Jefferson Lab > Physics > Detector Group
Privacy and Security Notice

Detector Group: Technical Notes

The detectors often put together relatively unpolished documents about their work. These documents are informal, and are not published. Below is a list of links to abstracts for the documents. If you wish to obtain a copy of the original paper(s) please email Stan Majewski ( or stop by his office in CEBAF Center, room C104.

Study of PMT sensitivity to X-rays

Date: 9/5/94


Response of different PMTs to X-rays and electrons was tested in search of solutions to make the arcing sensors installed in the accelerator RF cavities and based on 931B side window PMT insensitive or less sensitive to X-ray bursts, for example accompanying beam losses. The results show that all PMTs tested have similar sensitivity to X-rays and unless better discrimination by electronics vs X-ray (bremsstrahlung) bursts is possible, the only other technique available is to screen PMTs from X-ray radiation, for example by moving them away from the RF cavities and using light guides for optical coupling.

Studies of gain stability and linearity of XP2050 Photomultipliers

Date: 8/15/94


Two 5" XP2050 PMTs used in the Hall A shower counters were tested for gain stability vs load changes and for pulse linearity with a simplified linear CEBAF-type voltage divider. Both PMTs show good stability and linearity with this base. Fast amplifiers shall be used to compensate for a rather low gain of this 10 stage PMT in the shower counter application but primarily to prevent that PMTs are operated at too high average currents.

Test Results of the Hamamatsu Ultra Compact R5600U Photomultiplier Tube

Date: 2/23/94


The Hamamatsu R5600U is a compact photomultiplier tube in an OPTO-8 package whose characteristics are outlined in a preprint of a paper submitted for publication by Hamamatsu and in their product data sheet. The bialkali photocathode size is 8 mm in diameter and the outer diameter of the PMT is 16 mm. The PMT also comes in a slightly smaller package (R5600) where they removed the insulating enclosure over the outer kovar metal shell which is electrically at cathode potential. The compact "metal channel" eight stage dynode allows for a very small package. With a gain of 3 x 10^5 at 800V and a peak quantum efficiency of 15% (at 400 nm) the R5600 is lacking when compared to most PMTs with bialkali photocathodes. However its compact size may make it the ideal photo-detector for a particular application. The R4600 is being considered for the start counter in the CLAS detector of Hall B where only a very small PMT could fit.

All of the tests performed were in pulse counting mode with a negative biased cathode potential. Typically the output of the PMT was amplified with a fast amplifier before being connected to a CAMAC ADC (either a LeCroy 2249 ADC or Phillips 7166 QDC). The CAMAC data acquisition was done with Kmax software running on a Macintosh IIFX computer.

Since the E5780 voltage divider base assembly supplied by Hamamatsu does not have the high voltage ground and the signal ground common I needed to connect the two grounds with a short lead to eliminate any ringing on the output pulse. Hamamatsu reports that the risetime, falltime and pulse width are about 650, 1050, and 1500 psec respectively. Hamamatsu also claims a transit time spread of 280 psec FWHM at 800 V with a single photoelectron event. In addition it is not surprising that the response time of the PMT is 0.65 ns.

Highly Stable PMTs-news

Date: 2/20/94


New information became available about highly stable PMTs. The Detector Group is planning a study to finish our previous PMT tests.

CLAS Start Counter

Date: 2/20/94


The Detector Group is proposing to help with a study of a CLAS Start Counter. Two detector techniques are discussed.

Highly Stable PMTs

Date: 2/7/94


Inexpensive and slow box & grid type PMTs such as Hamamatsu R268 and Thorn/EMI 9258 are found to be very stable under changing average anode load. This fact may have some importance in some applications at CEBAF, for example in monitoring systems where gain stability is required.

X-ray Mini-Probes for the Digiray Setup

Date: 12/18/93


Preliminary results of the tests performed with a new X-ray miniprobe made with small-size lead-doped (5% and 10% Pb by weight) BC452 scintillator samples attached at the end of about 1 meter long plastic and liquid lightguides show that the probe produces surprisingly good quality images.

Embedded Waveshifting Fiber Readout of Long Scintillators

Date: Oct 93


We have tested a method of reading out light from long scintillators by embedding waveshifting fibers in longitudinal grooves created in the scintillators which are then connected to a photomultiplier (PMT). With PMTs on both ends of the fibers we have obtained attenuation lengths of over 5 m and light ouputs of over 5 photoelectrons per MeV from both ends of a 4 m long BC408 scintillator with five embedded 2 mm diameter BCF92 fast green waveshifting fibers using a Hamamatsu green extended R580-17 PMT. With a reflector at one end of the fibers, we have easily obtained over 10 me attenuation lengths. Also presented are results using inexpensive acrylic scintillators with this method and measurements of a new experimental scintillator from Bicron with over a 5 m attenuation length.

