Software Tools for Debugging: There are only a few programs that are useful for debugging the LGD HV system. These programs are in E852sccs under the subsystem DAQlgd. Note the LGD must be using "local gates" to use these programs. See Hut Electronics and Cables
for a discussion of this point.
gDAQ: This is a program which flashes the laser and reads out LGD ADC's. It then prints out messages concerning the status of individual channels. gDAQ runs independently of the E852 DAQ. The user may specify many command line flags. A list follows.
gDAQ -h
Usage : gDAQ options
Options are :
-b# Begin Excluded section of detector in the col.
-e# End Excluded section of detector in col.
-g Graphical mode - HIGZ pictures.
-w Wait mode. Pauses between events
-t# Sleep mode. Sleeps
Notes: The exclude section is used if you wish to suppress messages concerning a group of columns. (This is often useful if a splitter section or single splitter board is missing.) A message will printed for each channel above the on threshold and for each channel below the off threshold. gDAQ is generally used during the turn-on phase of operations. After data-taking has begun, it is of limited usefulness.
testLGD: This program is used mostly for LGD maintenance during the data-taking. testLGD, like gDAQ, flashes the laser and reads out the LGD ADC's. It differs, however, in that it allows the user to examine a particular (read suspect, broken, or otherwise flaky) channel and it's neighbors. Typically after replacing a base during maintenance it is examined with testLGD and the high voltage setting is modified until its output is similar to those of its neighbors.
Turning On After a Down Period: Turning on the LGD HV system is non-trivial, however rumors of it requiring the sacrifice of one's first-born are mostly unfounded. Here's a somewhat chronological list of tasks which must be completed:
If the system has been unused for some time, it is good to verify the functioning of the DC power supplies and distribution before going any further. To do this:
Go into the Darkroom and unplug the cables running from the grey box on the darkroom roof to the three individual splitter sections.
Verify that the MG set and UPS are both on. Also verify that after opening the service disconnect to the input of the UPS the alarm goes off in the MPS control room. If not consult the MPS personnel. It is important to make sure this alarm is working.
Using the key reset one or both of the crash switches. (If you don't have a key consult an LGD expert to obtain one.)
Using the test point panel located in the LGD hut verify that all of the voltages are present from all the supplies. If they are not, find the problem and fix it before continuing. These supplies normally operate up to two or three volts above their labelled values to compensate for the voltage drop in transferring power out to darkroom.
Remove all of the ribbon cables from the splitting boards in the darkroom.
Plug the cables back into the splitter sections, and check the voltages at testpoints on each section. If any voltages is missing consult an expert for your next course of action.
Using the HV control code (msuhv) select the option to Enable/Disable splitter boards. Using this turn off all 15 splitting boards.
Using the ribbon cables with the aluminum test point box that is kept in the back of the darkroom, verify the presence of -24, +5, and +10 as well as the absence of +200V on the test points. If none of them are present check that the ribbon cable is properly plugged in (i.e. with the arrow on the cable facing you). If +200V is found, try and turn the appropriate board off again. If it still does not work, consult an expert.
Now use msuhv to enable all splitters, then verify that all voltages are present on the test point box. If not, consult and expert.
Having gotten the DC power straightened out and working you are now ready to try and bring up the bases.
Verify the light tightness of the darkroom.
Disable all the splitting boards via software.
Plug in ribbon cables in the darkroom. IT IS CRITICAL that the ribbon cables be plugged in the correct orientation. This means that the number written on the ribbon cable must be facing you as you plug it in.
Verify that the laser on top of the transporter is turned on, ready, and enabled.
Attempt to turn on one of the splitter sections by doing the following:
from msuhv select Turn On/Off by Groups and Idle the group attached to the splitter section you are working on.
Enable the splitting board.
Turn on the group.
Use gDAQ to verify that at least some of the expected channels came up. If you see nothing consider the following:
Does the laser flash when you attempt to do a pulse_laser? You should hear a ticking when the laser fires.
