NEXT
User name Williamson
Log entry time 00:43:58 on February 20, 2004
Entry number 77135
keyword=Run Plan for Friday, 02-20-04
Run Plan Friday 02-06-2004Run Plan - Wednesday 01-28-2004
Run Plan for Friday 02-20-2004
Run Plan
- Today will be essentially 100% devoted to production data
taking on LH2 so the program is:
- Production data taking on LH2 (see below)
- Pay attention to beam position.
- Check the Halo detectors rates
- Pay attention to the beam trip rate.
- If there is a beam down period (more than 1 hour)
- Take DNL data (NA first, then French)
- Inquire about need for a controled access (Kellner, Damon,
Gary, ...)
- Check
that the feedback system is ON.
- Check convergence with Kaz's script
- Also check the G0Integrity
related information
- Page Kaz if you run
into
problems
- During the Swing shift, Dave McKee may do a 1 hour run with
GMS if the prerequisites are met:
- A HowTo is written
- Analyzer is updated for GMS analysis
- An analysis script is available to spit out gains
- Previous data has been analyzed.
- Interference with normal data
taking is demonstrated (to the degree possible) to be absent.
LH2 target measurement
Repeat
the following sequence:
- Take four hour-long runs with
prescales set normally (Fastbus
prescale
5000, Fast Clear Disabled, 30Hz see instructions below in this run
plan)
- Take a 5-10 minute
run
(Fastbus
prescale
5000, Fast Clear Disabled, 30Hz)
with the coil
modulation active by
following the instructions in the short Howto
for Coil Modulation below. (See
instructions below
in this runplan)
- Take one run of 50k events in
120
Hz oversampling mode every
8 hours only (Fastbus prescale
at maximum, Fast Clear Disabled, 120Hz). The shift worker is in charge of the
analysis of this file (see documentation below in this
runplan)
- Continue
this sequence until Saturday when IHWP is switched (or until your run
coordinator tells you otherwise).
On
a more general basis
-
We will likely change the IHWP
state every two days. At that time, Make a Moeller measurement
before and
after the change. Also do an RIP scan just after the change.
Next change is scheduled for Saturday
02/21.
-
Moeller measurements should be done every 2 days (normally
in connection with IHWP change). Next
measurement scheduled for Saturday 02/21 to be preformed by Dave
Gaskell.
-
BPM/BCM calibration should be done typically every
week. Next measurement to
be done on Wednesday 02/25. A lumi calibration for the
"frame" target will be carried out at the same time (Lars/Silviu/Greg
should be present. Greg must alter the administrative limits on
current)
-
Use time with no beam for DNL calibration (NA and French).
Also search
for helicity correlated ground loops (see
below).
REMINDERS
General
- Moller solenoid should be ON
- Hall C fast feedback energy lock is ON
(requested by PD since Hall A is down)
- Hall C Fast feedback position is OFF
- MCC can view the Halo
monitor
GUI.
Maintaining a low halo rate (see below) is part of beam tuning.
- G0 girder correctors
are
physically
disconnected
Useful Links
- The 6 mm hole target is in use.
- See references for positions and rates in entries #74605 (BPM)
and # 76406 (Rates)
- Automated procedure for IA/PZT scan
- See Kaz's entry # 76387 (IA/PZT)
- Shift worker
should perform the 120 Hz file analysis
- See Lars' entry #
76401
(120Hz)
- Shift worker should also
make an entry of Feedback system results
- Shift worker
(yes again, but the word work is associated with
you)
- Report the results of G0Integrity check)
- See an example in entry #76794 ( G0Integ-report)
Beam
conditions
40 uA with 2x2 mm
raster ON
A) Check
the NEW nominal beam
positions (see e-log entry
#76592 (BPM-NewReference))
|
3H00A |
3H00B |
G0 |
G0B |
X [mm] |
0.7 |
0.7 |
-1.0 |
-1.7 |
Y [mm] |
0.1
|
0.45
|
-1.9 |
1.8 |
We allow MCC a tolerance of 0.3
~ 0.5 mm for G0 and G0B. Ensure that
there is NO lock on 3C20 (check with MCC when position is off and
search for a reason)
B) Check the Halo rates
MCC has the same Halo rate monitor
GUI on
their screen as we do. They should be using it for beam tuning.
The following table is about the best that has ever been seen for the
rates. Since Friday 02/13, the rates have been typically 2 times
higher on 5 and 6 and up to 4 times higher on 3 and 4. This may
be due to increased "background" rather than increased halo. The
principal monitor is Halo3, which generally should be no more than
10kHz. The rule of thumb for the fraction of beam hitting the halo is
to subtract the "no halo" rate from the "6 mm halo target rate" for
Halo3 and then to multiply by 0.66 ppm per kHz. For example, the
40 uA data below yields 0.96 ppm.