Dark Counts Tests on XP4512b

Date: 9/10/93


The Phillips XP4512b was tested for dark current noise counts for various thresholds. The serial number of the PMT was 9173. A Phillips type VD105B negative bias base was used as the voltage divider. The PMT was biased to -1580 V which according to the Phillips supplied test ticket corresponds to a gain of 5 x 10^6 and a dark current of 12 nA.

Individual Photomultiplier Monitoring Using Small Scintillating Crystals

Date: 7/29/93


To monitor the gain of individual PMTs little buttons of plastic scintillators coupled with a radioactive source have been glued to the PMT window outside the useful central photocathode area or to the far end of the lead glass block or to the lightguide, etc.

Study of a FEU-115M Photomultiplier, Part III

Date: 7/22/93


The results of more additional measurements performed with fast 30 mm diameter FEU-115M PMTs are presented. Pulse linearity, dynode readout and timing properties were studied.

Studies of rate stability of XP2282B and XP3462B Photomultipliers

Date: 6/28/93


Samples of two types of PMTs used in Hall C were tested for gain stability vs rate changes and for pulse linearity. Both PMTs show gain shift effects seen also before in all other PMTs tested recently at CEBAF.

Study of a FEU-115M Photomultiplier, Part II

Date: 6/25/93


The results of several additional measurements performed with a fast 30 mm diameter FEU-115M PMT are presented. Rate behaviour, double-pulse performance, noise, and effects of external magnetic field were studied.

Study of a FEU-115M Photomultiplier

Date: 6/12/93


We had an opportunity to test two Russian PMTs: a fast 1" PMT FEU-115M and a new 2" PMT FEU-184 "Uspekh" (= "Success"). This report deals only with the first tested 12-stage device which was found to have a fast 2 ns rise time, uniform amplitude response over a 13 mm diameter photocathode and a good QE especially in the green spectral region (20% @ 500 nm).

Linearity Studies of Photomultipliers and Voltage Dividers for the CLAS Calorimeter

Date: 6/10/93


Two types of 2" photomultipliers (XP2262 from Phillips and 9954A from Thorn-EMI) were tested with about a dozen different voltage dividers. Two XP2262 and three 9954A PMTs PMT/divider sets were optimized for pulse linearity and tested for stability vs pulse rate (or average anode current load). The best results for XP2262 were obtained with the standard CERN 4244 base while 9954A performed best with a modified UNH/TOF base. Pulse linearity up to 200 ma (100 mA) was obtained for XP2262 (9954A) within +/-1%. The main problem detected which cannot be eliminated by a divider choice is the drift of photomultiplier gain accompanying each change of average count rate, even at the current levels as low as 1 micro-A. It confirms the need for a strict monitoring system of PMT gains in the experiment, especially in the case of rapidly changing ("flickering") average count rates. Both effects have origin in the intrinsic limitations in operations of presently available photomultipliers.

Evaluation of Linearity of PMT Bases for the CLAS Calorimeter, Progress report

Date: 3/27/93


Linearity was evaluated for a Thorn/EMI 9954 PMT with optical pulses produced by a green LED. LED light pulses were attenuated by a set of neutral density filters placed in a remote controlled filter wheel. The pulse range covered was a factor 1000. The detected signal was adjusted in the 5-10000 pe's ranges with about uniform illumination of a PMT entrance window. Linearity was measured versus the measured signal expressed in photoelectrons and for several signal amplitudes as measured on a 50 Ohm load.

Scan Test Results on Three 5 Inch Diameter Photomultiplier Tubes

Date: Sometime 2 years ago


Three different 5 inch diameter PMTs were tested for possible use in Cherenkov detectors. The three PMTs tested were the Hamamatsu R1584, the Phillips XP4512 and the Burle 8854. Three parameters of the PMTs were measured for different positions on the PMT face: 1) the variation of single photoelectron (SPE) amplitude, 2) the variation of relative SPE efficiency and 3) the time shift. In the SPE mode we were able to separate the variation of relative gain (amplitude) across the PMT face and variation of the photon detection efficiency (efficiency). The variation of the photon detection efficiency is a result of the combination of the variation of the photocathode quantum efficiency and the variation of the photoelectron collection efficiency as a function of position on the PMT face.

