1. Preliminary Performance Measurements for Streak Camera with Large-Format Direct-Coupled CCD Readout* 15th Topical Conference on High-Temperature Plasma Diagnostics San Diego, CA April 19 - 22, 2004 R. A. Lerche, J. W. McDonald, R. L. Griffith, D. S. Andrews, G. Vergel de Dios, A. W. Huey, P. M. Bell, O. L. Landen Lawrence Livermore National Laboratory P. A. Jaanimagi, R. Boni Laboratory for Laser Energetics, University of Rochester * This work was performed under the auspices of the U.S. Department of Energy by University of California Lawrence Livermore National Laboratory under contract No. W-7405-Eng-48. HTPD-2004-SC - 1 ral 040418

2. HTPD-2004-SC - 2 ral 040418 Abstract Livermore’s ICF Program has a large inventory of optical streak cameras built in the 1970s and 1980s. The cameras are still very functional, but difficult to maintain because many of their parts are obsolete. The University of Rochester’s Laboratory for Laser Energetics is leading an effort to develop a fully automated, large- format streak camera that incorporates modern technology. Preliminary characterization of a prototype camera shows spatial resolution better than 20 lp/mm, temporal resolution of 12 ps, line- spread function of 40  m (fwhm), contrast transfer ratio (CTR) of 60% at 10 lp/mm, and system gain of 101 CCD electrons per photoelectron. A dynamic range of 175 for a 2 ns window is determined from system noise, linearity and sensitivity measurements.

3. HTPD-2004-SC - 3 ral 040418 Summary: We have characterized a new prototype streak camera (ROSS) 1.The streak camera ROSS (Rochester Optical Streak System) Design effort led by Laboratory for Laser Energetics LLNL collaborated on design Streak tube: Photonis P-510 Direct coupled CCD (2K x 2K E2V 42-40 backside) LLNL input optics used for these tests 2.Camera performance Spatial magnification: 1.3 Spatial resolution: > 20 lp/mm (limiting visual) Temporal resolution: 12 ps @ 2 ns window Sensitivity: 101 CCD e - / photoelectron System noise: 5 e - / pixel Dynamic range: ~225 (at best resolution) 3.Camera appears to meet NIF optical streak camera requirements

4. HTPD-2004-SC - 4 ral 040418 The compact streak camera has a direct-coupled large-format CCD readout Prototype camera configuration S-20 Streak tube Photonis P-510 CCD Input optics (for our tests*): LLNL input hardware Tek C-27 lens 1-mm slit Mag: 1.167 Streak tube: Photonis P-510 S-20 photocathode CCD: Spectral Instruments SI-800 E2V 42-40 backside 2K x 2K (13.5  m pixels) Prototype streak camera (ROSS) (12” x 7” x 28” without input optics) * The prototype input optics module for the ROSS is not yet available.

5. HTPD-2004-SC - 5 ral 040418 Prototype camera magnification is 1.35, field-of-view is 20.5 mm Sample image for magnification and FOV measurements Test conditions: Light: Collimated (532 nm) Mask: 10-  m slit every 1.5 mm CCD: 2K x 2K (13.5-  m pixels) Time (CCD pixels) Space (CCD pixels) 02048 0 Notes: CCD (27 mm square) records central (high-res) region of 60-mm dia streak tube image. Magnification and FOV are referred to photocathode of the streak tube.

6. HTPD-2004-SC - 6 ral 040418 System line-spread function (LSF) shows a 40  m resolution (fwhm) 1. Measure system magnification 2. Illuminate 10-  m mask 3. Take spatial lineout Mask: 11 um (at photocathode) Binning: 2x2 FWHM: 2.0 super pixels (40  m) 40  m fwhm Line-Spread Function Gaussian

7. HTPD-2004-SC - 7 ral 040418 We calculate the system contrast-transfer function (CTF) from the line-spread function Calculation convolves LSF with square wave at various frequencies * = Contrast Transfer Function - Ronchi ruling measurements

8. HTPD-2004-SC - 8 ral 040418 Ronchi ruling measurements at 8.6 lp/mm confirm 70% CTR estimated from LSF Spatial Lineout and CTR CTR @ 8.6 lp/mm Measured: 68% LSF Prediction: 70% 8.6 lp/mm Ronchi Ruling Space Time

