Ground based photon counting detection for the 2010 Mars Laser Communications Demonstration

Transcription

1 Ground based photon counting detection for the 2010 Mars Laser Communications Demonstration William H. Farr Jet Propulsion Laboratory California Institute of Technology William Farr - 1

2 Optical Communications for Deep Space Operations * Mars Reconnaissance Orbiter Mars@ 0.6 AU Mars@ 2.4 AU Jupiter@ 6AU Saturn@ 10 AU Series5 ** Mars Laser Communication Demonstration 1.00E E+04 Planetary Images Streaming Video Difficulty (Mbps*AU^2) 1.00E E E+01 Cassini X- Ka- X- 1.00E+00 MRO * 1.00E-01 X- Mars Odyssey 1.00E-02 Opt MLCD ** Opt Hyper Spectral/ Synthetic Aperture Radar HDTV 1.00E E E E E E+03 Data Rate (Mbps) A. Biswas NASA is seeking orders of magnitude enhancement in deep space downlink capacity Optical communications systems can meet that goal Ground-based photon counting detectors are required that combine: Bandwidths in the 1 to 40 gigahertz range for active areas greater than 1 mm 2 Minimum photon detection efficiencies of 30% in the 1 to 1.5 micron wavelength range Saturation rates greater than 200 megahertz William Farr - 2

3 Photon Counting for Deep Space Optical Communications Detector Class Examples Photon Capacity Limit T s T w M Slot Width Frame (word) Time Alphabet Size T s T w = M*T s phase insensitive amplifier dual quadrature sensitive single quadrature sensitive parametric amplifier, Raman amplifier, laser amplifier coherent heterodyne coherent homodyne, degenerate parametric amplifier photon counting photomultiplier tube, cooled avalanche photodiode, hot electron superconducting hν/2ktln2 (for instance, 69 bits /photon at 1µm & 150K) C 1 1 B B log log2 = log M M M = from Brillouin s negentropy principle PPMencoding ( 1+ M ) C M 2 photon starved operations Deep Space Optical Communications requires data encodings that maximize the (bits/sec) per (Joule/sec) metric Photon counting can yield a higher channel capacity than phase-sensitive detectors PPM Encoding with photon counting detection is an attractive solution bits/second P avg average laser power = Bbit / s N η η link link loss photon / bit detector photon energy hν quantum efficiency Higher efficiency means a choice of: Lower transmitter power Smaller receive aperture Higher data rate William Farr - 3

4 Mars Laser Communications Demonstration NASA is planning a Mars Laser Communications Demonstration (MLCD) Laser terminal is to fly on the Mars Telecom Orbiter (MTO) with an October 2009 launch Laser terminal is to be boresighted with the MTO RF high gain antenna 5 watt average power at 1.06 µm transmitted through 30 cm aperture telescope Uplink beacon from Earth provides inertial pointing reference Fast steering mirror points and stabilizes transmitted laser beam with no unique requirements on spacecraft pointing or stability. Data rates up to 30 Mbps Not to exceed 70 kg mass, 130 W power Objective is to characterize laser communications from Mars under a variety of conditions (weather, Earth-Mars range, day/night, sun-earth-probe angle,...) William Farr - 4

5 Palomar Receive Terminal MLCD has selected the 5 m aperture Hale telescope for the primary ground terminal site 5-meter aperture supports high data rates Proximate to JPL with allocated JPL time Features 5 meter F/3.4 primary mirror Functional Adaptive Optics system Multiple accessible foci: Prime, Cassegrain, and Coude Facility will be modified to support daytime operations 5 m diameter solar blocking filter Additional dome air conditioning S/C tracking and light collection system Custom Optical Interface Detector Analog Conditioning Electronics To digital receiver assembly William Farr - 5

6 PRT Detector Size Requirements A large receiver requires a large detector area Angle-Area product is invariant, corresponding to number of spatial modes that must be processed Spatial and spectral filtering reduces background light contribution Atmospheric seeing controls the minimum required detector size Diffraction limit = 2.44 F λ F is focal ratio, λ is the receive wavelength Focal spot size (diffraction limit) x (D/r 0 ) D is telescope aperture [ F = (focal length) / D ] r 0 is atmospheric coherence size For instance, for a 5 m aperture at F/1 with worst case r 0 = 4 cm, the focal plane spot size is 0.65 mm In most atmospheric conditions focal spot diameters will be in the range of 0.2 to 0.5 mm Small arrays can be considered to meet simultaneous size and bandwidth requirements William Farr - 6

