Transcript
Visible Light Photon Counter Integrator
Group 48: Austin Jin Katie Nguyen Myoung Hyun Kim
Design Review for ECE 445, Spring 2016
TA: Ankit Jain
1 Introduction 3 1.1 Statement of Purpose 3 1.2 Objectives 4 1.2.1 Goals & Benefits 4 1.2.2 Functions & Features 4 2 Design 5 2.1 Block Diagram 5 2.2 Block Descriptions 5 2.2.1 Laser Source 5 2.2.2 VLPC 5 2.2.3 Integrator Circuit 6 2.2.4 Analog to Digital Converter 6 2.2.5 Data Processing 6 2.2.6 Graphical User Interface 7 2.2.7 Power Supply Circuit 7 2.2.8 EMI Shielded Enclosure 8 2.3 Circuit Schematic 9 2.3.1 Integrator Circuit 9 2.3.2 Analog to Digital Converter Circuit 9 2.3.3 Power Supply Circuit 10 2.4 Calculation 10 2.4.1 Integration for Single Pulse 10 2.4.2 Capacitance Calculation for Power Supply Circuit 11 2.5 Simulation 12 2.5.1 Single Pulse Integration 12 2.5.2 Power Supply Output 12 3 Requirements and Verification 13 3.1 Requirements and Verification Table 13 4 Tolerance Analysis 14 5 Safety Statement 15 5.1 General Safety 15 5.2 Laser Safety 16 5.2.1 Bullet Statement 16 5.2.2 Nominal Ocular Distance 16 5.2.3 Laser Safety Calculations 16 6.1 7.8 IEEE Code of Ethics 17 7 Cost and Schedule 17
1
7.1 Cost Analysis 17 7.1.1 Labor 17 7.1.2 Parts 19 7.1.3 Grand Total 20 7.2 Schedule 20 References 22
2
1 Introduction 1.1 Statement of Purpose With roots beginning since the studies of Bell's Inequalities, Quantum Information Science (QIS) is a young, yet established discipline that has broken ground on understanding the nature of information in quantum systems. Bell's Inequality showed that local realism, the principle that physical systems have preexisting "realistic properties" that can be revealed through measurement, failed to reproduce all the predictions of quantum mechanics. Classical models are no longer enough at the quantum level, where intrinsic uncertainty and entanglement both reject any version of local realism. Photons, being easy sources of quantum particles, will likely be important in largescale quantum information processing systems, but current prospective optical quantum information processing experiments are limited by the available methods of detecting complex multiphoton states. A simple quantum information processing system begins first with a source producing a quantum state. Afterwards, the produced quantum state will be manipulated, usually through a system such as quantum logic gates, which may also introduce entanglement on the initial state. Then the state must be detected, obtaining a measurement that may give some classical result such as secure communication between Alice and Bob [1] or efficient computing through quantum algorithms (e.g. integer factoring) [4]. The work presented here will concern the photonnumber resolution stage in detectors (i.e. photon counters). There is a need for reliable detectors that can count photons with high efficiency. In data collection, high detection efficiency η is useful for increasing yield. However, some schemes (such as linear optical quantum computing and loopholefree tests of locality) using highefficiency photonnumber resolving detectors absolutely require detection efficiencies of η > 66 − 83% [4, 5, 6]. Too low of an efficiency can bottleneck in largescale multiphoton experiments due to the probability of detecting N photons scaling as ηN 1 . This problem of
photon detection can be summarized as a decision problem: is it possible to correctly state the 1
If you compare 2 detectors detecting 6 simultaneous photons where one is η = 40% and the other is η = 90%, the 40%efficient detector would have to collect data 130 times as long.
3
presence of a photon incident on a detector given a pulse that may or may not contain a single photon? Considering the problem includes the possibilities of false negatives (a photon is not detected when it is present) and false positives (a photon is detected when it is not present). Current methods of analysis rely on an event from a single photon to trigger a pulse height and a FWHM2 of 1.5ns [3]. However, this presents some difficulties. Photonnumber resolved triggering requires calibration to set the discrimination thresholds between photon numbers, a measurement that could well take upwards of 3060 minutes [3]. Another problem appears when multiple pulses arrive slightly offset in time, skewing the pulse height analysis3. Accuracy is also a concern as the accuracy of the measurement is reliant on the accuracy of the voltage threshold comparison [3]. To take a more robust measurement, we would measure the total charge of an input pulse. An integration of the pulse would thereby resolve the problems inherent in pulse height analysis, as the total charge of a pulse will be constant regardless of the overlap of offset pulses. Creating the pulse height distribution will be much faster as well due to more information being gained from each pulse4 (as it is possible to discriminate empty time bins). Here, we will focus on optimizing the efficiency of photonnumber resolution through pulse area analysis.
