Dark Current Cancellation: The Art of Removing Darkness from the Signal |
Executive Summary |
This article provides a comprehensive, accessible exploration of dark current cancellation in barcode reader photodetector circuits. We examine why dark current is a critical problem that can corrupt the tiny signal from a photodiode and how leading companies have developed ingenious techniques to remove its effects. Rather than focusing on abstract theory, we ground every concept in concrete design examples and real products from industry leaders including Texas Instruments, Renesas, and various pioneering patent holders. We explore the physical origins of dark current, the use of shielded reference photodiodes, the design of differential amplifiers for subtraction, the application of current mirrors, and the implementation of digital dark frame subtraction. The article covers both analog and digital approaches, with special attention to the practical trade-offs between cost, complexity, and cancellation accuracy. The closing summary synthesizes the key lessons and offers practical guidance for anyone designing circuits that must distinguish signal from darkness. |
|
Chapter 1: The Problem of Darkness |
Even in complete darkness, a photodiode produces a small current. This current, known as dark current, flows through the device when no light is present. It is caused by thermal generation of electron-hole pairs within the semiconductor material. The dark current is typically in the nanoamp range, but it increases with temperature and with reverse bias voltage. |
Dark current is the enemy of precision optical sensing. In a barcode reader, the photodiode generates a current that is the sum of the desired photocurrent (from the reflected barcode light) and the dark current. The dark current adds a DC offset to the signal, shifting the baseline and potentially causing decoding errors. It also contributes noise, reducing the signal-to-noise ratio. |
The challenge is that dark current is not constant. It varies with temperature, typically doubling with every 8 to 10 degrees Celsius. It also varies from device to device, due to manufacturing tolerances. A dark current cancellation circuit must compensate for these variations, removing the dark current from the signal without affecting the photocurrent. |
A patent from a barcode reader manufacturer explains the importance of dark current compensation: 'Photodiodes operating at different biasing points can also have dark currents. Various embodiments of device can be designed such that the dark photodiode and the signal photodiode can operate at approximately the same bias point.' This ensures that the dark current from the reference photodiode accurately represents the dark current in the signal photodiode. |
|
Chapter 2: The Shielded Reference Photodiode |
The most common technique for dark current cancellation is to use a shielded reference photodiode. This is a photodiode that is identical to the signal photodiode but is covered with a non-light-permeable material, such as metal. The shielded photodiode produces only dark current, with no photocurrent component. |
A patent describes this approach in detail: 'The dark diode is formed using a photo diode that is equivalent (or identical) to photodiode, where the dark diode has its photo detecting area covered with metal or other non-light permeable material. With no photons from any source reaching the photo detecting area of dark diode, the signal produced by dark diode is based only on the environment in which the dark diode sits.' |
The key assumption is that the dark current in the shielded photodiode is the same as the dark current in the signal photodiode. This requires that the two photodiodes be matched in their electrical characteristics and that they be positioned close together so they experience the same temperature and bias conditions. The patent explains that 'with dark diode being positioned near photodiode, the dark current of both the dark diode and the photodiode should be substantially equal.' |
|
Chapter 3: The Subtractor Circuit |
Once the signal photodiode and the shielded reference photodiode are generating their respective currents, the next step is to subtract the dark current from the signal current. This is typically done with a subtractor circuit, also known as a differential amplifier. |
The patent describes a circuit that includes 'a subtractor that subtracts the current from the dark diode from the current from the photodiode to generate a signal that includes substantially only the photocurrent.' The subtractor takes the two currents as inputs and produces an output that is the difference between them. Since the dark current is common to both, it is canceled in the subtraction. |
The output of the subtractor is a current that is proportional to the light intensity, with the dark current removed. This current can then be amplified by a transimpedance amplifier and further processed. The patent notes that the light sensor of FIG. 4 'provides improved light sensing performance over a range of environmental parameters, such as varying volume and light source temperatures.' |
