These notes are intended to be a brief guide to operating the SCT tesbeam facility
at the CERN SPS H8 beamline (in the "North Area" Prevessin site, Building 887 (EHN1),
Phone 76733). We are in barrack HNA257, which is on the lower floor of the large
experimental hall EHN1 (building 887), on the side towards the Saleve.
The testbeam setup is highly changeable as equipment comes and goes,
so these notes may be incomplete or out of date. It is therefore advisable to visit
the control room before your first shift in any testbeam period for a briefing on
the latest details.
- Wear your film badge at all times ,
& generally take responsibility for your own safety.
- Please do not be late, especially when relieving the night crew in the morning.
- On arrival, make sure the departing shifters give you an update on the status
& show you the planned program (often on the whiteboard or in the logbook.)
- Note your arrival in the logbook
Generally during the course of the shift:
- Fill out the scan sheet, eye-balling all the items required to be checked
- Check the chillers
- Check the SCTHV VME modules: all cabled channels have their LEDs on ?
- Check the progress of the current scan by looking at the DAQ text console
- no errors, timeouts etc. ?
- Do the histograms look OK ?
- Are triggers occuring at a reasonable rate (approx. 300/burst) ? This is
particularly imporant after an access since sometimes one of the magnets
has tripped or the TAX (the beam blocker) hasn't actually moved clear..
- Have recent files been written correctly to the HP data disk /data3 or /data4 ?
- Do the files "tbrunNNNN_hsmcopy_finished" exist on /data3 or /data4, indicating that
the Central Data Recorder (CDR) is operating correctly ?
- If the magnet should be on, is it really at 5000A ?
- If you can't fix a problem quickly yourself, call an
expert . For DAQ or module problems, call me (in August 2001,
my hostel phone is 79224 - any time). For the beam or problems with access or
interlocks, call CRN (Control Room North, 75566) - they can often help.
The procedure for doing a scan is as follows.
- Before starting the scan, SCTDAQ should be running normally, with modules responding
to TriggerBurst without protocol violations or timeouts.
- If you have problems with modules, or starting SCTDAQ, use any of the usual
techniques for debugging module problems, including killing and restarting
SCTDAQ or crate power cycles.
- Prepare the next scan sheet filling in all start of scan details
- Make any changes to current conditions, e.g., Magnet Current, Angle, Bias Voltages.
- Final check of TriggerBurst to make sure all modules still alive
- If the Scan is to be a standard Threshold Scan, you can use the ROOT macro
TBSpecial(15000) , where 15000 is the number of events per run. This
can be different as noted in the planned program. Other special scans such
as Trigger Delay will have special macros or must be done as individual runs.
- In case of errors or timeouts, use the "Stop" button to terminate the scan. Fix the
problems. Restart the scan at the last point using the macro
TBSpecialSelect(5000) ,
which gives you the option of which scan point from which to continue.
- Fill out the scan check sheet as required at the end of the scan
- Paste the scan sheet into the logbook
TBDAQ is a companion to SCTDAQ, and uses SCTDAQ services whenever module configuration
as required. To start a TBDAQ session from scratch:
- Start SCTDAQ in the usual way, including running RESMAN if the crate or PC has
been power cycled. A ROOT console window and some histogram canvases should open.
- Special for 2001: Do not use the "Startup" button - it
prevents the TBDAQ VME mappping.
- Get all modules to respond using any of the usual SCTDAQ debugging methods -
TripRecovery, HardReset, Restart etc.
- Do ".L TB.cpp" to load the TB macro.
- Do "TBStart()" to load the TB dynamic library, do the module mapping etc.
- Do "TBRun()" to start a single run. If you have not changed module or chip
settings, these should be already on the defaults. The macro prompts for
a "Scan Type" and "Scan Value". This is simply for recording purposes. Use
the standard scan type integers listed in c:\sctdaq\stlib\sct_scans.h.
- Do "t->SetEventsRequested(15000)" to set the number of events per run to 15000.
Note: special macros like TBSpecial() ignore this setting.
- Use "TBSpecial(15000)" or "TBSpecialSelect(15000)" to do a scan of 15000 events,
or another number as required.