1.2 Objectives 1.2.1 Goals & Benefits ➢ Precise method of counting photons by using an op amp integrator ➢ Accurate Measurements when detecting a photon ➢ Fast and Robust Circuitry to continuously integrate 1.2.2 Functions & Features ➢ Upon receiving a single photon, the VLPC will output a pulse approximately 1 nanosecond long with a peak height of 150 mV. ➢ Taking the time integral of the signal yields a more robust measurement over measuring the peak height of the pulse, as each photon has a fixed amount of charge, meaning that 2
Full Width Half Maximum A singlephoton detection event will have a peak height that will be similar across all incidents. Simultaneously detected photons will have a peak height approximately linear with N photons. 4 Approximately 80% of the incoming signals will consist of empty time bins. 3
4
finding the total charge in a given time window will allow us to derive the number of photons. ➢ The calculated integral of the analog signal will be output into a log file which will be made accessible via Raspberry Pi 2 Model B or through software for all future and past measurements.
2 Design 2.1 Block Diagram
2.2 Block Descriptions 2.2.1 Laser Source5 The laser source will be used as a source of interesting quantum states (i.d. photons). Currently, the expected output from the laser source will be 5 to 0 photons at any given pulse. Photons are delivered from the laser source in set time bins (relative arrival time of laser pulses). It is estimated ~80% of the time bins will be empty. 2.2.2 VLPC6 Visible light photon counter (VLPC) was originally developed for highenergy level optics as a visiblelightoptimized version of the infraredsensitive Solid State Photomultiplier (SSPM). The VLPC has many attractive advantages for quantum optics, some of which being: close to perfect intrinsic quantum efficiency (95 ±5% ), fast and lowjitter response (~250 ps 5
Provided by Professor Kwiat in the Department of Physics of the University of Illinois at UrbanaChampaign. The laser source and VLPC can be simulated with a pulse generator for testing prior to experimentation. 6 Provided by Professor Kwiat in the Department of Physics of the University of Illinois at UrbanaChampaign. The laser source and VLPC can be simulated with a pulse generator for testing prior to experimentation.
5
jitter), photon number resolving potential, and a reasonable dark count rate (~10 kHz per m m2 detector area) [3]. Signal from the VLPC will be amplified with a Miteq Au1525 300MHz amplifier with a gain of 60 dB. After amplification, we will expect to see the pulse height of a singlephoton detection to be approximately 150mV and have a FWHM of 1.5 ns. Due to high amplification of the source signal, the VLPC will be extremely sensitive to noise from either of two reasons: 1) incomplete shielding on the cryostat/dunking probe or 2) coupling to the detectors via the heater/temperature sensor wires. However, these issues can be resolved through a wellgrounded Faraday cage and by applying low pass filters to the heater and temperature controller wires. 2.2.3 Integrator Circuit Current methods of resolving the photonnumber of the laser pulses rely on the incoming photon to trigger a certain trigger level (mV) responding to the number of photons for that energy level. However, this is an inherently faulty method as the detector can read the photon as be less than it should be. Our proposed design will utilize op amps in which we plan to integrate an analog signal and calculate the area under the curve. This will give us a much more robust measurement since it will give us a numerical value in which we can use to more accurately count the number of protons that passed per pulse. Even though the peak amplitude of a N photon event will fluctuation, the overall charge (area) under the curve will be approximately linear to the number of photons detected during the pulse. 2.2.4 Analog to Digital Converter The Analog to Digital Converter (ADC) will be used to convert the analog result obtained from the integrator into digital signal. Converted signal will be passed onto processing procedures, where the measured photons will be quantified and analyzed. Since we will want to do all of our digital processing with a Raspberry Pi Model B, we will need an external ADC since the Raspberry Pi does not provide one on the microcontroller itself. Using an 12 bit ADC chip with an I2C protocol, we will use a photocoupler to protect the input of the Raspberry Pi from voltages (noise) fluctuations. The ADC will also need to interface with a NChannel Digital FET to act as a I2C bidirectional level shifter to allow the Raspberry Pi to process various voltage levels. 2.2.5 Data Processing The data processing will be done using a Raspberry Pi 2 Model B. With a quadcore processor, the Raspberry Pi 2 Model B should have a processor speed of around 900 MHz which will be more than fast enough to retrieve our signal from the ADC circuit. Using the GPIO ports on the Raspberry Pi, the digital signal will be input into the Raspberry Pi. Since data will be processed every laser pulse, there will be a log file created to display the integration values from 6
our integrator. To help aid in the data collection process for our log file, a script will be written to constantly read the values of the digital pins of the Raspberry Pi. By taking the digitized value of the integral from the ADC, we will divided that number by the integral of one photon. Using the closest integer value from the division, the photon count will be calculated per pulse. 2.2.6 Graphical User Interface With the processed data, a graphical user interface will be shown on a computer monitor to display the count of the photons per pulse. The interface will also display the average number of photon per count and the overall total number of photons that was calculated over a succession of laser pulses. This count will constantly be updated when we detect a laser pulse.