|
Chapter 4: Texas Instruments' OPT101 - A Monolithic Solution |
Texas Instruments offers the OPT101, a monolithic photodiode with an on-chip transimpedance amplifier, as a practical solution for barcode readers and other optical sensing applications. The device integrates the photodiode and the amplifier on a single chip, eliminating many of the problems associated with discrete designs. |
The OPT101 datasheet highlights its key features: 'The integrated combination of photodiode and transimpedance amplifier on a single chip eliminates the problems commonly encountered in discrete designs, such as leakage current errors, noise pick-up, and gain peaking as a result of stray capacitance.' The photodiode operates in photoconductive mode for excellent linearity and low dark current. |
While the OPT101 does not include a dedicated dark photodiode, its integrated design reduces leakage current errors by minimizing the exposed high-impedance nodes that can pick up noise and leakage. The on-chip amplifier is specifically designed for single-supply operation from 2.7 to 36 volts, with a quiescent current of only 120 microamps. The device is suitable for barcode scanners, position sensors, and other applications requiring precise light measurement. |
|
Chapter 5: Renesas' ISL29009 - Dark Current Compensation on a Chip |
Renesas offers the ISL29009, a light-to-current silicon optical sensor that integrates dark current compensation. The device combines a photodiode and current amplifiers on a single monolithic IC. |
The ISL29009 datasheet highlights its dark current compensation feature: 'A dark current compensation circuit aids the photodiode to minimize temperature dependent leakage currents in the absence of light, improving the light sensitivity at low lux levels.' This on-chip compensation ensures that the device provides accurate light readings even in dim conditions where dark current could otherwise dominate the signal. |
The device's photodiode has a human-eye-like spectral response, with peak sensitivity at 550 nanometers. The integrated non-linear current amplifier boosts and converts the photodiode signal in a square root fashion, extending the dynamic range while maintaining sensitivity at low light levels. The ISL29009 is designed for display dimming in mobile devices and other applications where low power and small size are essential. |
Chapter 6: Dark Current Subtraction with Current Mirrors |
A patent describes an alternative approach to dark current cancellation that uses a current mirror. In this approach, the dark current from a shielded photodiode is mirrored and used to compensate for the dark current in the signal path. |
The patent explains: 'The dark current loop includes a reference generating circuit, a voltage regulator, the dark diode, and the current mirror circuit. The current mirror circuit connects the voltage regulator circuit and the dark photodiode to the signal diode, an IDAC, the TIA, and the feedback circuit.' The current mirror copies the dark current from the reference photodiode and injects a compensation current into the signal path. |
The current mirror is designed with a transfer ratio that ensures it does not over-compensate. The patent notes that 'the current mirror can be configured with a transfer ratio that results in the mirrored current that is less than or equal to the dark current when taking into consideration the manufacturing tolerances.' This prevents the mirror from removing part of the photocurrent along with the dark current. |
|
Chapter 7: The IDAC for Fine-Tuning |
In systems that use a current mirror for dark current cancellation, an IDAC (current digital-to-analog converter) is often used for fine-tuning. The IDAC compensates for the residual dark current that remains after the mirror's compensation. |
The patent describes the IDAC's function: 'The IDAC can be configured to compensate for the DC offset remaining in the signal current, so that only the photocurrent can enter a transimpedance amplifier.' The IDAC receives a feedback signal from the amplifier and generates a compensation current that removes the residual DC offset. |
The IDAC is a digital-to-analog converter that produces an analog current output in response to a digital input. The patent explains that 'the IDAC can be configured to convert a digital input, e.g., a binary number, into an analog output.' This allows the system to be calibrated digitally, with the compensation current set by a digital value. |
The use of an IDAC provides several advantages. It allows for precise, adjustable compensation that can be tuned to account for device variations. It also enables digital calibration, which can be performed during manufacturing or in the field. |
Chapter 8: Dark Field and Bright Field Subtraction |
