GAP Passive and Active PMT Biasing Networks II. S. Argiro, D.V.Camin, M. Destro and C.K. Guerard

Size: px

Start display at page:

Download "GAP Passive and Active PMT Biasing Networks II. S. Argiro, D.V.Camin, M. Destro and C.K. Guerard"

Willis Smith
5 years ago
Views:

1 GAP Passive and Active PMT Biasing Networs S. Argiro, D.V.Camin, M. Destro and C.K. Guerard Dipartimento di Fisica dell' Universita degli Studi di Milano and Via Celoria 16, 2133 Milano, taly Abstract n this note we examine the change of the PMT pulse gain as a function of DC bacground current using an active divider and compare its behavior to the one obtained using a passive networ. We show that the present version of the active divider results in: 1) constant gain over a large bacground current (rendering possible to operate the FD under the presence of some bacscattered moon light, and other sources of unpredictable occurrence), and 2) in a factor 1:9 lower power consumption. These results are also shown to be independent of temperature. 1 ntroduction. The bacground current in the PAO FD PMTs will depend collectively on the sy brightness, and individually on the particular section of sy viewed by a single pixel. Data from dar nights in Utah [1] indicate that 4 photons m?2 s?1 deg?2 will be arriving to the eye. Assuming an unobscured diaphragm area of 1:54 m 2, a pixel eld of view of 1:5 1:5 deg 2, and a cathode eciency of 2 %, this value will translate into 2:7 phel (1 ns)?1 or a DC current of 4.3 pa. This bacground could increase by a factor of 3 due to the presence of bacscattered Moon light [2], or even by a factor of 1, due to the Aurora and other UFO [3], maing it impractical, in this case, for the FD to operate. n the October 98 CERN meeting [4] it was agreed to operate the PMTs at relatively low gain, 5 1 4, in order to ensure a PMT life span of at least 15 years. Moon light would mae the DC anode current vary from a minimum of 215 na to 6:5 A. Other sources passing through the FOV of a particlular pixel [3] could originate currents which could easily reach a few tens of A and much more. Recording the DC anode current would give valuable information about each PMT history as well as statistical information on the behaviour of the 15, tubes to be used in each site. A current monitor system based on optoelectronics was developed for this purpose [6].

2 t would be highly desirable to eep the PMT gain for the uorescence pulses constant over a large bacground current. n a previous GAP-note (GAP-99-63) [7] we presented results obtained with a conventional resistive passive divider as well as an active divider passing similar standing current. The PMT used was a 1-dynode XP212. The measurements presented here were performed using tapered passive divider and active divider networs biasing an 8-dynode XP362 PMT (the Auger candidate). 2 The passive biasing networ Resistive divider chains are the most widely used technique to bias PMTs. t is the simplest way to ensure the required potentials at the dynodes. The resistors can have equal value (with exception of the rst dynode-to-cathode one, which typically is doubled), or tapered in order to reduce the non linearity due to space-charge eects [5]. When a DC current circulates in the PMT, the potentials at the dynodes change. This is illustrated in Fig 1, in which the balance of currents at the various nodes is indicated. t can be noted that the current at the last resistor,? A, diminishes. A is the anode current. The divider current, which can be calculated by superposition, P is = + 1 N N +1 i= i [8], where is the current at the divider for =, and i is the inter-dynode current. Since HV is constant, the potential of all dynodes increases when an anode DC current circulates g g g g g g g g 8 a C C C R a 3R R R R R R 1.25 R ( g 7 g 6 ) 1.75 R ( g 8 g 7 ) R 9 =1.25R g - g - g - g g g g - R= 31 a +HV Figure 1: The current distribution in a PMT divider when a DC anode current circulates. A value of R=31 was used in the measurements reported in this paper. An increase in the dynodes' potentials rises the PMT gain due to the Secondary Emission Ratio (SER) dependance on interdynode voltages. The standard way to eep

3 PMT gain variations small is to pass a large current, at least 1 times larger than the maximum DC anode current, through the voltage divider [5]. Typical behaviour is that at low anode current the pulse gain remains constant, while at ( A = ) = :1 it increases by about 7 % [8]. For a tapered bias networ the gain increase is less, about 3 % [5], however the gain rises rapidly with larger currents. An increase in bacground by a factor of 3 ( A 6:5A)rises the gain by 8 % when a standing divider current of 3 A passes through the tapered bias networ [9]. Typical counter measures to avoid the change in gain include the use of a Zener diode between the rst dynode and cathode, or across the last resistor. This method has the disadvantage that since the Zener voltage is xed, the ratio of potentials preserved by the resistive divider at any value of high voltage is lost. n addition, Zener diodes are noisy. 3 The active biasing networ The active biasing networ [5] uses bipolar transistors which can pass a variable current while eeping its collector-emitter voltage constant. Fig 2 shows the scheme of the divider used to bias the 8 dynode PMT. 3u u u u u u 1.25u 1.75u 1.25u C C C C 3R R R R R g 2R a - A -.5 Ra 2 - g 3 - g g - g R=511 2R 2.5R 3.5R 2.5R.5 - a +HV Figure 2: An active divider used to bias an 8 dynode XP362 PMT. Fig 2 shows the balance of currents at the various nodes. t was assumed that the current amplication factor h F E 1 and so the base current can be neglected in this analysis. A current? A? 1 2 passes through the last transistor while its collector-emmiter voltage is xed by the resistor chain and remains constant. The potentials at the dynodes are virtually undisturbed for any value of A providing that > A. There is now no need to mae > 1 A as it should be done with the passive divider. Nevertheless, for practical reasons has to have a minimum value to

4 ensure the transistors a minimum current amplication. The diodes are used to protect the base-emmiter junction against accidental application of reverse voltage. Regarding transistors availability, nowadays there is a variety of surface mounting devices with voltage capabilites of 4 V and more [1]. This is more than necessary as each transistor should sustain inter-electrode voltages of typically about 1-15 V. We have selected an FMMT 458 from Zetex which has a breadown voltage of 4 V and large h F E at low currents. n fact h F E 1 at c less than 1 ma. 4 Experimental results We biased a XP362 PMT with passive and active divider networs. The HV was xed at 94 V to set the PMT gain at The passive divider standing current was 272 A, and that of the active divider 145 A. Power dissipation was 246 mw and 131 mw, respectively. A green LED biased with an adjustable DC current illuminated the PMT. A second LED delivered pulses of constant amplitude and 1s width. The anode current was AC coupled to a load resistor of 511 Ohm, and the output signal readout with a Textroni TDS22 digital oscilloscope. The amplitude deviations were recorded as a function of bacground anode current, and plotted as deviation from linearity versus bacground current. Fig 3 shows the results for both passive and active biasing networs. A clear improvement of linearity is noted at large bacground currents when using the active divider. n fact variations in gain for this particular PMT are (as can be seen from gure 3).79% at 2 A, 1.59% at 3 A, increasing linearly with the logarithm of the anode current to a value of 7.4% at 11 A. Similar measurements using the active networ show no variation in PMT gain up to 1 A, and increases of 1.% at 2 A, 1.9% at 3 A, 2.9% at 4 A, 3.9% at 5A, and of 4.4% at 6 A. Briey, in the case of the passive divider the expected gain rise is seen at 1 % the divider current. With the active networ the same rise is measured only when the bacground current is 1 order of magnitude larger. This value corresponds to 14 % of the standing divider current. We repeated these measurements for a variety of temperatures ranging from room temperature to 55 C. n all such measurements variations in gain were constrained by the 1:6% sensitivity of the Textroni TDS22 digital oscilloscope. n other words, no variation in gain was measured as function of temperature. Noise referred to the input was measured and showed to be within 2% equal in both networs. 5 Conclusions The FD should wor under variable bacground illumination. Not only the bacscattered light from the Moon, but other sources, as well, could give rise to large bacground currents through the PMT, even during unpredictable circumstances. We want to eep the PMT gain independent of this current. The active divider is an eective and, to our nowledge, the only way to assure this. n addition, it allows a substantial reduction in power dissipation. The cost of the extra parts (3 transistors plus 3 diodes) is approximately $1:1 per channel, however power disipation is a factor 1.9 smaller for the active networ than

5 Gain Variation (%) Pass 27 C 3 C 35 C 4 C 45 C 5 C 55 C anode (ua) Figure 3: Gain variation, G=G (%), as function of bacground current, for both active and passive networs. The current through the passive networ was 272 A. The current through the active one was 145 A. for the passive one. The power saving implies a reduction of HVPS costs. Also LVPS requires less DC current per channel. References [1] R. M. Baltrusaitas et al, Nucl. nstr. and Meth. A24 (1985) 41.

6 [2] B. Dawson and A. Smith, GAP [3] See mail of David Kieda of July 3th 1998 to Fluorescence Group. [4] talian-german Fluorescence Group Meeting. CERN, October 25-26, Also see B. Dawson to Fluorescence Group of Dec. 2nd [5] See Hamamatsu and Philips PMT manuals. [6] S. Argiro, D.V.Camin, M. Destro, and C.K. Guerard. Monitoring DC Anode current of a grounded-cathode PMT. Submitted Nucl. nstr. and Meth., also GAP [7] S. Argiro, D.V.Camin, M. Destro, and C.K. Guerard. Passive and Active PMT Biasing Networs. GAP [8] See Philips PMT manual. [9] G. Matthiae and P. Privitera. Study of the Philips Hexagonal PMT XP362 for the FD detector (2/1/98). [1] For component description see

Monitoring DC anode current of a grounded-cathode photomultiplier tube

Nuclear Instruments and Methods in Physics Research A 435 (1999) 484}489 Monitoring DC anode current of a grounded-cathode photomultiplier tube S. Argirò, D.V. Camin*, M. Destro, C.K. GueH rard Dipartimento