Figure 1) Graphical User Interface Display 2.2.7 Power Supply Circuit The power source circuit will be taking 120V AC wall outlet as an input and convert it to the voltage ratings required by the parts in the other blocks. Specifically, +/ 5V DC will be provided to the integrator circuit for opamps, and 5V will be provided to ADC. Note that Raspberry Pi will be separately powered via micro USB power supply. First of all, as described in Figure 3, a step down transformer will be incorporated to lower the voltage from 120V AC to more manageable voltage around 12V AC. By choosing a centertapped transformer, it will be possible to generate positive voltages (5V) and a negative voltage (5V) in the same circuit without much inconvenience. Then, the lowered voltages will be halfwave rectified with one diode for positive and negative voltage, respectively. Filtering capacitors of 390 uF (+5V) and 220 uF (5V) were chosen to smooth out the ripples of rectified voltage. The reasoning behind this selection is described in Section 2.4.2. The filtered voltage will be fed into the linear regulators, which will output stable voltages. The additional 10 uF filter capacitors were added in parallel to the output, as recommended by the manufacturers.
7
2.2.8 EMI Shielded Enclosure The PCB will be protected by a metal conductor enclosure of aluminium diecast. This design will have a primary benefit of reducing interference on the circuit from external sources, but will have a secondary benefit of physically securing the circuit. Other designs, such mesh or plastics with copper coating, were considered but were found more permeable at higher frequencies. Holes will be made input and output wires and power, but the holes will be designed as small as possible to reduce tampering to the effectiveness of the EMI enclosure. To limit contact with the enclosure and ensure pieces from being damaged within the box, the PCB will be suspended through nonconductive foam padding.
Figure 2) Hammond R111083000 [10]
8
2.3 Circuit Schematic 2.3.1 Integrator Circuit
Figure 3) Integrator Circuit with Generic Op Amp [8]
9
2.3.2 Analog to Digital Converter Circuit
Figure 4) Analog to Digital Converter with Raspberry Pi
2.3.3 Power Supply Circuit
Figure 5) Power Supply Circuit with No Load
10
2.4 Calculation 2.4.1 Integration for Single Pulse
11
2.4.2 Capacitance Calculation for Power Supply Circuit
12
2.5 Simulation 2.5.1 Pulse Integration at Multiple Voltages
Figure 6) Pspice Simulation of Integration Circuit with Pulse Function Blue is the input pulse voltage at 150mV, 300mV, and 450mV. Green is the voltage reflected across the capacitor in the integrator circuit.
2.5.2 Power Supply Output A series of simulations was conducted based on the circuitry shown in Figure 5. Our goal was to find out whether rectifiers and regulators are capable of generating the specified voltages (+/ 5 V), given stepped down voltages (12V AC) from secondary windings of the centertapped transformer. As shown in Figure 7, LTSPICE simulation result demonstrates that the circuit outputs the correct, stable values.