A patent from Datalogic describes a different approach to dark current cancellation for image sensors. Instead of using a separate shielded photodiode, this approach captures a dark image when the illumination is off and subtracts it from the illuminated image. |
The patent explains: 'The illuminated image includes both desired data about the optical code to be read and fixed pattern noise that is approximated by the dark image. Accordingly, a compensated image is generated by subtracting the stored dark image from the illuminated image.' This is a digital technique, performed in the processor rather than in analog circuitry. |
The dark image is captured with a short exposure time, while the illuminated image is captured with a much longer exposure time. The patent notes that 'the exposure time of the illuminated image is between 100 and 1,000 times longer than the exposure time of the dark image.' This ensures that the dark image captures only the fixed pattern noise, while the illuminated image captures the barcode information. |
This technique is particularly useful for imagers with fixed pattern noise (FPN) caused by pixel-to-pixel variations. The dark image captures this noise, and it is subtracted from the illuminated image. The patent describes how this can be implemented on a linear imager: 'Capture a dark linear image using exposure time T1. Digitize and store pixels from dark linear image into array. Capture an illuminated linear image using exposure time T2. Digitize and store a fixed pattern noise compensated image.' |
|
Chapter 9: Two-Dimensional Dark Frame Subtraction |
The dark frame subtraction technique can be extended to two-dimensional image sensors. The Datalogic patent explains that 'method 100 is applicable, in various embodiments, to compensate for fixed pattern noise in both linear and two-dimensional imagers.' |
For two-dimensional imagers, the dark image is a full frame captured when no light is present. The illuminated image is captured with the illumination on. The dark image is then subtracted pixel-by-pixel from the illuminated image. This removes the fixed pattern noise that is common to both images. |
The patent notes that this technique can be performed using virtual scan lines: 'These techniques may also be used with virtual scan line arrays, where the dark image is a one-dimensional virtual scan line array and the illuminated image is a one-dimensional virtual scan line array.' This reduces the computational requirements, as only the scan lines of interest are processed. |
The patent also describes an advanced technique where two dark images are captured with different exposure times. This allows the system to separate threshold mismatch dark signal non-uniformity from dark current dark signal non-uniformity. The first dark image captures threshold mismatch, while the second dark image captures both threshold mismatch and dark current. By subtracting the first from the second, the dark current component can be extracted. |
Chapter 10: Temperature Compensation for Dark Current |
Dark current is strongly temperature-dependent, typically doubling with every 8 to 10 degrees Celsius. A dark current cancellation system that does not account for temperature will become inaccurate as the temperature changes. |
The Datalogic patent addresses this issue: 'Since dark current DSNU is influenced by temperature, in one embodiment the optical code reader may periodically determine the temperature, and when a change in temperature is detected, the dark image may be reacquired.' By reacquiring the dark image whenever the temperature changes, the system ensures that the dark current estimate is always accurate. |
Another patent describes a more sophisticated approach: 'The compensation circuit is configured to modify the dark current value of the shielded photosensitive device to provide a temperature compensated dark current value.' This uses a temperature sensor and a calibration factor stored in memory to adjust the dark current estimate based on temperature. |
The patent notes that 'dark current may be strongly temperature dependent. One approximation is that dark current doubles every 8 degrees Celsius.' A compensation circuit can use this relationship to predict the dark current at any temperature, based on a measurement at a known temperature. |
|
Chapter 11: The Problem of Photodiode Mismatch |
One of the challenges of dark current cancellation is that the signal photodiode and the shielded reference photodiode may not be perfectly matched. Manufacturing tolerances mean that even identical photodiodes can have slightly different dark currents. |
A patent addresses this problem: 'Manufacturing tolerances are present in nearly all manufactured electrical devices including photodiodes. Therefore, in the embodiments where the dark photodiode is designed to match the signal photodiode, the dark current from the dark photodiode can be more or less than the dark current from the signal photodiode.' |