Monitor |
No halo target
[Hz/uA] |
6mm Halo target
[Hz/uA] |
No halo target@40uA
[Hz] |
6mm halo target@40uA
[Hz] |
Halo3 |
18 |
25 |
1440
|
2080
|
Halo4 |
20 |
30 |
1600
|
2880
|
Halo5 |
250 |
300 |
20000
|
28000 |
Halo6 |
140 |
170 |
11200
|
16000 |
The 11 mm
hole position is at -40.5 mm (stepper motor position on the GUI), and
the 6 mm hole
position is at -57 mm. Out Position is "Home" or 0 mm.
Always use
the 6 mm hole target except when doing coil modulation.
If rates are too large,
ask MCC to improve on beam quality.
C) Watch Herberts paddle and associated anode
current
It
should
be giving a value of about 3280 nA for 40 uA of beam with
LH2 (see Paddle
)
General
settings for asymmetry
runs
- I_beam = 20uA for G0 empty target (Gas
Hydrogen).
- I_beam = 15uA for G0 target frame
(Aluminium)
- I_beam = 40uA for G0 full target (Liquid
Hydrogen)
- The nominal HV file is the proton reduced gain
file
(HV.halfgain_2nd_engineering_run.hvc).
- A separate file is used to
restore lumi and
halo settings (HV.2003-12-09-22.lumi_halo_on.hvc).
- Check to make sure that HV is on after a long
period down
especially
following
a magnet fast dump or Moller run.
- When beam is being restored after a long period
down make sure
that the
MCC tunes to the correct position at the four BPMs in the hall (posted
on the X-terminal adjacent to the DAQ computers). Tuning to within
0.3-0.5 mm is acceptable.
- Shift crews should watch the LTDs (visible on
the TV above the DAQ
computer
in the counting house) to make sure the red error lights are not
flashing.
The lights in the electronics cage have to be off, which should be the
default state, for these lights to be visible. If there are LTD errors
contact the NA electronics expert immediately.
- Here are the default DAQ settings that should always be used
unless a
run
plan specifically calls for something different.
- Configuration mode in RunControl GUI should be
"ts_full2"
- G0 DAQ Configuration Tool GUI
- 30 Hz mode enabled.
- Fastbus prescaler (PS4) = 5000
- FastClear Disabled
- GMS prescaler (PS5) = maximized
- NA CFD threshold setting = 7
- NA CFD width setting = 0
- DMCH Reference Config =
Reference_FiB_Buddy
- DMCH Threshold: 50mV
- For other configurations (Fastbus, DNL, ...) go to e-log
entry # 69965 (Config)
- Halo target (6
mm hole)
:
IN
- Insertable Half Wave Plate : IN until
Saturday 02/21
Instructions for 120
Hz data
taking and analysis
- Change G0 DAQ Configuration in the GUI Tool
- 120 Hz
mode enabled. (don't forget to submit this change)
- Fastbus prescaler (PS4) = To be set on Maximum (hit the
key)
- FastClear Disabled
- GMS prescaler (PS5) = maximized
- NA CFD threshold setting = 7
- NA CFD width setting = 0
- DMCH Reference Config = Reference_FiB_Buddy
- DMCH Threshold: 50mV
- When you are done use the Configuration GUI to :
- Go
back to 30 Hz
- PS4 back to 5000
(manually)
The 120 Hz analysis still has to be done by hand. From now the
shift
worker
has to perform this analysis so that any
problems
with 60 Hz noise is detected early.
The following analysis procedure works for 120 Hz
runs
that
have
their
data file(s) on the local disk.
It has been tested on cdaql2, butd similar machines have a
similar setup.
-
Log into cdaql2 (or similar machine) as user : gzero.
-
Type 'source ~/G0Analysis/setup_120Hz.csh'. This
will set
the
environment variables needed to run Jason's 120 Hz analyzer, instead of
the default G0Analysis. (Don't
use this window to do
standard 30 Hz
analysis!).
-
Type '~/G0Analysis/g0analysis120Hz.csh RRRRR', where
RRRRR
is the
run
number. This should start the 120 Hz analyzer and
eventually produce a
rootfile called ~/scratch/root/G0Scaler120Hz.RRRRR.root
-
When the analysis is done, use
Jason's script to
extract the
60
Hz
noise, and multiples thereof. To do this,
type
'~/G0Analysis/G0Scripts/extract60Hz RRRRR'.
-
MAKE A
LOGENTRY
containing the run number and the relevant text
output
from extract60Hz.