7 PRT Candidate Photon Counting Detectors Device Gain Gain Var. Geiger Mode InGaAsP Avalanche Photodiode Array Dark Noise (Kcps) 1.E+06 NA 20 for 10 micron active area Current Proposed Size (mm) 45% 45% 8x8 array on 0.1 mm centers Rise Time (ns) Dead time (us) Op Temp (K) 0.5 db Saturation Comment MHz requires microlens array (90% fill factor) and custom ROIC with > 500 MHz clocking Si:As Photon Counter 3.E <<1% 30-50% 1 < 0.8 NA MHz With increased PDE provides all desirable 1064-nm. PDE would be halved at 1550 nm Photomultiplier Tube (PMT) InP/InGaAsP Photocathode Hybrid PMT, InGaAsP/InP or InGaAs/InP photocathode 1064 nm 1.E+06 ~2 25 8% NA 3 x 8 3 NA KHz Low PDE, bandwidth and anode current limited. Difficult to meet even minimum success criteria 1-1.5E % 40% NA 240 > 50 MHz Very promising for both 1064 and 1550 nm Geiger mode array was decided to not be suitable for the PRT since a larger than 32x32 array would be required and the field-splitting optics would be complicated A conventional Near-IR PMT has low detection efficiency at 1064 nm, is too slow, too noisy, and has poor saturation characteristics The Si:As Photon Counter and Hybrid PMT were selected for further characterization and development William Farr - 7

8 Optical Communications Detector Characterization Facility Key Equipment: Optical Signal Synthesis Data Acquisition and Analysis Detector Bias and Environmental Control signal generation to 3 gigabits/sec data acquisition to 6 GHz ultra-low-noise amplifiers, 10 KHz - 20 GHz, operating from room temperature to < 4K detector environmental control from room temperature to < 4K optical modulation to 10 GHz at 1064 nm and 1550 nm biphoton absolute calibration at 1064 nm and 1550 nm photon counting to 4 GHz optical channel emulation Cryostat for Optical Detector Characterization Operates from 300 K to < 4K Free space or fiber coupled optical signal input DC and RF feedthroughs to >10GHz A dedicated laboratory has been established at JPL for testing optical detectors for deep space communications applications: Linear mode, Geiger mode, photon number resolving and photon counting detectors Emphasis on photon starved high bandwidth operations near theoretical channel capacity limits End-to-end support for laser transmitter, channel emulation, optical receiver and decoder William Farr - 8

9 NIPC Device Concept η = Fowler Relation for PtSi Quantum Efficiency C 1 2 ( hυ qφb ) C1 = 1.24 hυ λ + PtSi Intrinsic Region Gain Region Drift Region λ (1 ), λ c M. Petroff and M. Stapelbroek IEEE Trans. Nuclear Sci., 36, 158, (1989) - Spacer Region Substrate Contact Layer NIPC Structure Typical Si:As Absorption Spectrum The Arsenic doped Silicon detector has demonstrated single photon sensitivity over the 0.4 to 28 micron wavelength range This device is typically operated in the 6 to 10 K temperature range The device exhibits F near 1 with M > (localized avalanche gain process) Near-Infrared Photon Counter (NIPC) concept is to increase the near-infrared detection efficiency of the Si:As detector by adding a PtSi absorption layer Preliminary modeling (modified Fowler equation) and previous PtSi results have indicated that a detection efficiency from 20 to 50% should be possible Detection efficiency could be further enhanced by use of an optical cavity William Farr - 9

10 VLPC 1064 nm CW Performance Dark Counts Excess Noise M M 2 F = 2 KHz K 8.5K 9.0K 9.4K K 9.0K 8.5K 7.7K Bias Voltage (V) Optical Power (nw) Gain Counts Ke K 9.0K 8.5K 7.7K KHz K 9.0K 8.5K 7.7K Optical Pow er (nw) Optical Power (nw) Characterized the Visible Light Photon Counter optical communications performance Measured a detection efficiency at 1064 nm of < 0.05% William Farr - 10

