1: % Template article for preprint document class `elsart'

2: % SP 2001/01/05

3: 

4: \documentclass{elsart}

5: 

6: % Use the option doublespacing or reviewcopy to obtain double line spacing

7: % \documentclass[doublespacing]{elsart}

8: 

9: % if you use PostScript figures in your article

10: % use the graphics package for simple commands

11: % \usepackage{graphics}

12: % or use the graphicx package for more complicated commands

13: % \usepackage{graphicx}

14: % or use the epsfig package if you prefer to use the old commands

15: \usepackage{rotating}

16: \usepackage{epsfig}

17: \usepackage{cite}

18: 

19: % The amssymb package provides various useful mathematical symbols

20: \usepackage{amssymb}

21: 

22: % \parskip12pt plus 1pt minus 1pt

23: % \parindent 0mm

24: % \setlength{\floatsep}{0em}

25: 

26: \setcounter{bottomnumber}{4}

27: \setcounter{totalnumber}{9}

28: \renewcommand{\bottomfraction}{1.0}

29: \renewcommand{\textfraction}{0.0}

30: \renewcommand{\arraystretch}{1.1}

31: 

32: \def\mum{$\mu$m}

33: \def\be{\begin{equation} }

34: \def\ee{\end{equation} }

35: \def\ba{\begin{eqnarray} }

36: \def\ea{\end{eqnarray} }

37: \def\ban{\begin{eqnarray*} }

38: \def\ean{\end{eqnarray*} }

39: \def\epem{\mbox{e}^+\mbox{e}^-}

40: \def\arco{\mbox{ArCO$_2$}}

41: \def\mmrcm{\mbox{$\mu$m / $\sqrt{\mbox{cm}}$}}

42: \begin{document}

43: 

44: \begin{frontmatter}

45: 

46: % Title, authors and addresses

47: 

48: % use the thanksref command within \title, \author or \address for footnotes;

49: % use the corauthref command within \author for corresponding author footnotes;

50: % use the ead command for the email address,

51: % and the form \ead[url] for the home page:

52: % \title{Title\thanksref{label1}}

53: % \thanks[label1]{}

54: % \author{Name\corauthref{cor1}\thanksref{label2}}

55: % \ead{email address}

56: % \ead[url]{home page}

57: % \thanks[label2]{}

58: % \corauth[cor1]{}

59: % \address{Address\thanksref{label3}}

60: % \thanks[label3]{}

61: \title{Position Sensing from Charge Dispersion in 

62: Micro-Pattern Gas Detectors with a Resistive Anode}

63: 

64: % \author[ref_CU]{B.~Carnegie},

65: \author[ref_CU,ref_TR]{M.~S.~Dixit\corauthref{add_MSD}},

66: \ead{msd@physics.carleton.ca}

67: \author[ref_JD]{J.~Dubeau},

68: \author[ref_UM]{J.-P.~Martin}

69: % \author[ref_CU]{H.~Mes},

70: and

71: \author[ref_CU]{K.~Sachs}

72: % \ead{sachs@physics.carleton.ca}

73: % \ead[url]{home page}

74: 

75: \address[ref_CU]{Department of Physics, Carleton University, 

76:         \\ 1125 Colonel By Drive, Ottawa, ON K1S 5B6 Canada}

77: \address[ref_JD]{DETEC, Aylmer, QC, Canada}

78: \address[ref_UM]{Universiy of Montreal, Montreal, QC, Canada}

79: \address[ref_TR]{TRIUMF, Vancouver, BC Canada}

80: \corauth[add_MSD]{Corresponding author.

81:          Tel.: +1-613-520-2600, ext. 7535; fax: +1-613-520-7546.}

82: % \corauth[add_KS]{Corresponding author; 

83: %         tel.: +1-613-520-2600, ext. 1567; fax: +1-613-520-7546.}

84: 

85: 

86: % use optional labels to link authors explicitly to addresses:

87: % \author[label1,label2]{}

88: % \address[label1]{}

89: % \address[label2]{}

90: 

91: \begin{abstract}

92: % Text of abstract

93:  Micro-pattern gas detectors, such as the Gas Electron Multiplier (GEM) and the Micromegas 

94: need narrow high density anode readout elements to achieve good spatial resolution. A 

95: high-density anode readout would require an unmanageable number of electronics channels for certain 

96: potential micro-detector applications such as the Time Projection Chamber. We describe below a 

97: new technique to achieve good spatial resolution without increasing the electronics channel 

98: count in a modified micro-detector outfitted with a high surface resistivity anode readout 

99: structure. The concept and preliminary measurements of spatial resolution from charge 

100: dispersion in a modified GEM detector with a resistive anode are described below.

