The Compton polarimeter can be used to non-invasively monitor the beam polarization. It has three major components: an optics table with a laser that can lock to a Fabry-Perot cavity; an electron detector to count scattered electrons; and a photon detector to count scattered photons. At present, the electron detector is in commissioning, so only the photon detector is fully functional.
There is additional documentation at Hall A compton webpage
- 1 Standard Compton
- 2 Compton Upgrade
- 2.1 Green Fabry-Perot Cavity
- 2.2 Photon detector
- 2.3 Portserver assignement
- 2.4 FADC DAQ
- 2.5 Electron detector
- 2.6 DAQ setups for Fall 2015
- 2.7 ==IOC iochacp2 boot parameters
- 3 Compton Meetings
- 4 Compton Experts
- 5 Past Compton Experts
Monitoring the beam polarization using the Compton polarimeter requires two steps:
- Take Compton data. These can be long runs, but since you can only check the polarization after a run has ended, it is best to start a new run every hour or two. When you end a run, click "Submit" to dismiss the gray pop-up window.
- After a Compton run has analyzed (this takes about 10% the length of the run), check the logbook  for the Compton asymmetry.
If runcontrol shows that it can't communicate with a ROC or if the data won't analyze, stop the current run and type "coda reboot" into a compton@compton terminal window to reboot the ROCs and CODA. If the problem is with ER1, you will need to type
coda_er -n ER1
into a separate window.
See below for more details.
The Compton is run from the compton machine, which is the leftmost active monitor in the back room of the Hall A counting room. The login name is compton and the password is the same as the one for the adaq account.
To begin a new CODA instance, type
To recover from a CODA crash, type
To open a strip chart, type
To start a run, click the "Start" button in the runcontrol window. After a time, you should see the event count start to go up and the message "Transition Go Succeeded" at the bottom of the screen.
To setup the compton trigger parameters use the spyconf program :
One can select the different data taking modes in General Daq Setup.
On the right panel one can setup the photon detector parameters. We are using only the central crystal only.
On the bottom middle pannel, one can setup the electron detectors parameters.
number of planes needed to generate a trigger coincidence time windows between planes number of adjacent strips ored to be looked in the next plane for the trigger
When it is time to end a run, click the "End Run" button in the runcontrol window. A gray screen should pop up, giving you the opportunity to enter a message (say, if the beam was very trippy). Make sure you dismiss the window by clicking "Submit" -- not "Exit"! Hitting "Submit" sends the run to the analyzer, which is very important.
You may additionally get a pop-up window saying "End run failed!" on the runcontrol window. This is okay. Dismiss this popup window and click the "Reset" button on runcontrol. (Yes, despite the dire warning, you really do want to do this.) You should then be able to start the next run.
While running, you can monitor the beam current, photon rates, cavity status, etc using a strip chart. Here's what it might look like:
- Purple line: Beam current (uA). In the figure you can see that it has tripped a few times.
- Yellow line: Vertical beam position. This should remain roughly constant. A beam trip sends it to the top of the chart (position 0).
- Red line: Cavity power. This should be a square wave: the cavity is locked in the right polarization state, then turns off, then locks in the left polarization state, then turns off, then repeats. The powers for the right and left states appear different but are actually the same. The power should stay comfortably above 400 W. (Cavity will not lock during extended beam trips.)
- Blue line: Scattered photon rates in the detector. As in the figure, these should track the cavity on/off states; if you can tell from the blue line alone whether the cavity is on or off, then you are probably getting good Compton data.
- If this is not the case, this may be due to a bad tune resulting from very low current, or the Compton chicane is off. If one of these is not the cause, contact a Compton expert.
- If your blue line is zero or not displaying, there are three possible causes:
- Beam trip
- HV trip. Check the Beamline HV card 11. Channel 1 should be set to -1800 V for the GSO detector. If the HV is tripping over and over again, call a Compton expert.
- DAQ is not running. Start a CODA run and the blue line should come back. If CODA is already running, it may be in a strange state;
end the run and type "coda reboot" into a terminal to restart the ROCs and CODA, then start a new run.