9. HTPD-2004-SC - 9 ral 040418 Contour plots show position dependence of spatial resolution in the streak camera image 1-mm slit 100-  m slit 2x2 binning Super-pix = 20  m FWHM (CCD super pixels) Position (CCD super pixel) 1000 0 20 0 10 Position (CCD super pixel) 8 0 4

10. HTPD-2004-SC - 10 ral 040418 Contour plot shows position dependence of time resolution in the streak camera image Contours showing position dependence of time resolution Test conditions: Collimated light (530 nm) Slit: 100  m CCD: 2K x 2K Sweep: Static (no sweep) Binning: 2x2 Time resolution: < 4 super pixels (fwhm) over 90% of image Position (CCD super pixel) FWHM (CCD super pixels) Position (CCD super pixel) 1000 0 0 5 10 Sweep (ns)  t (ps) 212 623 1246 30112

11. HTPD-2004-SC - 11 ral 040418 Camera gain is high enough to detect individual photoelectrons Determine total ADUs in signal Convert with CCD gain Determine number of pe - generated Energy at photocthode times QE Correct for streak camera time window Gain = 101 CCD e - / pe -  2x2 = 6 CCD e - * CCD gain = 1.09 e - /ADU FWHM: 11.3  = 4.81 pix Laser pulse Time (ns) Amp Noise for 2 sec exposure BinningNoise(e - ) 1x15.13 2x25.97 3x36.69 4x47.92 Histogram of Background Number of pixels Counts (ADUs) FWHM: 11.3 ADUs  noise : 4.81 ADUs Image of swept slit (3 mm x 0.5 mm) Time Space

12. HTPD-2004-SC - 12 ral 040418 We use 20% temporal broadening to define the maximum usable current density J 0 = 2.2 amp/cm 2 2.C-L current depends on geometry and voltage 1.20% broadening occurs near 1% of Child- Langmuir space-charge limited current (J 0 ) for laser pulse  t = 45 ps FWHM vs Current 1% C-L Window pc -15 kV Extraction grid -12.5 kV x Expect reduced performance for J > 1% of C-L current (J 0 )

13. HTPD-2004-SC - 13 ral 040418 We use SNR versus photoelectrons per resolution element as a figure-of-merit SNR = s 2 N / [s 2 N (  C /C) 2 + (s/s b ) 2 (  b /C) 2 ] 1/2 s 2 = # of detector pixels / image resolution element N = # of photoelectrons per detector pixel s b 2 = number of detector pixels per binned pixel  b = read plus dark current noise for one binned pixel C = # of CCD electrons / photoelectron  C = standard deviation in C For the ROSS streak camera we have: s 2 = 32 pixels (4 space, 8 time) s b 2 = 1, 4, 9, 16  b = 5.13, 5.97, 6.69, 7.92 C = 101 CCD e - / pe -  C = Unknown Time Space 1 pixel s b 2 (2x2) Image PSF (32 pixels) Maximize SNR by: Increasing s (more averaging) Increasing N (reduce sweep speed) Increasing C (more efficient pe - detection) Decreasing  read (improve CCD)

14. HTPD-2004-SC - 14 ral 040418 A SNR plot establishes the dynamic range (DR) at ~60 for an image resolution element SNR too low (avoid to ensure quality data) SNR versus Photoelectrons per Detector Pixel Sweep DR* 2ns175 6ns405 12ns855 * Binning: 2x2 Dynamic Range = f(s, sweep speed) >1% C-L 2ns 6ns 12ns C = 101

15. HTPD-2004-SC - 15 ral 040418 Summary: We have characterized a new prototype streak camera (ROSS) 1.The streak camera ROSS (Rochester Optical Streak System) Design effort led by Laboratory for Laser Energetics LLNL collaborated on design Streak tube: Photonis P-510 Direct coupled CCD (2K x 2K E2V 42-40 backside) LLNL input optics used for these tests 2.Camera performance Spatial magnification: 1.3 Spatial resolution: > 20 lp/mm (limiting visual) Temporal resolution: 12 ps @ 2 ns window Sensitivity: 101 CCD e - / photoelectron System noise: 5 e - / pixel Dynamic range: ~225 (at best resolution) 3.Camera appears to meet NIF optical streak camera requirements