101: \end{abstract}

102: 

103: \begin{keyword}

104: % keywords here, in the form: keyword \sep keyword

105: Gaseous Detectors \sep 

106: Position-Sensitive Detectors \sep

107: Micro-Pattern Gas Detectors \sep

108: Gas Electron Multiplier \sep

109: Micromegas

110: 

111: % PACS codes here, in the form: \PACS code \sep code

112: \PACS 29.40.Cs \sep 29.40.Gx 

113: 

114: \end{keyword}

115: \end{frontmatter}

116: 

117: % main text

118: \section{Introduction}

119: \label{sec:intro}

120: 

121:   A new class of high-resolution multi-channel gas avalanche micro-detectors has been 

122: developed during the past decade for charged particle tracking. The Gas Electron Multiplier 

123: (GEM) \cite{cit:GEM} and the Micromegas \cite{cit:MM} are examples of some of the new micro-pattern gas 

124: detectors \cite{cit:mpgd} already in wide use. The GEM and the Micromegas sample the avalanche charge 

125: using arrays of closely spaced long parallel anode strips to measure a single co-ordinate.  

126: Spatial resolutions of 40 to 50 \mum\ are typical with anodes strips at 200 \mum\ pitch. 

127: Micro-detectors do require more instrumented channels of electronics than multi-wire proportional 

128: chambers. However, the number of readout channels has not yet become an issue for most 

129: experiments using micro-detectors for charged particle tracking. 

130: 

131:  There are potential micro-detector applications, however, where the electronics channel count 

132: may become an issue. The Time Projection Chamber (TPC) \cite{cit:TPC} used in high energy physics 

133: experiments is one such example. A single endcap detector is used in the TPC to measure 

134: both the radial and the azimuthal co-ordinates of the ionization charge cluster. In the 

135: conventional TPC, read out with a multi-wire proportional chamber endcap, the radial 

136: co-ordinate is obtained from the anode wire position. A precise second co-ordinate along the 

137: anode wire length is measured by sampling the induced cathode charge with a series of 

138: several mm wide rectangular pads.

139: 

140:  The spatial resolution of a TPC in a high magnetic field is dominated by the 

141: {\boldmath $E\times B$\unboldmath} and track 

142: angle systematic effects \cite{cit:clif,cit:amend}. Replacement of the usual anode wire/cathode pad readout with 

143: one based on micro-detectors with anode pad readout would almost entirely eliminate the 

144: systematics and has the potential to improve the TPC performance significantly. However, the 

145: suppression of the transverse diffusion in a high magnetic field may often result in the 

146: collection of most of the avalanche charge within the width of a single anode pad resulting in 

147: a loss of TPC resolution. For better resolution, a micro-detector readout TPC will need either 

148: a finely segmented anode pad structure with a prohibitively large number of instrumented 

149: channels of electronics, or perhaps the complication of specially shaped pads which enhance 

150: anode charge sharing \cite{cit:chevron}. 

151: 

152:  We describe here a new technique which can be used to measure the position of a localized 

153: charge cluster in a micro-detector using pads of widths similar to those used in wire-pad 

154: systems. Most of our tests have so far been done with a modified double-GEM detector. 

155: However, the new technique appears to be sufficiently general to be applicable to the 

156: Micromegas.

157: 

158: 

159: \section{Charge dispersion in a micro-detector with a resistive anode}

160: 

161:  With certain modifications to the anode readout structure, it is possible to measure the position 

162: of a localized charge cluster in a micro-detector with pads wider than have been used so far. A 

163: thin high surface resistivity film is glued to a separate readout pad plane and is used for the 

164: anode (Fig. \ref{fig:scheme}). The resistive anode film forms a distributed 2-dimensional resistive-capacitive 

165: network with respect to the readout pad plane. Any localized charge arriving at the anode surface 

166: will be dispersed with the $RC$ time constant determined by the anode surface resistivity and the 

167: capacitance density determined by the spacing between the anode and readout planes. With the 

168: initial charge dispersed and covering a larger area with time, wider pads can be used for signal 

169: pickup and position determination. 

170: 

171: This spatial dispersion, which can be explained by a simple physical

172: model described below, does not attenuate the pad charge signal.

173: As the initial charge cluster diffuses, the electron density on the 

174: resistive surface, varying with time, is capacitively mirrored on the readout 

175: pads below. Since the readout pads are directly connected to the preamplifiers, 

176: there is no charge signal amplitude loss.

