EMC Slow Controls
Last Update July 21 1999
do I know that the alarm I see is real and not a monitoring artifact ?
Before we go into details, let me just say that any Real alarm will
have consequence on data taking (loss of efficiency in a group of channels,
increase of noise, etc). It is therefore always wise to check the fastmonitoring
if there is new data coming in to confirm the diagnostic delivered by the
Here's a list of know monitoring artifacts.:
And here's a list of known real alram symptoms
A single channel value dropping close to zero. This applies to all
types of channels, but especially to the low voltage monitoring (currents
and voltage).Be careful though, it could still be a real alarm. Check with
A single channel oscillation. This is definitely an artifact. Watch
it closely. If it's a temperature and it goes out of range periodically,
it should be masked out or fixed quickly.
A group of 8 channels oscillation. This is usually fixed by
switching of the gmb board power for a few seconds and switching it on
A current value of more than 10 Amps. This is definitely a monitoring
problem. The GMB full scale for current readings is around 10 amps. If
this happens: mask the alarm and contact the monitoring expert by e-mail.
All readings for one power supply drop to zero(V(-5V),I(+5,-5,10), except
the temperature reading. A power supply has tripped or is dead. GO to the
EH and try to switch it on again. If this doesn't work try replacing
the power supply.
monitoring display is blinking white, what does that mean ?
Communication was lost between the workstation and the Epics I/O controller.
This can be caused by network problems or loss of one monitoring
hardware device (GMB). Wait for at least 30 seconds before doing
anything. That's about the time it takes for the update of the sensors
value to be completed. If after a minute or two the display is still white,
Try cycling the GMB power in the EH (recipe).
display is full of white boxes with the "Not Connected" mention, what does
that mean ?
This is a network problem or the IOC has been rebooted.. In this case
only one thing to do. First make sure we do not experience afs problems.
This could cause the monitoring not to restart because the boot file is
located under afs space. Then go in the electronics house and reset the
IOC if it hasn't been done yet. You can see that it restarted when the
green LEDs on the CAN module start blinking again. This might take a while.
I start the monitoring
display by typing ~paf/mon.sh but it dies right away with an error message,
what do I do ?
If netscape is running, kill it. Restart netscape with the -install
option. restart the monitoring.