All PMTs were tested in pulse mode using a UV/blue spark pulser coupled to a square 2x2 mm plystyrene optical fiber with a 1mm pin hole mask. The pickup pulse from the spark gap was used to generate a 80-100 ns pulse for the LeCroy 2249A ADC to obtain pulse amplitude spectra and the LeCroy 2228A TDC to do time shift measurements. The stop pulse for the TDC was generated in a 300 MHz Phillips 708 leading edge discriminator with a threshold at about 1/5th the average amplitude for the amplified SPE pulses. In all measurements a fast Phillips 777 amplifier was used. The positioning of the fiber was achieved by an X-Y scanner controlled by the Macintosh computer running a CAMAC based data acquisition called Kmax.

Flexible Liquid Lightguide

Date: 11/18/92


Continuing with some thought from the memo of 11/12, it might be interesting to consider a liquid lightguide "knee" inserted between two straight pieces of a cladded fiber/cane/rod of the same diameter that is being tried right now by the CLAS shower group (5 mm or more). I t might be both time saving and economical to bypass the bending procedure while retaining all the other advantages of the fiber readout (speed, light, output, good attenuation length). The added advantages of flexibility are obvious during assembly when mechanical tolerances are relaxed because flexibility of the lightguide allows for compensating any errors, shifts, etc contrary to the plastic lightguide case when flexibility is decreasing substantially with the increasing diameter of fibers or rods.

Test of the New Yerevan Grooved Scintillators

Date: 11/9/92


We have finally received the six pieces of about 2 meter long Yerevan scintillators produced with grooves already in the extrusion process with the help of the special attachment designed and made at CEBAF and used by Stepan in the extrusion process.

These pieces have the "old" uniform distribution pattern of grooves with a step of 16.7 mm between fibers. This results in about a 5% drop in signal at the scintillator edges but it is not that important in this first evaluation of this unique grooving technique developed in Yerevan.

The production procedure in Yerevan was this time intentionally simplified in order not to go through the regular purification steps to test the advertised by us immunity of our method to the attenuation quality of the plastic. We have already demonstrated in the past that a poor quality acrylic scintillator with attenuation components of 105 and 130 cm wen used with WVS fibers was showing much longer attenuation length characteristic of the BCF92 waveshifting fibers used. This was the second attempt to confirm this advantage of the WVS fiber technique. All six received pieces were grooved. The quality of the groove surface was found to be good and the same as of the flat surfaces.

Comments on Time Resolution of the CLAS Shower Counters

Date: 11/8/92


In this note we addressing the issues of the test methodology of the timing resolution of the scintillator strips for the CLAS shower counters and of the comparison of the performances obtained for the two fiber readout methods.

First it is our belief that the most appropriate method to test timing resolution is with minimum ionizing particles either using cosmics (as we were doing) or using electron sources (as the INFN group is doing). From collected information and from our own measurements, the tests done with a nitrogen laser or with 1 MeV Bi207 electrons underestimate time jitters observed in the realistic conditions when energy deposits follow Landau distribution and not simple (and much narrower) Poissonian or Gaussian curves. Also, there is a possible geometric effect due to the point-like energy deposition in the case of Bi207 electrons vs. the cosmic test conditions with a larger active area involved. In the laser case the first step in the energy transfer chain (plastic to primary fluor) is bypassed, therefore "improving" the time jitter. Because it is advantageous to use lasers or sources in evaluation procedures consisting of many repetitive measurements, the correction factors are introduced to correct for the above mentioned underestimate.

Preliminary Amplitude and Time Response Scans of an XP4312B PMT

Date: 11/4/92


We have tested a Phillips 3" PMT type XP4312B, SN# 5027 form uniformity of amplitude and time response by performing two linear X and Y scans. The axi were selected arbitrarily with the only requirement that they do not coincide with the evaporated conductive strips on the inner side of the entrance window. HV was set at -2.5 kVolt and the PMT was pout in the mu-metal shield extending about 1" beyond the PMT surface. Potential of the mu-metal shield was left floating.

We have used out UV/blue spark gap pulser in the tests. It produces short, under 5 ns, pulses at a 70 Hz rate. Light pulses were injected by a 2 mm square clear fiber attached to an X-Y computer-controlled scanner. Standard LeCroy ADC and TDC were used int he measurements. TDC start pulse and 80 ns wide ADC gate were provided by a pickup pulse from the spark gap. Constant fraction discriminator was used to time stop PMT pulses.