177: 

178: \subsection{A model for charge dispersion in a micro-detector with a resistive anode}

179: 

180:  The resistive anode and the readout plane together can be looked upon to form a distributed 

181: 2 dimensional $RC$ network in the finite element approximation. Consider first the 1 dimensional 

182: problem of a point charge arriving at $t=0$ at the origin in the middle of an infinitely long wire 

183: grounded at both ends.  For small inductance, the space-time evolution of the charge density 

184: $\rho$ on the wire is given by the well-known Telegraph equation:

185: \be

186: \frac{\partial \rho}{\partial t} = h \frac{\partial^2 \rho}{\partial x^2}

187: \hspace{1cm}\mbox{where} \hspace{1cm}

188: h = 1/RC \; . \label{eq:T-1dim} \ee

189: Here $R$ is resistance per unit length and $C$ the capacitance per unit length for the wire. 

190: 

191: The solution for charge density is given by:

192: \be

193: \rho(x,t) = \sqrt{\frac{1}{4\pi t h}} \exp(-x^2 / 4 t h)\; . \label{eq:1dim}  \ee

194: 

195: 

196:  In analogy with the 1 dimensional case, we can write the Telegraph equation for the case of a 

197: resistive surface. At time $t = 0$, a point charge is collected at the origin by a resistive anode 

198: surface of infinite radius (for simplicity). The 2-dimensional Telegraph equation for the charge 

199: density is:

200: \be

201: \frac{\partial \rho}{\partial t} = h \left[ \frac{\partial^2\rho}{\partial r^2} +

202: \frac{1}{r}\frac{\partial \rho}{\partial r} \right] \; , \label{eq:T-2dim}  \ee

203: where in this case, $R$ is the surface resistivity and $C$ is capacitance per unit area.

204: 

205: The solution for the charge density function in this case is given by:

206: \be

207: \rho(r,t) = \frac{1}{2 t h} \exp(-r^2 / 4 t h)\; . \label{eq:2dim}  \ee

208: 

209: 

210: The charge density function (equation \ref{eq:2dim}) for the resistive anode varies with time and is 

211: capacitively sampled by the readout pads. Fig. \ref{fig:charge} shows the time evolution of the charge density 

212: for an initially localized charge cluster for our detector. The charge 

213: signal on a pad can be computed by integrating the time dependent charge density function over 

214: the pad area. The shape of the charge pulse on a pad depends on the pad geometry, the location 

215: of the pad with respect to the initial charge and the $RC$ time constant of the system. 

216: 

217: 

218: \subsection{Charge dispersion signal in micro-detectors with long readout strips}

219:  

220: The charge dispersion measurements were carried out with a modified GEM detector with long 

221: anode strips. Since a spatial co-ordinate measurement for long strips is meaningful only in a 

222: direction transverse to the strip length, we can use the 1-dimensional Telegraph equation 

223: (\ref{eq:T-1dim}) to 

224: describe the situation. However, the solution for the charge density must account for the finite 

225: size of the resistive anode in contrast to the solution given by equation (\ref{eq:1dim}) in the long wire 

226: approximation. 

227: 

228: The boundary conditions to solve equation (\ref{eq:T-1dim}) in this case are: 

229: \be

230: \rho(x=0,t) = \rho(x=s,t) = 0\; ; \; 0 \leq t \leq \infty  \; , \label{eq:boundary}  \ee

231: where $s$ is the size of the resistive foil (assumed square) held at ground potential along its 

232: boundary.

233: 

234: The solution satisfying the finite boundary conditions is:

235: \be

236: \rho(x,t) = \sum_{m=1}^\infty A_m \exp [ -h(m\pi/ s)^2 t] \sin(xm\pi/s) \; , \label{eq:simsol}  \ee

237: where the coefficients $A_m$ can be determined from the knowledge of the initial charge density:

238: \be

239: A_m = \frac{2}{s}\int_0^s \rho(x,t=0) \sin(xm\pi/s) dx \; . \label{eq:coeff}  \ee

240: 

241:  The signal on a readout strip can be computed by integrating the charge density function over 

242: the strip width. Furthermore, the finite extent of the initial charge cluster, the intrinsic 

243: micro-detector charge signal rise time as well as the rise and fall time characteristics of the front-end 

244: electronics determine the shape of the measured signal shape. All these parameters need to be 

245: included in the model to compare to the experiment.