13
Figure 7) LTSPICE Simulation of Power Supply Circuit
3 Requirements and Verification 3.1 Requirements and Verification Table Block
Requirement
Verification
Points
Integrator
1. Integrates a delivered signal within a 5 ns range +/ 1 ns
1.Verification Process for Item 1: 50 a) Generate a periodic mock pulse (sawtooth… etc.) using a function generator. b) Observe a transient response of the integrator via an oscilloscope c) Integrates a given signal within a 5 ns time from by analyzing output saved in a log file
Power Supply
1. Provides a set of required
1.Verification Process for Item 1:
10
14
voltage (+/ 5V) in a stable manner. Output ripple must be under 1% and DC values should be within 5% of proposed voltages.
Data Processing
1. Process the data and calculate the total photon count per pulse with an accuracy of +/ 4 photons
a) Connect the power supply circuit to 120V AC wall outlet. b) Measure and observe the output voltages when there is no load via an oscilloscope. c) Measure the output voltages when there is load. (i.e, when the circuits are all connected and operating) via an oscilloscope. 1.Verification Process for Item 1: a) Use channel 1 of an oscilloscope to view the output of the integrator. b) Use channel 2 of the oscilloscope to view the output of the ADC (before data processing). c) Compare channel 1 and channel 2 to see that one analog signal corresponds to one digital signal. d) Observe and compare photon count on the GUI (post processing). e) Compare the change in count on the GUI with the output signal from channel 1 on the oscilloscope.
40
4 Tolerance Analysis Our critical component for our integrator is the operational amplifier which ensures speedy and reliable integration. Our accepted tolerance is such that the integrator circuit can reliably continue to integrate within an approximately 5ns long time window. 15
We will be using the OPA847 operational amplifier which has a high slew rate of 950V/µs . Using dimensional analysis:
4300V 4300V μs 4300mV = = 4300mV μs μs 1000ns V ns which shows that the OPAMP is suitable for our needs. Approximating the peak voltage magnitude of one photon to be 150mV, we expect the maximum detected peak (of 5 photons) to be 750mV. We can also calculate the minimum slew rate via equation: Slew Rate = 2πfV where: f is the frequency of the swing V is the amplitude voltage We use 200MHz as the frequency as our sampling time window is approximated to be 5ns. We use 750mV as the amplitude because that is the highest magnitude voltage we expect to see. Using the equation stated above, we calculate our minimum slew rate to be 942.478 mV ns , which shows that our OPAMP meets our needs.
5 Safety Statement 5.1 General Safety ● All group members will receive appropriate laboratory training as required by the coursework ● Voltage/power rating of components in the circuitry and power supply will be checked and confirmed prior to implementation. ● In order to ensure the safety of transformer, a fuse will be added in the power supply circuit.
16
5.2 Laser Safety 5.2.1 Bullet Statement ● All members of Group 48 will be trained and authorized to operate a laser to the specification of Professor Kwiat. ● This project requires a laser as a primary source therefore, protective eyewear is necessary to ensure safe interactions. ● All reflective watches and jewelry will be removed before operation with a laser. ● A laser will not be left unattended to ensure safety of every individual in Professor Kwiat’s lab.
5.2.2 Nominal Ocular Hazard Distance Nominal Ocular Hazard Distance (NOHD) is the calculated distance from a photon source at which the intensity (energy per surface area) becomes lower than the Maximum Permissible Exposure (MPE) on the cornea and on the skin. In layman's terms, NOHD is the point where the viewing distance becomes eyesafe. The calculation of the NOHD is necessary to determine the Nominal Hazard Zone (NHZ), which is the area inside which personnel are at risk of ocular damage due to exposure from the laser. Anyone within the NHZ is required to wear the appropriate eyewear for protection by ANSI Std. z136.6.
