For Source Runs use the special Source Runs instructions

To (re)start the DAQ from the beginning

1. Close any GUI windows or DAQ windows.
2. Click the "Rocket Ship" icon to start the Shifter menu program.
3. Select the type of run you wish to take from the menu.
• For most beam running, choose option (1).
• For once-per-shift pedestals, choose option (2).
• For once-per-shift LED pulser runs, choose option (3).
4. In the same area, or another desktop, start the GUI by clicking on the GUI icon (screen with wavy traces).
5. If the DAQ is restarted, the GUI should also be restarted. The DAQ should always be started before the GUI (the DAQ window should say "State changed to LOAD".

Normal (non-source) Run Taking

1. Restart DAQ (with cntrl-C)
2. Select option (5) "Setup for source test runs".
3. From the submenu, select option (3) "Load Beam Mode firmware into a HTR"
4. Enter the HTR# you want to reprogram. Answer 'y' to any "Do you want to/Are you sure" questions.
5. Reprogramming each HTR will take approximately 4 minutes.

1. Start the DAQ program from the menu by selecting option (4) "Normal Data run".
2. Open a GUI.
3. Check any advanced parameters which may have changed in the "Details..." window. The "Details..." window produces a tree of modules and parameters. Some of the most important/useful are:
• RunManager/Events -- number of events to take
• RunManager/PerSpill.BeamData -- Whether to take beam data (or only calibration)
• RunManager/PerSpill.PedInSpill -- Number of in-spill pedestal events per spill to take
• RunManager/PerSpill.PedOutSpill -- Number of out-of-spill pedestal events per spill to take [observationally, 200 is about largest number achievable]
• EventBuilder/EventBuffers -- maximum number of events to take in a spill
4. Put pre-run comments into the comments box. Any comments from after the run must be added using the (currently non-existant) web page.
5. Hit INIT
6. If the initialization is successful, hit RUN. The DAQ should run and eventually stop. [at this point all the modules will return to a STOP/OK state]
7. If the initialization is not successful, try it again. This will usually fix problems with ECALReadout or TDCReadout. It will sometimes fix problems with HCALReadout. If INIT fails several times, take a look below for debugging tips.
8. If the run was successful, proceed to data plotting.
9. Run numbers less than 200 are not saved, and can be used for test/debugging runs. Run numbers larger than 200 are auto-incremented by the database.

Plotting the data

1. Create another text/terminal window.
2. Start ROOT : root
3. Load the data access library [ once per ROOT session ] : .L ~/lib/libHTBDAQ_Data.so
4. Your data should be available as "/data/tmp/HTB_[your run number].root"
5. To run a ROOT script: .x [path to the script, as below]

Some useful ROOT scripts

• ~/src/HTBDAQ_data/tools/plot_wire_chambers.C("filename") -- look at wire chamber data
• ~/src/HTBDAQ/tools/checkCAPIDs.C("filename") -- verify CAPIDs in the file.
• ~/src/HTBDAQ/tools/plotPulse.C("filename",HTR#[,optional event number]) -- plot the pulse (hopefully) in the HCAL
• ~/src/HTBDAQ/tools/plotPedsPerCapid.C("filename",HTR#,Input#) -- plot the pedestals for a given HTR and HTR input (1-24).
• For pulser data, try .x ~/src/HTBDAQ/tools/plotPulse.C("/data/tmp/HTB_192.root");

Debugging

Symptoms and solutions

No "Beam Energy" shows up in the GUI even after INIT, or only very old data

Try restarting the BeamDataServer using the "Wrench" icon on the toolbar. Wait for 2 minutes, then try INIT again.

ECALReadout or TDCReadout fails to INIT.

Try INIT several times.This will usually clear the failure.

HCALReadout fails to INIT.

Try INIT once and if this does not solve the problem, you will have to exit the DAQ and try again.

During the run, many "SLINK appears to be stalled" messages appear.

You will have to STOP the run and restart the DAQ. This cycle may have to be repeated more than once.

Very few successful spills are acquired, with many "Missing HTRs" messages.

One of the HTRs is missing triggers. This may be cured by restarting the program or power-cycling the crate. If the currently taken data does not require the presence of the malfunctioning HTR, you may want to make it inactive, using the "Details" window. Any changes made in the "Details" window will be overriden each time the DAQ restarts!

Please send an email to the experts with a copy of the several of the "Missing HTRs" lines.
 

To look at raw data from the HTR:

On the DAQ machine of course, you cd to:  cd /home/daq/drew and run a program that was originally written by Eric Hazen and modified by Drew Baden:

./htr_test -s [slot] -t [fiber]

This will printout a formatted text of what came on the fiber you specified.  It is basically just decoding Terry Shaw's data format.  If all's well you should see the CAPids rotating, and the Mantissa and Exponents of all 3 QIEs for all 4 CAPids behaving reasonably.  It should look like this:

  Data   Mant2  Exp2   Mant1  Exp1  Mant0  Exp0  C2 C1 C0
0x11803                                12     0   0  0  0
0x12832     25     0      20     0
0x12AAB                                21     0   1  1  1
0x12422     17     0      18     0
0x12B53                                21     0   2  2  2
0x12C26     19     0      22     0
0x125FB                                18     0   3  3  3
0x13020     16     0      24     0

Things to note:
 

Each word delivered is 16 bits, but there are 32 presented to the QIE.  So the format is:

1st 16 bits:  capid for all 3 and the data from QIE0 (1st QIE)

2nd 16 bits:  data from the next 2 QIE (2nd and 3rd)

The capid bits should all be the same, and they should rotate as above.

See Terry Shaw's data format documentation for details.