In 2001, we use SCTHV for detector bias. These are red or pink modules in the
VME crate with four channels.
To set the bias voltages for a Bias Scan point, use the macro "TBSetBias(Vunirr, Virr)".
This will configure the modules to the required voltages for the unirradiated and
for the irradiated sensors, and then execute a HVRamp().
E.g., for nominal conditions, "TBSetBias(150,350)".
Bias voltages can also be set using SCTDAQ functions, e->ConfigureVariable(mod,typ,val)
followed by e->HVRamp(0) or pressing the HVRamp button. For Vdet, typ is 30; for
Idet (current limit in uA) the typ is 60.
Do not forget to use the "HVStatus" button to check that the HV is really ramping.
Also monitor the LEDs on the actual SCTHV modules - they should be on, and not blinking..
The Magnet terminal is in the next control room, near the Pixel electronics. It is
an old PC in rack R25 with an aluminium panel above.
On the panel, the LEDs for "CT4" should be ON, and for "CT5" OFF. If CT4 is off,
press "CT4 START".
On the PC, there should be a large LabVIEW window. If not, click the icon to start it.
For all magnet operations, several steps are required. ALWAYS
wait till one step is finished before starting the next.
To switch ON the magnet:
- Click "On Standby" --- wait until the "OnStb" LED is illuminated
- Click "On" --- wait until the "On" LED is illuminated
- Set the target current, eg, 5000A
- Click "Execute Ramp", and watch to see that something happens.
To switch OFF the magnet, reverse the above procedure.
- Set the target current to 0A
- Click "Execute Ramp", and watch to see that something happens.
- Click "On Standby" --- wait until the "OnStb" LED is illuminated
- Click "Off" --- wait until the "Off" LED is illuminated
If there is an interlock problem, go to the Interlock panel, and click "Read". If
this doesn't help. do "Load" and load the "intlcks" file. If this still doesn't
help. call CRN or one of the magnet experts .
Ramping the full 5000A, either up or down, takes a long time, about 30 minutes.
To set the angle, first take an access. Climb to the rear of the black SCT chamber,
and turn the small handle until the scale reads the required setting in millimetres
for the angle inside the box. The scale is read at the join between the aluminium
and brass support pieces, on the side further from the box.
Note: if you are unhappy about climbing into the magnet with it switched on, then
take the time to ramp the magnet down.
To open the zone, go to the SPS terminal, "eanorth>". There should be a "NODAL"
window. If not, restart the terminal. In the NODAL window, you should have a
"TREE" prompt. If not, do "RUN TREE" (in capitals).
Do "ACCESS", then "OPEN". Our door is 158. Don't forget to enter your name. Check that
something is happening on the terminal screen.
After about 5 minutes, the door goes to "Access with Key". Take a key, go inside,
come back, return the key, and press "End of Access".
Then go to the terminal, do "ACCESS", "BEAM ON", door 158 and your name.
If there seem to be problems, eg, a very low raw trigger rate (scalers 1 and 2)
after the beam should be on, then call CRN. Likely that a magnet has tripped
or the beam blocker hasn't properly moved.
If you see the message "Veto on Chain H8" after ending access, don't worry. This
just means you haven't waited the few minutes it wants to ensure that anybody left
inside the zone can see the warning lights and get out. Wait one minute and then
try again.
In 2001, we use two Huber chillers, one to cool the chamber base plate and N2 gas,
and the other to supply cold fluid directly to the modules. These chillers will
usually be set to -25C, resulting in displayed fluid temperatures around -20C
depending on heat load and insulation.
The Huber temperature is set by pressing "Set", then up or down arrows, then "Set"
again. Make sure the Huber power switch is in "II" mode (cooling & pump),
rather than "0" (off) or "I" (pump only).
In 2001, environment temperature monitoring is by:
- Huber fluid temperatures, visibile as red numbers on the side of the unit.
- Two small white stand-along monitors, only visible when inside the zone,
one recording the base plate and one the air inside the chamber.
- An HP multiplexer reading several Pt1000 sensors in the chamber. This is
monitored by a software on the PC, "HP BenchLink".
- The SCTLV3 hybrid temperature monitors recorded by SCTDAQ
suggestions/corrections