5.2.3 Laser Safety Calculations
−1
R NOHD = θ
√
4ΦP π*MP E
− dout 2
where R NOHD is the Nominal Ocular Hazard Distance ( cm ) θ is the beam divergence ( radians ) ΦP is the beam output radiant power ( watts ) M P E is the Maximum Permissible Exposure ( watts/cm2 ) dout is the output beam diameter of the laser ( c m )
R NOHD = (0.4 millirad)−1
√
4(1 mW ) π*0.001W /cm2
− (0.1 cm)2
17
R NOHD = 856.3146 cm ≈ 8.56 metres
6 Ethical Considerations 6.1 7.8 IEEE Code of Ethics High frequency signals tend to be very sensitive. Our design will have to take into consideration that physical circuits are not ideal. Our calculations and simulations govern our design at its best and that physically constructing it in hardware may lead to undesirable results. Item 3 of section 7.8 of the IEEE Code of Ethics states “to be honest and realistic in stating claims or estimates based on available data [9].” We have to be mindful and accept any faulties in our hardware that could alter the results of our simulation and calculation. As the capstone class to end our undergraduate experience, we have to acknowledge and accept any feedback given to us from our teaching assistant, instructors, and classmates. We have to keep in mind that any criticism or suggestion we are given is only meant to help us achieve our design goals for this project. Item 7 of section 7.8 of the IEEE Code of Ethics states “to seek, accept, and offer honest criticism of technical work, to acknowledge and correct errors, and to credit properly the contributions of others [9].” As a continuation of a project that was attempted Spring 2015, we will revise and credit any contributions that were previously made in hope of a more refined and successful design. Since we will be conducting experiments in Professor Kwiat’s lab, we have to respectful of his equipment and laboratory. Working with a laser could also be hazardous so we will be abiding by all safety precautions specified by Professor Kwiat. Under item 9, it states” to avoid injuring others, their property, reputation, or employment by false or malicious action [9].” We will ensure that our group along with all colleagues in the lab are safe while we are conducting experiments and debugging in his lab.
18
7 Cost and Schedule 7.1 Cost Analysis 7.1.1 Labor
Name
Hourly Rate
Total Hours Invested
Total = Hourly Rate x 2.5 x Total Hours Invested
Austin
$34.50
225
$19,406.25
Mike
$34.50
225
$19,406.25
Katie
$34.50
225
$19,406.25
Total
675
$58,218.75
7.1.2 Parts Item Name
Manufacturer Name/ Part Number
Quantity
Unit Cost
Cost
High Speed Operational Amplifier
Texas Instrument/ OPA8471DBV
1
$6.19
$6.19
EMI Shielding
Hammond Manufacturing/ R111083000
1
$56.73
$56.73
1K Resistor
Bourns/ CHV2512FX1004ELF
2
$0.61
$1.22
10K Resistor
Panasonic/ ERA8AEB103V
4
$0.79
$3.16
1.5 uF Capacitor
Vishay/ VJ0805V155ZXJTW1BC
3
$0.10
$0.30
100 nF Capacitor
TDK/ CGA2B3X7R1H104K050 BE
1
$0.20
$0.20
10 uF Capacitor
Murata Electronics/ GRM155R60J106ME44D
4
$0.33
$1.32
Panasonic/
1
$0.84
$0.84
390 uF Capacitor
EEEFK1E391SP
19
220 uF Capacitor
Panasonic/ EEEFC1E221P
1
$0.85
$0.85
Transistor Output Optocouplers Prototrans 4Channel
LiteOn/ LTV847
1
$0.70
$0.70
24V CT 3A Power Transformer (12V012V)
Parts Express/ 671243, 120210
1
$24.80
$24.80
Schottky Diodes & Rectifiers 30V 5A
NXP Seimiconductors/ PMEG3050EP
2
$0.48
$0.96
Linear Regulator (Negative 5V)
Linear Technology/ LT1964ES55#TRMPBF
1
$2.02
$2.02
Linear Regulator (Positive 5V)
Texas Instrument/ LM1085ISX5.0/NOPB
1
$2.09
$2.09
Linear Regulator (Positive 3.3V)
Texas Instrument/ LM1085ISX3.3/NOPB
1
$2.09
$2.09
1 x 40 Header Pins
All Electronics
1
$0.85
$0.85
4G MicroSD Card
Sandisk
1
$2.98
$2.98
HDMI to HDMI Cable
Adafruit
1
$4.95
$4.95
USB Cable A/MicroB
Adafruit
1
$2.95