246:  

247: Model calculations were done for a GEM detector with a resistive anode readout with 1 mm 

248: wide strips. The anode resistivity and anode-readout gap limits

249: the computed spatial spread of the charge dispersion over pads to about 700~\mum\ comparable to 

250: transverse diffusion in a high magnetic field TPC. Simulated signals for the readout strip directly 

251: below the initial ionization charge cluster and for the next four adjacent strips are shown in 

252: Fig. \ref{fig:simsig}. The same figure also shows the simulated pad response function or equivalently the 

253: spatial spread of an initially localized charge cluster due to charge dispersion.

254: 

255: 

256: \section{Spatial resolution measurements in a GEM detector with a resistive anode}

257:  

258: The charge dispersion test measurements were made with the modified double-GEM detector 

259: (Fig. \ref{fig:scheme}) filled with Ar/CO$_2$ 90/10. A 50 \mum\ thick Mylar film with a 

260: surface resistivity of 2.5 M$\Omega$

261: per square was glued to the pad readout board. The spacing between the anode and readout 

262: planes was close to 100~\mum\ including the thickness of the glue. 

263: 

264: The initial ionization is provided by x-ray photon conversion in the gas. 

265: The average photon energy was 

266: about 4.5 keV as low energy bremsstrahlung photons from the copper target x-ray tube, run at 7 

267: kV, were absorbed by the material in the x-ray tube and detector windows. A 

268: \mbox{$\sim$40 \mum} pinhole in 

269: a thin brass sheet was used to produce a miniaturized x-ray tube focal spot image in the GEM 

270: drift gap. The size of the x-ray spot at the detector is estimated to be on the order of 

271: 70 \mum . 

272: After avalanche multiplication and diffusion, the RMS size of the electron cloud reaching the 

273: resistive anode was \mbox{$\sim$400 \mum}. The gas gain was about 2500.

274: 

275: Signals were read out from 7 cm long and 1.5 mm wide strips. The front-end electronics 

276: consisted of Aleph \cite{cit:amend} TPC wire charge preamplifiers followed by receiver amplifiers. Signals 

277: from 8 contiguous strips were digitized using two 4-channel Tektronix digitizing oscilloscopes. 

278: A computerized translation stage was used to move the x-ray spot in small steps over the width 

279: of the centre strip. One thousand event runs were recorded for each x-ray spot position on an 

280: event by event basis. 

281: 

282: For a given anode surface resistivity and readout geometry, the observed shape of the charge 

283: pulse depends on the strip position with respect to the location of primary charge cluster on the 

284: resistive anode.  Fig. \ref{fig:scope} shows an event where the x-ray ionization spot is located directly above 

285: the centre of a readout strip. A fast charge pulse is observed on the strip peaking in time with the 

286: maximum of the charge density at the anode surface above. Pulses on strips farther away have a 

287: slower rise time and peak late because the local charge density on the anode surface nearest the 

288: strip reaches its maximum later. Also, an early short duration induced pulse is visible for strips 

289: adjacent to the main strip. The induced pulses \cite{cit:orsay}, produced by electron motion in the GEM 

290: induction gap, have demonstrated position sensitivity \cite{cit:LCWS} but require the use of high-speed pulse 

291: shape sampling electronics. In addition, measurable induced pulses are specific to GEM 

292: detectors with sizeable induction gaps. For charge dispersion measurements described below, the 

293: induced GEM pulse information has not been used. 

294: 

295: 

296: 

297: \section{Data analysis and results}

298: 

299: The charge dispersion signals were confined to a narrow region on the readout board in the 

300: present setup. There were measurable signals above noise only on three 1.5 mm wide strips. The 

301: analysis to determine the space point resolution from the event by event data consisted of 

302: following steps: a) Determine the pulse heights of signals on the strips, independent of rise time; 

303: b) Compute a centre of gravity position for the event from the measured pulse heights; c) Correct 

304: for the bias in the computed centre of gravity to obtain the position of the ionization cluster for 

305: the event. The standard deviation of the bias corrected centre of gravity position with respect to the 

306: known x-ray beam spot position for the run gives a measure of the space point resolution. 

307: 

308: The data from each 1000 event run were sub-divided into two equal data sets: one used for 

309: calibration and one for resolution studies. The pulse heights were obtained by fitting polynomial 

310: functions to the digitized pulse shape data. The calibration data set was used to determine and fix 

311: the coefficients of polynomial functions (see Fig. \ref{fig:poly}) used subsequently in the analysis of events 

312: in the resolution data set. 