Checking a Run
Compton analysis results are written to the Compton logbook, which can be accessed onsite at 
Find the run number (newest runs are at the top) and click the html link. (The pdf link does not work.) This will open a new window with the analysis results.
The full results will not display right away (because the analyzer does not finish immediately). It takes the analyzer about 10% of the runtime to complete. So an hourlong run will be analyzed after about 6 minutes.
The most important graph is about a third of the way down the page. You can find it quickly by searching the page for "Counting Rates". It's the graph titled Asym in ADC bins:
Here are the important things to look for:
- The green line (cavity off) should be consistent with zero. This shows that there is no asymmetry in the background.
- The red/blue lines (cavity on, right/left) should be roughly symmetric about zero. This reflects the fact that flipping the cavity polarization changes the sign of the asymmetry. (It's ok for there to be slight differences as above.)
- At the extremes (highest/lowest point) we should see an asymmetry greater than about 5%. (Here, we see extremes of about 8% and just over 6% for right and left.) If the asymmetry drops below 3%, we worry. Call an expert.
Getting good data sometimes requires a CIP (Compton interaction point) scan, where the beam tune is adjusted slightly in order to maximize its interaction with the photons in the cavity. Call a Compton expert to perform such a scan if:
- There has been a new beam tune through the Compton chicane.
- The beam has been shut down for longer than an eight-hour shift.
- The photon counting rates drop drastically during steady running. (Compton rate (blue line) in strip chart -- see below)
Green Fabry-Perot Cavity
to be updated soon...
The new photon detector crystal is a cylinder of GSO with a diameter of 6 cm and a length of 15 cm. It was installed in Hall A in December 2008.
In October 2008, the crystal was tested at incident photon energies of 20 MeV, 22 MeV, 25 MeV, 30 MeV, and 40 MeV. We are working on extracting the response function.
PREx requires an integrating DAQ system, which will run on a Struck FADC. We are currently commissioning this DAQ with a prototype FADC.
Homing the photon detector table
You need to put the photon detector in the right position before you can take data: How to home the table
New DAQ flags file controls
The FADC photon detector crl code lives in the franklin/crl directory. To change the running mode, link a run file to the fadcScaler.flags file. This is done by deleting the old fadcSclaer.flags file and then creating a link. To do this, type:
rm fadcScaler.flags ln standardRun.flags fadcScaler.flags -s
There are several different configuration flags files. These are:
standardRun.flags, for production data Etrigger.flags, for triggering the FADC on the electron detector miniMegan.flags, for miniMegan LED pulser runs test.flags, for playing around
There are several flags which can be changed in the file. These are:
- readoutWaveformLength (number of samples read out per trigger, should be 100 for production data)
- readoutWaveformStart (sample to start read out -- trigger should be near 0, so -50 centers a pulse in a window of 100)
- maxStoredWaveforms (number of waveforms stored per MPS)
- maxStoredSums (number of sums stored per MPS)
- bitPMTtrig (1 means triggering on PMT signal)
- bitEDetTrig (1 means triggering on electron detector)
- bitPulserTrig (1 mean triggering on 10kHz pulser)
- bitMiniMegan (0 means LEDs not flashing, 1 means flashing)
- bitDelta (0 means delta LED not flashing, 1 means flashing)
- VarDACsetting (0 means variable LED changing pulse amplitude; set it to some value to leave it constant, 35200 corresponds to a 250mV signal on the scope in the counting house)
- numADCchan (number of FADC channels read out as sums; generally 1 for the GSO, 3 reads out the fingers also)
- prescaleTrig (prescale on the trigger for sums; 1 turns off prescale)
- prescaleRand (random prescale on the samples; 0 turns off random prescale, 4 means that every 4 samples is random in time)
(Courtesy of Megan Friend )
5 cptaq5 6 cptps1 7 cptps2 ( RPC5 ROC1 on port 1 and ROC2 on port 2)
10 iochacp2 11 TS crate
We are commissioning a new FADC-based DAQ for the photon detector. See run information.
