PVSS trigger clock monitoring

Accessing the panels

The PVSS project can be accessed connecting to the alitrion001

  • from the consoles in the ACR, type rdesktop -a16 -g1280x1024 alitrion001
  • user dcsoper, password (ask the DCS shifter or check on the printed version of this guide)

The same panel can be accessed from the DCS console, as main panel for the TRI_DCS node.

  • Trigger main panel:
    main.png

Panel description

The upper left part describes the clock mode (LOCAL or BEAM1 and the BEAM MODE).

  • Clock monitor:
    clock_monitor.png

When the clock is LOCAL:

  • the measurements and the plotted values have random oscillations, this is normal
  • just check that all the machinery is working and be ready whenever a clock transition is announced

  • Clock transition:
    clock_transition.png

When the clock is BEAM1:

  • it has been observed that the clock jitter appears when the beam mode is RAMP; if it's not fixed, the jitter usually stays present in the next beam modes (FLAT TOP, SQUEEZE, ADJUST and STABLE BEAMS).

The upper right bozes show the bits for the different CTP clocks (BC1, BC2, BCref, BCmain).

  • the error bit should always be green
  • the locked bit is green when the clock is locked correctly and orange if it gets unlocked. Unlocking should happen on clock transition and adjustment, and should lock again after 10-20 s (a historical plot is accessible clicking on the QPLL bits button)
  • the TTRrx bit should always be green

The bits status is logged and CTP experts can check at their convenience.

  • QPLL chip monitor:
    qpll_bits.png

The oscilloscope measurements

The small table on the center of the panel shows the measurements done in the scope:

  • the time difference between BC1_fanout1 and BCmain_fanout3 clocks
  • the time difference between BC1_fanout1 and BC1_CORDE clocks
  • the time difference between BC1_fanout1 and BCmain_fanout2 clocks

The values are measured every 10 s in BEAM1 clock, 50 s with LOCAL clock.

The second row shows the standard deviation (updated every 2 minutes):

  • the boxes are grey during LOCAL clock (these values are irrelevant)
  • in clock BEAM1, the values are OK if they are smaller than 1e-19 (the boxes are white)
  • in clock BEAM1, the values are NOT_OK if they are ~1e-17 or 1e-18 (the boxes are orange)

  • Oscilloscope values:
    scope_values.png

The two plots show:

  • the three measurements described above
  • the clock mode and the beam mode with numeric codes (a legend is placed at the right)

The trend buttons on the right will open a new big panel with all the variables in the same plot.

  • Plots:
    2plots.png

Recognizing a jittering clock

Clock behaviour:

  • during LOCAL CLOCK the measured clock timings oscillate with an amplitude of 25 ns (two oscillates - blue and red curves, while the black curve is always constant); the standard deviation is ~1e-15 or 1e-16 and the boxes are greyed out
  • during BEAM1 clock the three measurements are very stable:
    • the measurements shown in the table are around 1e-8 - 1e-9
    • the standard deviation is around 1e-19 or 1e-20
  • the clock is jittering when the clock mode is in BEAM1 and there are oscillations of an amplitude of ~2.5 ns
    • the measured values show a standard deviation of ~1e-18 and the boxes are orange

  • Clock jittering:
    clock_jitter.png

Troubleshooting

In order to provide correct data, the oscilloscope alidcsaux008 must:

  • be opened in rdesktop, in an active workspace, with the correct geometrical constrains (1280x1024)
  • the Infiniium application (the scope monitor) must be running

If the application is stuck, or the oscilloscope is not responding anymore, the project will not update the measurements anymore. In this case, you'll have to click on the emergency button on the upper right part of the panel. A new panel will be opened, giving you access to the emergency procedures.

  • Emergency button:
    emergency_button.png

Try following the first procedure:

  • first kill the application (even if it's already dead on the rdesktop, the process might still be alive in the background)
  • then try to restart the application

  • Emergency recovery of the oscilloscope:
    scope_recovery.png

Wait for ~1 minute and decide if the monitoring is working again or not.

If not, go to the second procedure, and try to reboot the oscilloscope completely. The booting takes 2-3 minutes at least, but the Infiniium application should restart automatically. After 3 minutes you can try to connect with rdesktop to the scope again (command scope, or rdesktop -a16 -g1280x1024 alidcsaux008).

If the reboot is impossible (problem with the network connection or oscilloscope not alive anymore), the last resource is to power off and on the whole rack. The rack is controlled by the DCS shifter: ask him/her to open the SERVICES/RACK and click on the rack C21.

  • Rack control on the DCS panel:
    rack_control.png

-- OmbrettaPinazza - 10-Nov-2011

Topic attachments
I Attachment History Action Size Date Who Comment
PNGpng 2plots.png r1 manage 27.0 K 2011-11-10 - 19:23 OmbrettaPinazza Plots
PNGpng clock_jitter.png r1 manage 20.1 K 2011-11-10 - 19:02 OmbrettaPinazza Clock jittering
PNGpng clock_monitor.png r1 manage 1.9 K 2011-11-10 - 18:14 OmbrettaPinazza Clock monitor
PNGpng clock_transition.png r1 manage 2.1 K 2011-11-10 - 18:26 OmbrettaPinazza Clock transition
PNGpng emergency_button.png r1 manage 1.1 K 2011-11-10 - 18:16 OmbrettaPinazza Emergency button
PNGpng lhc_cycle_ramp.png r1 manage 12.5 K 2011-11-10 - 18:42 OmbrettaPinazza LHC cycle, beam mode RAMP
PNGpng main.png r1 manage 69.6 K 2011-11-10 - 18:14 OmbrettaPinazza Triger main panel
PNGpng plots.png r2 r1 manage 27.0 K 2011-11-10 - 19:22 OmbrettaPinazza Plots
PNGpng qpll_bits.png r1 manage 2.1 K 2011-11-10 - 18:15 OmbrettaPinazza QPLL bits
PNGpng rack_control.png r1 manage 55.0 K 2011-11-10 - 19:48 OmbrettaPinazza Rack control on the DCS panel
PNGpng scope_recovery.png r1 manage 18.8 K 2011-11-10 - 18:16 OmbrettaPinazza Emergency recovery of the oscilloscope
PNGpng scope_values.png r1 manage 3.7 K 2011-11-10 - 18:15 OmbrettaPinazza Oscilloscope values
Edit | Attach | Watch | Print version | History: r4 < r3 < r2 < r1 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r4 - 2011-11-18 - OmbrettaPinazza
 
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    Main All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright &© 2008-2024 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback