if we run out of cable.
Quad shunt controls
• To the control system, the shunts will look like generic
digital devices with a corresponding analog readback.
–
–
–
–
With names like: PR04 SHUNT1, 2, 3…
One digital control (IDOM bit) per shunt (command ON/OFF)
One status (ON/OFF) reflecting that IDOM bit
One ASTS value reflecting current through the shunt resistors
• Not presently calibrated well enough to tell more than whether the shunt is
shunting or not, really.
– SCP Panel and analog status display group exist, though not linked
to a PEP panel yet.
– We’ve requested that the shunt on/off status be on the SDS CUD
under LGPS with ON being RED.
SCP Panel & Analog display
How it all works
• We are creating a flat file to hold the information of which
shunt is on which quad, with a list at the bottom of which
shunts can go to which quads.
– This file will live on the VAX somewhere with an entry in e-log
(which e-log to be determined) with its location.
– The person who moved the shunts is required to change this file
• For data acquisition, a button macro will prompt which
magnet or which shunt you wish to do BBA on, then load
up a CRR and take data.
• Data to be analyzed offline by the Woodley/Wolski coupled
analysis matlab code.
The Crew
• The project has been brought to you by the usual
suspects:
– Marc Ross, Doug McCormick, Tonee Smith, Justin May,
Janice Nelson, Nicole Varela, Juan Cruz
• with the invaluable input of Those Who Know:
– Paul Bellomo, Jerry Yocky, Bob Fuller, Nancy Spencer,
Mario Ortega, Mark Woodley