Webcams were added to monitor the :
- hacp2 EthernetBlaster
- hacp3 webcam trigger
- hacp4 webcam on Beam Safety module, Front End power supply
- hacp5 webcam on silicon strips detector current
- hacp6 APC
- plug 1 Front End power supply
- plug 2
- plug 3
- plug 4 Webcam current
- plug 5
- plug 6 Ligh for viewer
- plug 7
- plug 8 Light for cameras
How to reboot stuffs
APC : hareboot4
*plug 2 webcam *plug 3 EPICS Edet *plug 4 ETROC *plug 5 ts crate *plug 6 FADC crate
login hareboot4 using the usual password ( if you don't know it ask the run coordinator or a Compton expert )
select the outlet you want to reboot with the pull down menu "Immediate reboot"
and wait for the IOC to come back
DAQ setups for Fall 2015
Updated Sept 29, 2015
There are mainly two DAQ systems for Fall 2015.
FADC_prod runs on the cpt1 session. It is for the CMU DAQ.
cntFadc1 runs on cpt2 session. It is the new counting-mode DAQ.
To start the CMU DAQ, type "startcoda" on compton@compton. To start the counting-mode DAQ, type "countFadc"
These scripts try not to stomp on each other, in particular not killing each other's processes. I don't have a lot of experience with this yet, but it looks, so far, like these sessions run independently. The "kcoda" script will, however, kill all CODA processes, regardless of the session.
==IOC iochacp2 boot parameters
boot device : dc unit number : 0 processor number : 0 host name : eosfs file name : /eossite/op/iocs/iochacp2/vx/vxWorks.st inet on ethernet (e) : 126.96.36.199:fffffc00 host inet (h) : 188.8.131.52 user (u) : vxwrks flags (f) : 0x80 target name (tn) : iochacp2 startup script (s) : /eossite/op/iocs/iochacp2/startup
Electron detector signals on patch panel
Output cable patch
2 48 24 5(5) 79 10 6(6) 56 20 7(7) 25
Analog 24 Analog 26
N6-38L Plan A ch 96 left N7-39L Plan A ch 96 right
Electron detector input signals
0 P0 Half of helicity flip frequency
1 P1 Helicity
2 P2 Laser on/off
3 P3 Laser helicity
4 Laser helicity inverted
7 Photon from integrating
Electron detector motion
- Friday, December 7, 2012 11am EST - Compton
- Monday, January 14, 2013 1pm EST - Compton
- Tuesday, February 12, 2013 3pm EST - Compton
- Thursday, March 28, 2013 1pm EST - Compton
- Wednesday, May 1, 2013 2pm EST - Compton
- Thursday, September 26, 2013 2pm EST - Compton
- Thursday, November 14, 2013 2pm EST - Compton
- Tuesday, December 10, 2013 11am EST - Compton
- Thursday, February 27, 2014 2:30pm EST - Compton
- Friday, April 4, 2014 2:00pm EDT - Compton
- Thursday, May 21, 2015, 11:00am EDT - Compton
- Thursday, June 4, 2015, 11:00am EDT - Compton
- Thursday, July 2, 2015, 11:00am EDT - Compton
- Thursday, July 23, 2015, 11:00am EDT - Compton
- Thursday, August 6, 2015, 11:00am EDT - Compton
- Thursday, September 3, 2015, 11:00am EDT - Compton
- Thursday, September 17, 2015, 11:00am EDT - Compton
- Monday, September 28, 2015, 11:00am EDT - Compton
- Monday, October 19, 2015, 11:00am EDT - Compton
- Monday, November 9, 2015, 11:00am EST - Compton
- Monday, November 23, 2015, 11:00am EST - Compton
- Monday, December 7, 2015, 11:00am EST - Compton
Past Compton Experts
|Abdurahim Rakhman||x5626 (Lab)||315-391-4622 (cell)|
|Megan Friend||412-302-0576 (cell)|