313: 

314: The centres of gravity for the events in the calibration data set were computed from the measured 

315: pulse heights. The correction function for the bias in the centre of gravity method 

316: (see Fig. \ref{fig:bias}) 

317: was determined by plotting the mean value of computed centres of gravity against the known 

318: x-ray spot position for the individual runs. 

319: 

320: The pulse heights and the peak positions for the events used for resolution studies were 

321: determined by fitting the pulses to the fixed polynomials shapes obtained from the corresponding 

322: calibration data set. The computed centres of gravity were converted to ``true position'' by 

323: interpolation using the bias correction function determined from the calibration data set. 

324: Fig. \ref{fig:resolution} 

325: shows the measured resolution function for the 1.5 mm wide strips at two different positions of 

326: the x-ray ionization spot over the strip. Fig. \ref{fig:summary} shows the measured spatial resolutions and the 

327: position residuals: i.e. the deviation of measured positions from the micrometer settings. The 

328: small systematic trends apparent in the plot appear to be related to an imprecise knowledge of the 

329: experimental parameters in the present setup and any remaining biases in the present method of 

330: analysis. Nevertheless, the standard deviations of the position measurements, in the range of 50 

331: to 80 \mum , are all consistent with the size of the collimated x-ray spot at the detector.

332: 

333: 

334: \section{Outlook and summary}

335: 

336: We have demonstrated that a controlled $RC$ dispersion of the avalanche charge makes it possible 

337: to measure its position with a micro-detector with strips wider than have been used previously. 

338: The pad response function and signal shapes are determined by the anode surface resistivity and 

339: anode-readout plane gap. With the proper choice of the $RC$ time constant of the system, the 

340: charge dispersion technique will not compromise the counting rate ability of the detector. Nor 

341: should it compromise the 2-track resolving power of the detector which should be limited only 

342: by the diffusion effects in the gas. Once the characteristics of the charge dispersion signal are 

343: properly understood, it should also be possible to simplify the technique replacing the pulse 

344: shape measurement system with less expensive charge integrating electronics. 

345: 

346:  The charge dispersion space point resolution studies described here were done for long readout 

347: strips in a modified GEM detector with a resistive anode. Further experimental and simulation 

348: studies are in progress to investigate the spatial resolution capabilities of the charge dispersion 

349: technique with rectangular pads similar to those used in wire-cathode pad TPCs. We are also 

350: testing the concept of position sensing from charge dispersion with the Miromegas where the 

351: high anode resistance may help improve the detector HV stability as well as protect the front-end 

352: electronics from spark damage.

353: 

354: \clearpage

355: 

356: \section*{Acknowledgments}

357:  

358: The charge preamplifiers used in these measurements came from the Aleph TPC at CERN and 

359: we wish to thank Ron Settles for making these available to us. Ernie Neuheimer lent us his 

360: expertise in designing, building and troubleshooting much of the specialized electronics used for 

361: these measurements. Mechanical engineers Morley O'Neill and Vance Strickland helped with 

362: the detector design and in improving the clean-room facility used for the detector assembly. 

363: Philippe Gravelle provided technical assistance when needed. Our CO-OP students Alasdair 

364: Rankin and Steven Kennedy made significant contributions to all aspects of this research from 

365: hardware construction to improving the data acquisition software as well as writing some of the 

366: data analysis code. Finally, one of the authors (MSD) would like to express his thanks to 

367: V.~Radeka for an illuminating discussion concerning the phenomenon of charge dispersion. This 

368: research was supported by a project grant from the Natural Sciences and Engineering Research 

369: Council of Canada.

370: 

371: \begin{thebibliography}{00}

372: \bibitem{cit:GEM} F. Sauli, Nucl. Inst. Meth. {\bf A386} (1997) 531.

373: \bibitem{cit:MM} Y. Giomataris et al, Nucl. Inst. Meth. {\bf A376} (1996) 29.

374: \bibitem{cit:mpgd} F. Sauli and A. Sharma, Ann. Rev. Nucl. Part. Sci. {\bf 49} (1999) 341.

375: \bibitem{cit:TPC} D. R. Nygren, PEP 198 (1975).

376: \bibitem{cit:clif} C.K.Hargrove et al, Nucl. Inst. Meth. {\bf 219} (1984) 461.

377: \bibitem{cit:amend} S. R. Amendolia et al., Nucl. Inst. Meth. {\bf A283} (1989) 573. 

378: \bibitem{cit:chevron} M. Schumacher, LC-DET-2001-014.