-
Inform the shift leader, especially if the values are out
of
spec.
For more info about the 120 Hz analysis, see logentry
#70368
(120Hz-analysis).
Instructions for Coil modulation
data taking
-
Ask MCC to
retract
the Halo target (home position or 0. mm). In the mean time Start
a run
-
Check with general Tool that
the fast position
feedback is OFF
-
From the Monticello screen, open the modulation coil
response screen
by doing the following from the Accelerator Main Menu :
- Under Magnet, Open "magnet
commander"
- Open "3C"
- Open "3 C Position
Modulation"
- "Hall C Position/Energy Modulation" is
displayed
- Note that this item 2/ is just to check
the Coil
modulation is turned ON
-
Logon and pop up a window on gzerol3 as user
gzero
-
On the gzerol3 window type:
- "cd/home/gzero/users/rutledge"
- "bash"
- "rm -f simple-in.txt"
- "ln -s grid-0.25sec.txt
simple-in.txt"
- "./auto-mod2 +g"
-
When the script exits (takes 5 minutes), end the run
-
Ask MCC to put back the 6 mm hole target (-57 mm)
You're done.
Remark : The option + g generates a grid shape. Witout
this option, the shape will be a cross
** NOTE that if you experience a long (more than 2
minutes) beam trip during the data
taking, the data are
useless. In order to save time, the best is to
keep the run going and just restart Gary's
script. If this a single
trip,
then the script will repeat the grid 20 times, so this should be
fine.
Things
to do during shift
- Check whether the feedback is still working. If
it fails, call
the
RC and Kaz.
- Check the IA, PZTX, and PZTY strip charts to
be sure they are
not
railed,
or oscillating wildly. These strip charts are created with data of
rather poor statistics so do not try to draw conclusions about feedback
convergence from them. Instead perform the next step.
- After about two hours of data has been taken
with the feedback
turned
on,
run Kaz's script to do the integrated charge and position asymmetrie
check
- If the feedback is not converging, notify MCC
that the beam is
unacceptable
because the feedback is not working; you should
call Kaz, who will probably instruct you to try a set of IA
and PZT
scans. If the slopes are less than ~150-200 nm/V for the PZTX and PZTY
or
less than 300 ppm/V for the IA, or one or more of the devices do not
give a zero crossing, the feedback cannot be made to converge;
Kaz will provide further guidance on what to try if it is needed.
- Also
use G0Integrity to see if beam parity quality parameters are within the
specifications
- Shift crew should watch the LTDs (visible on
the TV above the DAQ
computer
in the counting house) to make sure the red error lights are not
flashing.
- The lights in the electronics cage have to be
off, which should
be the
default state, for these lights to be visible. If there are LTD errors
contact the NA electronics expert immediately.
- Use
g0realtimemonitor to check :
- beam parity quality
- Detector health (CFD_TDC, MT_TDC, TOF,
...)
- Once an hour, look at the last complete
g0Integrity file.
- After each run, update the purpose of each run
within Good_for
GUI.
- Once a shift, don't forget to fill out the
check list (normally this the job of the shift worker).
- Check
once a shift and especially
after
beam restoration, reboot of IOC SE20 etc.
that
the Halo
target position hasn't changed
Things to Do When There is No
Beam
- Take NA DNL runs with the white noise box (see
procedure in the
Beam
How-to).
- Take FR DNL ("green block") runs (see RC for
the procedure).
- Load the HV full gain file:
HV.fullgain_2nd_engineering_calibrated.hvc
from cvxwrks@cdaqs2:$EPG0HV/tk
- Select French
configuration as FoB (using g0daq_config
GUI) and set
the
threshold to 20 mV for French. 30 Hz, ps4 maximum,
fastclear
disabled.
- Take a long ~2hr DAQ run.
- Restore
standard
configuration (HV file and
DAQ) when you are done
- Take hour long runs in the nominal DAQ mode to measure
pedestal
differences.
The following IOCs must be in forced gain mode at the listed settings
(see
BPM/BCM Calibration section of Beam How-to):
IOC |
X Gain |
Y Gain |
iocse14 |
1900 |
1900 |
iocse17 |
1850 |
1800 |
iocse18 |
1800 |
1800 |
iocse20 |
1900 |
1900 |
Make sure the IOCs get set back to auto gain when beam is restored
to
the
hall.
Steve Williamson
Cell: 876-1791
Appartment: 7468 (Residence facility room 2C)
e-mail: sew@UIUC.EDU
A copy of this log entry has been emailed to: biselli@ernest.phys.cmu.edu,carlini@jlab.org,dmckee@jlab.org