To avoid over-compensation, the current mirror is designed with a transfer ratio that ensures it cannot remove the photocurrent. The patent explains: 'To avoid the dark current loop from compensating for the photocurrent, the current mirror circuit can be configured to mirror less than the dark current from the dark photodiode.' This ensures that even if the dark currents are mismatched, the photocurrent is not affected. |
The residual dark current that remains after the mirror's compensation is then removed by the IDAC, which can be adjusted to provide exactly the right compensation. This two-stage approach provides robust compensation even with mismatched photodiodes. |
Chapter 12: Switch-Based Transient Reduction |
When a pulsed light source is used for illumination, the photodiode can become saturated during the light-on period. This saturation can cause a long settling time after the light is turned off, making it difficult to measure ambient light during the dark period. |
A patent addresses this problem with a bypass switch that shorts the photodiode during the light-on period. The patent explains: 'A bypass switch is configured to bypass the first light sensor at a first point in time when the light source is on, and to not bypass the first light sensor at a second point in time when the light source is off.' |
When the bypass switch is closed, it provides a short circuit from the photodiode to ground, allowing the generated current to flow away without saturating the photodiode. 'When the light source is on, a substantial portion of current generated by the first light sensor as a result of the light source is shorted to ground.' This prevents saturation and reduces the settling time. |
The patent notes that 'the settling time of a light sensor is proportional to the level of light detected. As measured light decreases, the settling time increases.' To combat this in low-light scenarios, a reference current is introduced, and the subtraction operation is reversed. This results in shorter settling times in low light detection scenarios. |
|
Chapter 13: The Differential Amplifier in Color Sensors |
A patent describes a color sensor that uses a differential amplifier to cancel dark current. The color sensor includes separate photodiodes for red, green, and blue, plus a shielded dark photodiode. |
The patent explains: 'The color sensor includes a dark color sensor circuit. The dark color sensor circuit produces dark photocurrent and outputs a dark current offset voltage. The dark color sensor circuit converts the dark photocurrent into some form of reference offset voltage for later cancellation from a corresponding voltage output from the color sensors.' |
A differential amplifier subtracts the dark current voltage from each color channel: 'The differential amplifier circuit receives the voltage output from the color sensor circuit and the dark current voltage offset from the dark color sensor circuit and outputs a final output that cancels contributions of the offset voltage due to dark photocurrent.' |
The patent claims that this approach 'cancels contributions of the offset voltage due to dark photocurrent' and 'increases the temperature robustness of the color sensor.' By using a single dark photodiode for all three color channels, the sensor saves space and cost. |
|
Chapter 14: Current Subtraction vs. Voltage Subtraction |
Dark current cancellation can be performed in either the current domain or the voltage domain. Current subtraction is typically performed before the transimpedance amplifier, while voltage subtraction is performed after the amplifier. |
The differential amplifier approach described in the color sensor patent performs voltage subtraction. The output voltages from the signal and dark photodiodes are amplified by separate transimpedance amplifiers and then subtracted by a differential amplifier. |
The current mirror approach described in another patent performs current subtraction. The dark current from the shielded photodiode is mirrored and injected into the signal path before the transimpedance amplifier. This removes the dark current in the current domain, so the transimpedance amplifier only sees the photocurrent. |
Current subtraction has the advantage of removing the dark current before it is amplified, reducing the dynamic range requirements of the transimpedance amplifier. Voltage subtraction is simpler to implement and can be used with standard operational amplifiers. |
Chapter 15: The Adaptive Bias Technique |
A patent from a barcode reader manufacturer describes an adaptive biasing technique for photodiodes that reduces dark current. The technique uses a current source to provide a small reverse bias voltage across the photodiode during low light conditions, and a higher reverse bias during high light conditions. |
The patent explains: 'The adaptive biasing circuit produces a small reverse bias voltage across the photosensor during low photodiode current measurements, reducing the photodiode's dark current and improving the measurement's accuracy. During the measurement of high photodiode currents, a higher reverse-bias voltage is developed across the photodiode, reducing the photodiode's effective capacitance, increasing the circuit bandwidth, and providing a faster transient response.' |