379: \bibitem{cit:orsay} M.S. Dixit et al, Proceedings of Workshop on Micro-Pattern Gas Detectors, Orsay France 

380: (1999).

381: \bibitem{cit:LCWS} D. Karlen et al, Physics and experiments with future linear $\rm e^+e^-$ colliders, LCWS2000, 

382: American Institute of Physics Conf. Proc. Vol 578.

383: 

384: % \bibitem{label}

385: % Text of bibliographic item

386: % notes:

387: % \bibitem{label} \note

388: % subbibitems:

389: % \begin{subbibitems}{label}

390: % \bibitem{label1}

391: % \bibitem{label2}

392: % If there is a note, it should come last:

393: % \bibitem{label3} \note

394: % \end{subbibitems}

395: % \bibitem{}

396: \end{thebibliography}

397: 

398: \clearpage

399: 

400: 

401: 

402:  

403: \begin{figure}[bp]

404: \centerline{\mbox{\epsfxsize=\textwidth \epsffile{fig1_GEM_foil.eps}}}

405: \caption[]{\label{fig:scheme}

406: Schematics of the resistive anode double-GEM detector used for charge dispersion 

407: studies.

408: }

409: \end{figure}

410: 

411: \begin{figure}[bp]

412: \centerline{\mbox{\epsfxsize=\textwidth \epsffile{fig2_dispersion.eps}}}

413: \caption[]{\label{fig:charge}

414: The evolution of the two dimensional charge density function (equation \ref{eq:2dim}) 

415: on the resistive anode with a resistivity of 2.5~M$\Omega$ per 

416: square and anode-readout plane separation of 100~\mum . 

417: For the model calculation the initial 

418: charge was point-like and localized at the origin at time = 0.

419: }

420: \end{figure}

421: 

422: \begin{figure}[bp]

423: \centerline{\mbox{\epsfxsize=\textwidth \epsffile{fig3_padres.eps}}}

424: \caption[]{\label{fig:simsig}

425: Simulated signals for 1 mm wide strips for an anode resistivity of

426: 2.5~M$\Omega$ per 

427: square and anode-readout plane separation of 100~\mum . The pad response function for charge 

428: dispersion is shown on the right.  The diffusion effects were neglected for this simulation.

429: }

430: \end{figure}

431: 

432: 

433:  

434: \begin{figure}[bp]

435: \epsfxsize=\textwidth \epsffile{fig5_ScopePulses.eps}

436: \caption[]{\label{fig:scope}

437: Observed charge dispersion signals on three adjacent strips for a single x-ray photon conversion in the 

438: double-GEM detector. Tektronix scope pulses with \mbox{400 ns/div} on the 

439: a)~right strip (20 mV/div), 

440: b)~central strip (50 mV/div) and

441: c)~left strip (20 mV/div).}

442: \end{figure}

443:  

444: \begin{figure}[bp]

445: \centerline{\mbox{\epsfxsize=\textwidth \epsffile{fig6_scope.eps}}}

446: \caption[]{\label{fig:poly}

447: Polynomials were fitted to the calibration data set for each run to fix the functions used to 

448: determine the pulse height of events for resolution studies. The top figure shows the polynomial 

449: fit to the average of fast charge pulses for the centre strip. The figure below shows the 

450: polynomial fit to the average of late charge dispersion pulses for an adjacent strip.

451: }

452: \end{figure}

453: 

454: 

455: 

456:  

457: \begin{figure}[bp]

458: \centerline{\mbox{\epsfxsize=10cm \epsffile{fig7_bias.eps}}}

459: \caption[]{\label{fig:bias}

460: The bias correction function shown in the figure was experimentally determined. The bias 

461: function was used in converting the computed centroid of signals on three strips to the true 

462: position on an event by event basis. 

463: }

464: \end{figure}

465: 

466: 

467: 

468:  

469: \begin{figure}[bp]

470: \centerline{\mbox{\epsfxsize=\textwidth \epsffile{fig8_centroid.eps}}}

471: \caption[]{\label{fig:resolution}

472: The resolution function for two x-ray beam spot positions.

473: }

474: \end{figure}

475: 

476: 

477:  

478: \begin{figure}[bp]

479: \centerline{\mbox{\epsfxsize=\textwidth \epsffile{fig9_resolution.eps}}}

480: \caption[]{\label{fig:summary}

481: The summary of the spatial resolutions and the position residuals for the x-ray scan across 

482: the readout strips.

483: }

484: \end{figure}

485: 

486: \end{document}

487: 

488: