Short description of the French DPDs and associated PMT/bases.
Some hints for Troubleshooting
This document describes some basic features and instructions relative to the French detectors and associated devices like power supplies. This report has to be read along with the report on NA detectors http://g0web.jlab.org/manual/documents/NAdetectors/howtonadetectors8-22-02.ps and on the CAEN HV
written by Julie Roche .Serge Kox, ISN-Grenoble
Latest update made by C. Furget : October 23th 2002
French Detectors
French FPDs
Description of the french FPDs can be found in the Grenoble http://isnwww.in2p3.fr/hadrons/hadrons.html#g0#dev_detect
and Orsay http://ipnweb.in2p3.fr/~rdd/G0/RDD/index.html Web pages.For those interested, reports on the tests of the 4 French octants with cosmic rays are available :
http://www.npl.uiuc.edu/exp/G0/secure-g0/docs/Cosmiques_rapport_F1.ps.gz
http://www.npl.uiuc.edu/exp/G0/secure-g0/docs/Cosmiques_rapport_F2.ps.gz
http://www.npl.uiuc.edu/exp/G0/secure-g0/docs/Cosmiques_rapport_F3.ps.gz
http://www.npl.uiuc.edu/exp/G0/secure-g0/docs/Cosmiques_rapport_F4.ps.gzThe cable path for the signal and High Voltage are being described in Julie's report. The mapping for the French detectors can be found in the first 5 pages of the following document http://www.jlab.org/~pking/G0/cabling/channel_layout_2002aug22.ps
French PMTs
The french PMTs are Photonis XP2282B/04 type. Technical information can be found in the report on the manual G0 web page http://g0web.jlab.org/manual/documents/FRdetectors/PMT_Relative_gain_measurement.ps
Relative gain calibration has been performed for all PMTs and a database is available on the Grenoble web site ( HTTP://larbookkeeping.in2p3.fr/G0_DB/). Also an internal report is available at http://www.npl.uiuc.edu/ftp/G0/detectors/config/Tables_HV.ps with the most recent tuning of the tubes for a gain of 107.
SPARE French PMTs are located in the French cubicle (10) in the left cabinet. Look at the database to find measured gains in order to replace the tubes with ones having similar characteristics. Also we will try to let a spare mechanics so people can have a look before changing a base or a tube !
The replacement of a PMT is supposed to be done by an EXPERT. It requires all the PMT of the octant to be turned off as a light leak will be produced. Also it is safer to turn off the LV power supply. Mechanically, first one should remove the EM copper shielding (one connection on the base and one on the back plate of the octant). Then one has to remove the 4 cables (signal, HV and 2 LV) from the base. The black backelite tube containing the base and tune can be dismounted at 2 levels. When you want to change the phototube then the optical grease has to be changed. The optical grease is available in the same location than the spare PMTs. This might then require to remove the complete Bakelite black tube. Clean the grease left on the light guide and remount the Bakelite tube back but this time without the part containing the base. Put some grease on the phototube and insert it in the tube with the base attached to it. Finally mount the second part of the Bakelite tube and turn it until you reach the end. On the base, the three screws should now be a little (few mm) off the aluminum plate (springs in the base are then in tension to obtain a good coupling with the optical grease). Then remount the EM shielding, not forgetting to connect to the two grounding point (plate and base) and then the cables (LV first, then you may have to rotate a bit the base to be able to plug both signal and HV cables).
When the PMT is installed first test to be done when changing a PMT is that no light leak is observed for it (use lower values like 1300 V for this test). When ramping up the voltage a flash will appear around 1100 V but this is only the crossing of the Zeener working point. Finally the fine tuning of the PMT gain has to be done with particles.
PMT HV power supply
The PMTs bases are powered with CAEN HV supplies (SY 403) located in the G0 cage. A technical instruction manual is usually available near these crates.
For the French PMTs the crates 32, 34, 36 and 38 are used.CAEN 16 channels cards (A 503) should be adjusted with HV hardware limits in order not to exceed 2250 Volts (knots located on the back of each card). The value of the maximum current can be controlled in the CAEN with the crate map menu. When a card is replaced, make sure that this limit is set at 2250V. Never try to set voltage above 2050 V to a French PMT (call EXPERT if this appears necessary).
For the forward angle measurements, the guessed gain to be used is about 2-4 106, so this corresponds to the HV value for a gain of 107 decreased by approximately 150 V. [HV(G/2)=HV(G)-100V)]
A file to be restored with the HV control program will be kept updated along with the NA detectors. Julie Roche and Serge Kox will try to take care of this and the file name will be : defaultg0.hvc
This file is not to be modified by other people. Backup is supposed to be made with other names and any changes in the values of French or NA PMT HV values should be notified (via mail or e-log reports).Typical current read by the CAEN power supply for a 2000 V setting is 600 µA (around these values, the relation is quite linear, i.e. a decrease of 10 % in voltage should translate into a decrease of 10 % in current). When settings are finalized, it is good to set a limit at about 10% above the read current. If you change/increase the HV values, be aware of checking that the new current may not be exceeding the limitation in current !! Also the sensitivity of the reading of the current should be the same for all cards
(to be checked when changing a 16 channels board, only by EXPERTS).
Changes of a HV values should be motivated (replacement of PMT, loss of more than 10% in gain, need to lower the current in bases ...). When beam development is going on or expected call an EXPERT to decide to turn Off the HV or use a set of lower voltage.
Report on G0-log and Notify EXPERT when changing a HV value. If you have any concern with values displayed or to be applied, call an EXPERT. Before restoring an old setting, check with BOTH EXPERTS (NA and French) that the values are satisfactory.
French bases
The technical report (G0-99-014) on custom made french bases can be found in the manual G0 web page http://g0web.jlab.org/manual/documents/FRdetectors/Embaseletter.ps .
They make use of active amplifiers which require a specific low Voltage power supply described in the next section. A failure of this amplifier prevents a good functioning of the base. Noise and a lower gain will be observed in such a case. If the failure is observed on a large number of bases, the low voltage power supply or/and the distribution boxes could be involved (call EXPERT in this case). Notify EXPERT or the RunN Coordinator if a base or a tube is failing so he can decide the best time period for repair. If this is happening during a no-beam period, call Immediatly the EXPERT for repair.
To change a bases, the procedure is the same than the one used to change a PMT. In this case however, only the second part of the Bakelite tube can de dismounted. PMT will often come with the base. Check the quality of the optical gel and decide whether it is necessary to add some grease.
SPARE BASES are located in the french cubicle (1) in the left (labeled) cabinet.
Low Voltage power Supply
For technical information, read the report available on the detector manual G0 web page, http://g0web.jlab.org/manual/documents/FRdetectors/PMTsupplyletter.ps
The location of this power supply is in Hall C, Rack HC01Z19 (near magnet controls, third blue rack from left). Bottom position in the rack. A brief description of the supply :
-- FRONT PANEL --
Button ON/OFF to be set ON
LED display +12V and -12 V for Voltages, can be checked with the LED display, switches set to U_V
LED Display 5.41 A (switch set to I_A) for the 12V and -5.05 A (switch set to I_A) for the -12V
These currents are with no beam and all octants connected (256 bases). Values might changes a bit (+/- 0.1 A) with temperature or with the beam. However if you observe only a fraction of the current (1/4 to 3/4 of 5 A), this might say that one or more octants are not connected. Then call an expert.Signals monitoring the current on +12V and -12V and to be sent to DAQ/ADC are available from 2 BNC connectors located on the bottom left part of the front panel. These cables are first sent to the patch panel located on the back of the rack HC01Z18. The cables arrive in the G0 cage on rack CH03C09 and are labeled I(12V) and I(-12V). From there they are sent to attenuators located near a the dedicated box used to fit the signals into a ribbon cable required by the Fastbus ADCs and located in the middle of rack CH03C04. These currents are in the data stream and are monitored in real time with the g0acm code.
-- BACK PANEL --
There are 3 SOCAPEX connectors available. Use LV1 and LV2 SOCAPEX cables on two first connectors (the third one is a spare, remove the fake connector to use it).
Selection of power voltage (EXPERT ONLY : 110 V or 220 V) has been done for JLab, i.e. is now set to 110 V.
Fuse for LED Display is located inside the supply (EXPERT ONLY)
In case of serious problem, Call French Detector EXPERT. A spare complete LV supply is available in the CLEAN ROOM.
SOCAPEX cables LV1 and LV2 are send on 2 distribution boxes located on the French octants 4 and 6. Two additional SOCAPEX cables are used to send LV to octants 2 and 8 respectively from these distribution boxes. For the first connexion of the SOCAPEX cables, in order to test the correct polarity, two LED lights (red and green) should light up on the distribution boxes. These LED are clearly visible only on octant 8 from the platform. LV is distributed from the box to each PMT by a daisy chain of small
bipolar cables. It's better to terminate the daisy chain on the box. In case of a failure in one base, the other ones should continue to work.Call French Detector EXPERT in case you want to stop the experiment and change the base. Spare bases are located at present in French cubicle (10) as well as spare bipolar cables and a test box for bipolar cables. The spare supply is located in the Hall C cage in the storage building with SOCAPEX cable and a spare connector.
Anode Current measurement
-- Picoamperemeter --
Available in Electronic room, second drawer of the desk. MUST use with 50 ohms in series (special box). EXPERT level required.
Allow to measure from patch panel currents from the bases or amplifiers (French and NA).
Turn ON. Select sensitivity (200 µA).
-- Modified French Bases --
All NA bases anode current can be measured. This is not the case for the French bases due to a Base Line Restorer associated with the amplifier. Thus for the french, only LV +12V and -12V and anode current for the 4FL and 16 FL FPDs of 4 french octants equipped with modified bases can be measured. Cables are connected in the main patch panel below the Ferris wheel and sent upstairs in the G0 cage (signal cable numbers : 513 to 520 FPD 4FL and 16FL, then octant 2, 4, 6, 8). From the patch panel, they are sent to a dedicated box to fit the signals into a ribbon cable required by the Fastbus ADCs and located in the middle of rack CH03C04. The signals are now measured after an x10 amplifier (green color, 8 channels) located in rack CH03C04. One of the two outputs is free and can be used to plug a lemo cable and check/monitor the current with a scope.
The input/output from top to bottom of the amplifier are respectively
2-04 FL
2-16-FL
4-04-FL
4-16-FL
6-04-FL
6-16-FL
8-04-FL
8-16-FLWith this amplifier, the sensitivity with a scope and under 50 ohms is a DC signal with 10 mV corresponding to an anode current of 2 µA.
-- With DAQ --
goacm program for both NA and french (real time).
Calculate current from pedestal shifts (NA/French) in Fastbus ADC spectra French (2+8 channels, Fastbus ADC 10 channels 88 to 95 for bases and ?? for LV)
Calibration is 10 mV/µA of DC level observed on the scope under high impedance and 5 mV/µA observed under 50 Ohm after the fast amplifier and 25mV/µA for the LV supply. At the ADC level, calibration is about 300 channels/µA for the anode (30 channels/µA if we remove the amplifier) and about 800 channels/A for the LV power supply.
GOAT has been implemented for the off-line analysis of these currents (ACM button).
-- CAEN snipper --
The eight Anode Current outputs will be used to kill the CAEN in case several PMTs have a current above a fixed threshold.
For that purpose the second outputs from the fast amplifiers discussed above are send into a modified eight channels discriminators, which is located in the NIM crate under the VXI crate (ROC3). The "VETO" output of this module has been modified to provide a NIM output of about 100 ms of width
when a given number of the eight discriminators channels have an output. This multiplicity can be changed and is actually tuned three and more.
The threshold of each discriminator has to be adjusted depending the limit on the anode current is choosen. The actual tuning is 60 mV readout
(corresponding to a true threshold of 75 mV) give a limit of 15 µA ( 15 µA *5 mV=75 mV ).
The NIM signal of 100 ms is send into the NIM kill input of the CAEN (each CAEN are daisy chained and terminated by a 50 Ohm adaptator.
Light leak PMTs on French Octants
4 Hamamatsu PMTs can be used to check large light leaks in French octants.
French Grenoble EXPERT only as the signal is now noisy and difficult to interpret.
HV to be set between 1100 and 1200 V. CAEN supply CRATE 39, channels 33, 34, 35 and 36. With these voltages, signals of 5 mV are observed for the single p.e. in the electronic room.
HV cable and signals cables originate from the back plate of the octants in the Ferris wheel
HV cables are connected to a panel located between the Ferris wheel and the magnet (16 channels are available, use 1,2,3,4 (channel 0 for laser) on the left side of the panel).
For the signals BNC cables, they connected to the main panel below the octants, channels 521, 522, 523, 524. In the electronic room, these cables are not sent in the data stream.
---> Use a scope to look at the signals. These information can be checked after some moves of the Ferris Wheel or after any incident which have involved the light leak boxes of the french octants.
French FastClear signal
Modules have been installed in the NIM crate located on the left side of the VXI crate (ROC3) in order to provide the "trigger" for French octants
used for the Fastbus. This signal has to arrive within the FastClear gate (avbout 100 ns every 3 µA) and is used for vetoing the
Fasclear signal (then accept the event).This kind of "trigger" is performed using the meantimer outputs which are send into the Fastbus TDCs. The modules four modules provide
eight multiplicity signals (each corresponding to 4 detectors Front and Back and for one detector and its buddy --> 16 meantimers) which send into
a discriminator. The multiplicity signal is 50 mV/hitted meantimer. A threshold of 30 mV is applied during standard G0 running. Specific tests
like cosmics requires a multiplicity 2 corresponding to a threshold of 75 mV. The sum of the disrciminator is send to built the "accept" accept.