Comparison of Waveshifting Fiber Readout vs Clear Fiber Readout of Yerevan Scintillators

Date: 10/19/92


We were asked to make some additional comparative measurements of the Yerevan scintillators plus we performed more timing studies to resolve the issue of timing resolution of the waveshifting (WVS) fiber readout.

New Results with the Tagger Prototypes

Date: 10/14/92


We were asked to look more into the possible solutions of tagger counters with the flexible fiber type readout. This short note presents some new measurements performed recently after a long waiting period for samples with "the right" scintillator/waveshifter combination. As a byproduct of out studies of the waveshifting fiber readout of long scintillators, we have tested several new combinations. This study is not yet finished, but some of the results are interesting enough to make this partial report.

Embedded Waveshifting Fiber Readout for the CLAS Shower Counters

Date: 10/13/92


We are proposing that the embedded waveshifting fiber readout method is considered for the sectors 5 and 6 of the CLAS shower detector. The results of the R&D effort which is in its final phase, demonstrated the advantages of the technique over the "standard" techniques. the few remaining technical questions are still being studied right now. However, because of the pressing time with decisions for the CLAS we have decided to open the discussion about the use of this technique at CEBAF prior to having these final issues settled.

This novel technique offers improved operation at a reduced cost and with simplified mechanical solution over the clear fiber lightguide technique. The main original complication of grooving was solved by developing the integrated groover allowing for a fast and well controllable production of good quality grooves. Also, the Yerevan group is testing production of grooved scintillators during the extrusion process with the (made at CEBAF) attachment fixture to their extruder. In this paper, we present only a selection of our results to justify our claim that the technique became practical. Two papers are in preparation and will be ready in about one to two months from now.

Update on the single photoelectron response of mesh dynode PMTs

Date: 8/12/92


We have tested two Hamamatsu mesh dynode PMTs, the 2-inch R2490-05 and the 3-inch diameter R2238. Both were tested for single photoelectron (SPE) response. As mentioned in the 6/7/92 memo, we hope that the 5-inch PMT of this type might be a possible solution to the well-know problem of high sensitivity of regular PMTs of this size to magnetic fields. The particular application in mind is the Hall B Cerenkov counters. The mesh dynode PMTs have a well-documented immunity to magnetic fields, including in the direction transverse to PMT axis. However, there are two problems to solve: (i) 5-inch mesh dynode PMTs are unavailable at this time, and (ii) Single photoelectron response is not expected to be particularly good in the mesh dynode structures. The former problem can be solved, as Hamamatsu is producing 5-inch mesh dynode position sensitive PMTs and a modification (actually, a simplification) of their standard design is possible.

Further Crosstalk Reduction in Hamamatsu R4760 Multi-channel PMT

Date: 4/6/92


As a follow up to the Detector Meisters memo dated 3-9-92, concerning the reduction of crosstalk in the Hamamatsu 4760 PMT, we have explored the various techniques used to shape the pulse seen on the crosstalk channels.

In our setup cannel 6, an anode that is near the center of the PMT is set up to receive ~10 (on average) photoelectron (10PE) pulses. Channel 4, an adjacent cannel, is covered so that it receives no illumination. Also other channels are covered. The signals are measured directly from the anodes of the R4760, at -1.1KVolt bias, using a fast 400 MHz scope with a 50 ohm input impedance.

PMT Base for use with the BURLE 8854 PMT

Date: 4/6/92


This is a brief memo concerning the Burle 8854 PMT. The detector group was asked to assemble a PMT base for the Burle 8854 using a printed circuit board that was designed by George Lolos.

Tests of operational inefficiency of the CLAS drift chambers in the end-plate/feedthrough region-part I

Date: 3/5/92


The Detector Group was asked to help with the measurements of the width of the inefficient region in the CLAS drift chamber close to feedthroughs by using a highly collimated beam from our X-ray generator (as advertised by us for some time). A short description of the experimental method and results obtained is presented. Mac Mestayer was participating in the preliminary tests and gave us the test voltage combinations for the measurements.

Reduction of Ringing in a Fast Sidewindow PMT

Date: 2/8/92


The Detector Group was asked to clean up the signal of a sidewindow PMT (see memo dated 2/8/92) As a continuation of this effort the ringing was greatly reduced and the results are presented in this short report.

Memo on Waveshifting Fiber Readout of Long Scintillators

Date: 11/13/91


This is a follow-up to my two previous memos dated 10-17 and 10-27 on scintillator readout of long scintillator pieces with waveshifting fibers imbedded in grooves. Along with further discussion of the idea and of some new concepts, first experimental results of a small test counter are shown.