$2.95
GPIO Ribbon Cable for Raspberry Pi Model B
Adafruit
1
$2.95
$2.95
Raspberry Pi 2 Model B
Adafruit
1
$35.00
$35.00
ADC with I2C Interface
Microchip/MCP3424
1
$2.92
$2.92
Dual NChannel, Digital FET
FairChild/FDC6301N
1
$0.44
$0.44
USA Raspberry Pi Micro USB Power Supply Charger 5v 1500mA
The Pi Hut
1
$5.19
$5.19
Total
$161.70
20
7.1.3 Grand Total Section
Total
Labor
$58,218.75
Parts
$161.70
Grand Total
$58,380.45
7.2 Schedule Week
Task
Responsibility
02/08/2016
Finalize Project Proposal Design Circuit for ADC on Pspice Design Circuit for Integrator on Pspice Design Circuit for Power Supply on Pspice
All KN MK AJ
02/15/2016
Finalize Document for Mock Design Review Revise/Simulate Design for ADC on Pspice Revise/Simulate Design for Integrator on Pspice Design Power Source (Stable Voltage Setup)
AJ, KN KN MK AJ
02/22/2106
Simulate Power Supply Circuit on Breadboard Simulate ADC Circuit on Breadboard Simulate Integrator Circuit on Breadboard Prepare Final Documentation Design Review
AJ KN MK AJ, KN
02/29/2016
Create Integrator Circuit PCB on Eagle Create ADC Circuit PCB on Eagle and Order Parts Create Power Supply Circuit PCB on Eagle
MK KN MK
03/7/2016
Setup Raspberry Pi Environment Write Code for Graphical User Interface Prepare Equipment Setup in Kwiat’s Lab
KN MK AJ
03/14/2016
Solder and Verify Integrator in Senior Design Lab Solder and Verify Power Supply in Senior Design Lab Solder and Verify ADC in Senior Design Lab Initial Integrator Testing in Kwiat’s Lab
MK AJ KN All
03/21/2016
Spring Break
03/28/2016
Test / Debug Integrator in Kwiat’s Lab
AJ
21
Test / Debug ADC with Raspberry Pi in Kwiat’s Lab Write/Debug Data Processing Code
KN MK
04/04/2016
Final Testing / Debugging for Entire Design
All
04/11/2016
Mock Demo Optimize / Fix Remaining Issues with Design
All All
04/18/2016
Optimize /Fix Remaining Issues with Design Prepare for Final Demonstration Prepare Final Paper
All AJ, MK KN
04/25/2016
Final Demonstration Mock Presentation Prepare Final Paper Prepare Final Presentation
All All AJ, MK KN
05/2/2016
Final Presentation Finalize Final Paper Lab Checkout
All All All
22
References [1]
[2]
[3]
[4] [5]
[6] [7]
[8]
S. Lomonaco, "A Talk on Quantum Cryptography or How Alice Outwits Eve", Coding Theory and Cryptography , pp. 144174, 2000. J. Kim, K. McKay, P. Kwiat, K. Zielnicki and E. Gansen, "Novel Semiconductor SinglePhoton Detectors", Experimental Methods in the Physical Sciences , pp. 147183, 2013. K. Zielnicki, “Pure Sources and Efficient Detectors for Optical Quantum Information Processing”, Ph.D. Dissertation, Dept. Phys., University of Illinois at UrbanaChampaign, Champaign, IL, 2014. E. Knill, R. Laflamme and G. Milburn, "A scheme for efficient quantum computation with linear optics", Nature , vol. 409, no. 6816, pp. 4652, 2001. P. Eberhard, "Background level and counter efficiencies required for a loopholefree EinsteinPodolskyRosen experiment", Phys. Rev. A , vol. 47, no. 2, pp. R747R750, 1993. P. Kwiat, P. Eberhard, A. Steinberg and R. Chiao, "Proposal for a loopholefree Bell inequality experiment", Phys. Rev. A , vol. 49, no. 5, pp. 32093220, 1994. Waks, E., Inoue, K., Oliver, W. D., Diamanti, E., & Yamamoto, Y. Highefficiency photonnumber detection for quantum information processing. Selected Topics in Quantum Electronics, IEEE Journal of , 9 (6), 15021511, 2003. P.Tapashetti, A. Gupta, C. Mithlesh, A.S Umesh, "Design and Simulation of Op Amp Integrator and Its Application", International Journal of Engineering and Advanced
[9]
Technology , pp. 1219, 2012. ieee.org, "IEEE IEEE Code of Ethics", 2016. [Online]. Available: http://www.ieee.org/about/corporate/governance/p78.html.
[10]
Hammond Manufacturing, “Hammond Manufacturing R111083000”, R111083000 datasheet, 2016.
23
