Forward Hadron Calorimeter

Testing Manual

I've tried to include in this web page documentation on the testing of modules in the physics high-bay.  Some parts are kludgy and others will make you cringe I'm sure, but if you see anything wrong, please let me know so I can fix it.  Good luck!

Orientation:

Here is a simple diagram of the high bay area.

There is a curtain between our working area and that of the TEC to the west and there is a small room (in blue) that has been used by the MVD.  The two pallets on the east side of the hall are those that are being tested, while the other pallets have already been tested.  There are a set of RG58 cables (with LEMO ends) and a thin blue ten line cable that run from the computer area to the testing pallets.  This is how the room should look when you first arrive to test a brand new stack of modules.

Beginnings:

First you should write down the 3 digit number of each of the modules in the stack (the white numbers on the opposite end as the light guide).  Draw a diagram that makes it clear which modules are on which layer and place it in the logbook.

Then you need to find a new set of phototubes to attach to the testing modules.  There may be a box or two already in the high bay.  If not, they are found in the basement of the physics building -- Rob Pisani (x5301) knows where.  There should be a stack of already tested phototubes in boxes that are clearly labeled.  Don't use those.

Unpack the phototubes from their packaging and aluminum foil wraps and attach them to the light guides on the testing modules.  Be careful with the phototubes because they have optical cookies attached to the face of the phototube to improve the optical contact between the light-guides and the PMT.   Recall, we're only trying to test the basic response of the tube and the attenuation of the modules.  Therefore a very tight light seal is unnecessary.  There are aluminum shims in a plastic bag that I have used to attach the phototubes to the light-guides.  I use two shims to hold the phototube on the end of the light guide with some electrical tape.  If you put the shims on top and bottom, it's straight forward to attach the tube so that it doesn't droop on the end of the light guide.

Finally, wrap the PMT-light-guide interface with aluminum foil to decrease the light-leakage.

Cable it up:

The testing modules are stack in layers.  There are two layers on each pallet and each layer is oriented perpendicular to the layers above and below.  For example, if layer one is oriented north-south, then layer two is west-east, layer three is north-south, etc.

There are two outputs on each PMT.  The one labeled 'D' is the discriminator output.  The other one (labeled 'A'?) is the analog output.  Connect the cables to the latter.

There are a series of labeled cables running from ADC's and fan in-fan outs.  Those labeled 1A, 1B, 2A, 2B, etc are trigger cables while those labeled 0, 1, 2, 3, etc. are readout cables.  Here's an example of an appropriate cable layout:

In this configuration, any combination of trigger pairs (1A-1B, 2A-2B, 3A-3B, 4A-4B) will correspond to a cosmic that passes through one of the modules in layer 2.

In order to test all 4 layers, we will need to re-cable the stack 4 times.  Theoretically we could cable all of them simultaneously, but it required a much more complicated trigger logic and more NIM and CAMAC modules than we could scrounge in short order.

There is also a thin, blue 10-line cable coming from the NIM crate that is used to power the phototubes.  Be sure you orient the PMT powerline appropriately in in the connectors on the blue line (line up the arrows!).  This line is flakey and, while it can handle powering 18 PMT's simultaneously ... well ... it wouldn't come close to passing a review in some experiment.

There are a couple saw horses which you should place along the phototubes so that, when you drape the tarp over the stack, it relieves the stress on the light-guides and PMT's.

Be sure you write down the serial number on each PMT that you are connecting to each module.  There tend to be two number on each PMT.  I've kept track of both of them.  If there is no number on the tube, add a label to it in the format: UL080201 -- UL = UnLabeled, 08 = month, 02 = date, 01 = the first phototube.  (If you find a second unlabeled phototube, add the label UL080202, etc.)

Power it up:

There are two power supplies above the NIM crate that supply the PMT discriminator threshold and the the Cockroft-Walton control.  The first we usually set at -.5V.  It really doesn't matter what you set it at -- if the voltage is too low, then the on-board electronics in the PMT base will suck a lot of power over that blue line -- that's bad.  So we set it at some reasonably high number since we're not looking at the discriminator output anyway.

The Cockroft-Walton voltage should be less than 10V.  I tend to run the system between 8 and 8.5 Volts.  We've run into trouble in this system with cross-talk between the trigger PMT's when we increase it to 10V.  I don't think it makes a difference in the analysis, but it's just easier to run it at a lower voltage.

Also turn on the NIM, CAMAC, and VME crates.

Debugging:

Now you should be seeing triggers in the format of a blinking red light on the gate generator.  You should be able to look at all 4 combinations of triggers by simply toggling them in and out on the 4 input coincidence unit.  Try looking at the rate of all four triggers and make sure they're reasonable.  Often one trigger pair is especially high or non-existent.  In this case you'll want to take a look at the signals on the digital scope (take copies from the fan in-fan out and the associated discriminator) and tweek the threshold on the discriminator until it looks good and you have a reasonable trigger rate (~Hz) for each trigger combination.  In some cases the tube may be flakey and you'll need to replace it.

When you're done debugging, be sure to toggle all triggers in.

Starting a run:

Now you're ready to start taking data. 

Now a run has been started.  You can type 'Status' in the minicom window to check the status of the run.

Analyze a run:

In the ~calo/test1/ directory there is a program called afterburn.cc which includes all of the code for reading in the data, creating a series of histograms and an nutple, and analyzing the results to extract the attenuation.  It's still a little rough around the edges and, if you're so inclined, feel free to update the code -- just let me know what you did so I can keep up.

There is some documentation for this code here.  Hopefully it's straight forward.

At the beginning of a run you'll want to take a look at the Calo nutple the afterburner module makes to make sure all of the phototubes all acting reasonably.  You can watch the data as it comes in by opening up the et pool (poncsetopen("/tmp/Maincalo") instead of the datafile (poncsopen("/calodata/run_005xxx.evt").  In this case, the ntuple will be continually updated as events are taken.

If there are any flakey tubes you may need to stop the run (daq_end), turn down the voltage, and take off the tarp to replace a tube.

If everything looks reasonable, then you'll need to wait about 4-6 hours till there are enough statistics to be able to run pend(xxx) (see the afterburner documentation) to see the attenuation.  Again, the code isn't perfect and there need to be some tweeks.  Some of the fits don't converge which causes unreasonable attenuation values.  The most important thing is to convince yourself that the data for all 10 tested tubes in each run is reasonable.  We can fine-tune the analysis as we go.

Rinse and repeat:

Once you've done the first run, stop the daq (with daq_end in the minicom window), turn down the voltage, take off the tarp and move the readout cables to the next layer.  (Don't forget to move the power cable too.)

Repeat.

When the second run is complete, switch the position of the trigger and readout cables and repeat.

etc. etc. etc.

And now?:

When all four runs are complete, close the port of the daq on minicom with the daq_close command.  Then power off the NIM, CAMAC, and VME crates as well as the power supplies above the NIM crate.

Remove the tarp and make sure all of the modules and phototubes are clearly documented in the logbook.  Further verify that it is clear which layer (and, therefore, which modules and tubes) were tested in each run.

Send an email to Al Pendzick asking him to schedule the riggers to bring in a new stack of modules.

Remove all the cables and phototubes.  Cover the ends of the phototubes with aluminum foil and wrap them nicely in a box.  Be sure to label the box as 'tested phototubes'; include the date and number of PMT's in the box.

Move things out of the way for the riggers and wait for the next set of modules to arrive.

Thanks for your help.