11 Photon Counting Linearity Photon Interarrival Times Single Photon Pulse Response Afterpulse rate < 1 %, within 20 ns of main pulse VLPC Arrival Rate Histogram Exponential interarrival time distribution is expected for CW illumination William Farr - 11

12 Near-Infrared Photon Counter Performance Dark rate Bias in V nm Laser Illumination Bias in V However, NIPC performance to date has been poor Devices exhibit positive feedback at the PtSi - Si interface, a problem which has only been partially solved William Farr - 12

13 HPMT Detector R. A La Rue, et al., IEEE Trans. Elec. Dev., 44, 672, (1997) HPMT cathode bias Photocathode photocathode bias Vb + Baffle focus bias Anode APD Vk 50 Ohm transmission line Bias-T Vcc LNA OUTPUT Vf + + Preamp Anode SMA black blue yellow red photocathode anode photocathode cathode baffle electrode (electron focus) case common & APD anode Ammeter anode bias + Va Bias-T The HPMT uses a two-stage gain process to achieve single photon sensitivity: ultra-low noise multiplication on the order of 10 3 via energetic impact of kinetic (8 KV typical) electrons onto a GaAs semiconductor anode avalanche multiplication with gains on the order of 10 within a high-field region of the semiconductor anode (avalanche diode) 1 mm diameter photocathode; 3 cm diameter by 3 cm long cylindrical package William Farr - 13

14 InGaAsP HPMT Measured Performance Detection Efficiency (%) K K 270K K K 224K Photocathode Bias (V) Dark Rate (MHz) Vb Temperature (K) K K K K K K Photocathode Bias Count (V) Rate Linearity Quantum Efficiency (%) Photocathode Bias (V) 16-Sep Sep-04 1-Nov-04 Single photon response measured with CW illumination at 1064 nm Pulse gain and variance measured by integrating temporal pulse areas Mean gain <M> near 8000 with F = 1.03 ( <M 2 > / <M> 2 ) William Farr - 14

15 InGaAsP Linearity Optical Power (pw) Exponential Interarrival times under CW illumination Results show no significant after pulsing effects Dark counts also show exponential interarrival time distribution Count Rate (MHz) William Farr - 15

16 Slot Energy Histograms n s = 0.3 n s = 1.1 n s = 2.1 Ts = 1.6 ns A.U. A.U. A.U. note: horizontal scale on slot energy histograms varies between plots Further characterized the InGaAs HPMT with a PPM encoded signal Slot histograms represent the sum of signal and noise generated charge in a PPM slot Thresholding converts the slot integrated charge levels ( energy ) into a photon number Subsequent processing of the photon number signal is essentially noise free comparison of thresholded vs. non-thresholded signal processing for an example MLCD link K. Quirk William Farr - 16

17 End-End Demonstration Architecture TSG PC Test Signal Generator RS170 Hardware MPEG Encoder 1 Mbps TCP/IP Software FEC Encoder 2 Mbps USB PPM64 Mapper Video Input MHz ECL serial Display PC megabit/sec real-time link PPM-64 half-rate Forward Error Correction code CRC for code block validation Operates at near the nominal MLCD Operating Point 0.1 < n b < 2 2 < n s < nm Laser Modulator Optical Channel Emulator Output Display Software MPEG Decoder 1 Mbps TCP/IP FPGA Decoder digital FPGA Receiver analog HPMT Detector Decoder Host PC Receiver Host PC William Farr - 17

18 Performance Validation Characterized the real-time link performance using a live video source Evaluated frame loss rate using embedded Cyclic Redundancy Check code The link operates at better than one bit per photon (after decoding) William Farr - 18

19 Capabilities Roadmap Targeted detector developments are leading to rapid increases in demonstrated data rates for photon starved deep space optical communications links Specified performance is at better than one bit per photon Flight demonstrations such as MLCD are now required to retire operational risks Major risks are pointing and mission operations An Optical Deep Space Network can be in place by 2020 Providing an interplanetary internet for deep space probes and manned operations Year 1064 nm Detection Efficiency Analog Bandwidth (GHz) e- noise (FWHM) Saturation Rate (GHz) Data Rate (GHz) % GHz ultralow noise amp Linear Mode Photon Counting InGaAs and HgCdTe Year Nonlinear Optical Components % % % % NbN SSPD Array GHz Photon Counting William Farr - 19