This technique provides the benefits of reverse bias (reduced capacitance, faster response) while minimizing the dark current penalty. By reducing the bias voltage when the photocurrent is low, the dark current is minimized. By increasing the bias voltage when the photocurrent is high, the capacitance is reduced, allowing faster response. |
|
Chapter 16: Dark Current in Laser-Based Scanners |
Laser-based barcode readers face the same dark current challenges as imaging readers. However, the scanning nature of laser readers allows some unique approaches to dark current cancellation. |
A patent for a laser-based barcode reader describes using the laser's off-time to capture a dark field signal. The patent explains: 'During the off-time of the laser, a dark field signal is captured and stored. This dark field signal represents the ambient light noise. During the on-time of the laser, a combined signal is captured. The dark field signal is then subtracted from the combined signal to extract the laser reflection signal.' |
This approach is analogous to dark frame subtraction in imaging readers. By capturing a dark signal during the laser off-time, the system obtains an estimate of the ambient light and dark current. This estimate is then subtracted from the signal captured during the laser on-time. |
The patent notes that this approach can be implemented with a single photodiode, rather than requiring a separate shielded reference photodiode. This reduces the cost and complexity of the reader. |
Chapter 17: Fixed Pattern Noise Compensation in Imagers |
In imagers, dark current appears as fixed pattern noise (FPN) --- a fixed pattern of pixel-to-pixel variations that is present in every image. FPN is caused by variations in dark current and threshold mismatch among the pixels in the image sensor. |
The Datalogic patent describes a method for compensating for FPN in an optical data reader. The method involves capturing a dark image (with the illumination off), capturing an illuminated image (with the illumination on), and subtracting the dark image from the illuminated image. |
The patent states: 'The dark image comprises minimal scene information and an estimate of fixed pattern noise. The illuminated image comprises optical data from a read area. Subtracting the dark image from the illuminated image generates a compensated image comprising optical data that is, at least partially, compensated for fixed pattern noise.' |
The patent also describes a more advanced method that separates threshold mismatch dark signal non-uniformity from dark current dark signal non-uniformity. This allows the system to apply different compensation for these two sources of FPN, improving the overall compensation accuracy. |
|
Chapter 18: The Dark Image Capture Process |
The capture of a dark image is a critical step in dark frame subtraction. The dark image must be captured with minimal scene information so that it accurately represents the fixed pattern noise. |
The Datalogic patent explains: 'Special equipment and special conditions are not required in order to obtain an estimate of FPN using a dark image. For example, in certain embodiments using a trigger pull scanner, a new dark image may be obtained each time the trigger is pulled.' This ensures that the dark image is always up to date. |
The dark image is typically captured with a short exposure time. The patent notes that 'the first exposure time is between 5 microseconds and 30 microseconds' for a linear imager. This short exposure minimizes the contribution of scene information, while still capturing the fixed pattern noise. |
The illuminated image is captured with a much longer exposure time. The patent states that 'the exposure time of the illuminated image is between 100 and 1,000 times longer than the exposure time of the dark image.' This ensures that the illuminated image captures sufficient light to read the barcode. |
|
Chapter 19: Temperature Compensation for Dark Current |
A patent describes a system for compensating for dark current that uses temperature information to improve cancellation accuracy. The system uses a compensation circuit that modifies the dark current value based on temperature. |
The patent states: 'The compensation circuit is configured to modify the dark current value of the shielded photosensitive device to provide a temperature compensated dark current value.' This is more sophisticated than simple subtraction, as it accounts for the temperature dependence of dark current. |
The patent notes that 'dark current may be strongly temperature dependent. One approximation is that dark current doubles every 8 degrees Celsius.' By using temperature information, the system can predict the dark current more accurately, improving compensation. |
The patent describes a system where 'a first one-time programmable memory is provided which holds a compensatory factor. A second one-time programmable memory provides two compensation coefficients.' These coefficients are used to calibrate the temperature compensation for the specific photodiodes in the system. |