Do the gate leds on the ROC's fire?
If nothing came up: remove all of the ribbon cables from the splitting board you are working on and re-check the voltages on with the aluminum test point box. If they are okay, then plug in one of the cables at a time
and try turning them on. (Note that when you plug a cable in, the current surge will in all probability cause the splitting board's protective circuitry to kill the 200V output until it is re-enabled via software.) If things are still not working call an expert.
Repeat this procedure for the reset of the groups.
You are probably now faced with some missing channels. Here is a summary of a few known failure modes:
Absolutely NOTHING TURNED ON --- call an expert and pray I have forgotten something trivial when writing this.
Missing 1/3 of the Wall!!! (i.e. one entire splitter section is not working) Here's the suggested debugging technique:
Using the test points on the splitter section, check the voltages being supplied to that section. If any are missing there is probably a blown fuse in the grey box on the darkroom ceiling. DO NOT ATTEMPT TO WORK INSIDE THIS BOX WITH THE POWER ON!!!! Exposed 200V DC capable of sourcing lethal current is contained within the box. POWER OFF BEFORE SERVICING, then check the fuses.
If the fuses were okay, then check the heavy grey cable connecting the splitter section to the grey box on the darkroom ceiling. If this cable shows any sign of damage replace it.
If the above steps haven't fixed the problem call an expert.
Missing 5 Columns in a row!!!
Sounds like splitter board problems. Here's the rough guide for figuring out what may have gone wrong, but most of the fixes have to be left to an expert (i.e. Victor Bodyagin, Victor Bodyagin, or Victor Bodyagin).
Try unplugging the cables and checking the voltages with the splitting board turned on and with it turned off. If the 200V line does not come on then it's broken, call the expert.
The 200V line seemingly switches on and off correctly. With the cables still unplugged try using software to turn on the group of bases associated with the splitter. A common logic failure in the splitter boards will result in their switching the 200V line on and off wildly when you do this. If it does so call the expert.
Okay so the 200V line is steady and only changes when you tell the splitter to enable or disable. Now try adding one column at a time and check the voltages and that the columns you are adding keep working. If you can maintain 3 or 4 columns but not all five, it may be that the current limiting features on the splitting boards are set too sensitively. Call the expert.
ELSE call the expert
Missing a single column!!! Try the following:
Is it plugged in? Start at the splitting board ribbon cable connection and make sure the cable actually goes to the bases. Specifically, there is a connection in the middle of the cable which can be undone during base repair work etcetera. Make sure the polarity is correct here or else you can destroy the entire column full of bases and probably do considerable damage to the splitting board. The polarity is indicated by small arrows on both connectors. Make sure the arrows are aligned.
If the column is missing sometimes (meaning it comes and goes), call an expert immediately. In the past the has been do to poor soldering of splitting board connections.
ELSE call an expert.
Missing a single base or a bases scattered throughout the wall!!!
But enough of this happiness, now for the typically long and drawn out process of repairing bases one at a time. Here's roughly the way to procede:
If the ADC's have been thoroughly checked out assume the problem lies in the signal cable and/or the base. (See IU ADC Maintenance for information about how to do check the ADC's.)
Start off simply: verify that at the LGD the base number and cable number agree with the map. It may also be useful to check that the cable plugs into the correct ADC channel in the LGD hut.
If you find a discrepency in the cable number it is probably due to a pair of swapped cables. Figure out exactly what happened, find the missing cable: DO NOT JUST CHANGE THE MAP!
If you find a discrepency in the base number remove the base from the wall and verify the internal address. At your option replace the base with another or change the map and put this one back in the wall.
If he cable and base numbers check out then remove the base from the wall, it may be broken. If you have spare bases and are in a hurry you may want to find a replacement for the base in question before removing it. Click here for help in finding a spare.
The whole thing just @#$^&*(&^ turned off!!! Dunno. Call an expert. Prayers may be offered at this time.