Chapter 20: The Feedback Circuit for DC Offset Control |
A patent describes a feedback circuit that controls the DC offset in a transimpedance amplifier. The feedback circuit senses the offset voltage at the amplifier output and adjusts a compensation current to cancel the offset. |
The patent explains: 'The feedback circuit can be configured with control logic that sends a digital signal informing the IDAC the amount the voltage inside the TIA is offset after the signal enters the TIA.' The IDAC then produces a compensation current that removes the offset. |
This feedback loop allows the system to compensate for DC offset that varies with temperature or aging. The feedback circuit continuously monitors the offset and adjusts the compensation current as needed. |
The patent notes that 'the IDAC can be configured to compensate for the DC offset remaining in the signal current.' This includes not only dark current but also other sources of DC offset, such as amplifier offset voltage and leakage currents. |
|
Chapter 21: The Dark Current Problem in High Sensitivity Applications |
In high sensitivity applications, such as long-distance barcode reading, the photocurrent from the barcode reflection can be very small. In these cases, dark current can be a significant fraction of the signal. |
A patent explains: 'The improvement is typically 2x and may not be enough to mitigate for the 100x reduction in light level due to for example the housing of the sensor.' This illustrates the magnitude of the dark current problem in high sensitivity applications. |
In these applications, dark current cancellation must be very accurate. Even a small residual dark current can degrade the signal-to-noise ratio, making decoding difficult. |
The patent describes a system that uses 'temperature information and temperature dependent calibration information to compensate for dark current.' By using calibration data stored in memory, the system can provide accurate dark current cancellation across a wide temperature range. |
Chapter 22: The Digital Domain Advantage |
Subtracting dark current in the digital domain offers several advantages over analog subtraction. Digital subtraction is more precise, more flexible, and less sensitive to component variations. |
A patent explains: 'This arrangement has dark current subtraction in the digital domain. This contrasts with the arrangement of FIG. 1 where there is dark current subtraction in the analog domain.' The digital approach uses an analog-to-digital converter to digitize the signals before subtraction. |
Digital subtraction allows the use of sophisticated algorithms, such as averaging multiple dark images to reduce noise. The patent notes that 'averaging the plurality of preliminary dark images in order to generate the dark image' can improve the accuracy of the dark current estimate. |
Digital subtraction also allows the compensation to be adjusted dynamically based on conditions. For example, the system can recapture the dark image when a temperature change is detected, ensuring that the dark current estimate is always accurate. |
|
Chapter 23: The Optical Data Reader Architecture |
A patent describes the architecture of an optical data reader that incorporates fixed pattern noise compensation. The reader includes an image reading engine, a processor, and a fixed pattern noise compensation subsystem. |
The patent explains: 'The optical data reader may incorporate an image reading engine to read optically encoded symbols. The image reading engine may encompass a CCD-based or CMOS-based imager.' The processor executes programs to control the operation of the reader. |
The fixed pattern noise compensation subsystem subtracts the dark image from the illuminated image. The patent states: 'The processor may, for example, be configured to perform the subtraction of a dark image from an illuminated image to generate an image that is, at least partially, compensated for fixed pattern noise.' |
The reader may also include a temperature sensor. The patent notes: 'The portable data reader may further include a temperature sensor. Information from the temperature sensor may be utilized in calculating an estimate of dark current DSNU.' This allows the compensation to be temperature-aware. |
Chapter 24: Dark Current in PIN Photodiodes |
PIN photodiodes, which are commonly used in barcode readers, have dark current characteristics that must be considered in circuit design. The dark current in a PIN photodiode is typically in the nanoamp range, but it increases with temperature. |
The datasheet for the BPW34FAS, a popular PIN photodiode, specifies a dark current of 2 nanoamps. The Hamamatsu S8385 has a dark current of 0.1 to 1 nanoamp. These low dark currents are important for barcode reading, especially when the reader is used in warm environments. |
The dark current doubles with every 10 degrees Celsius increase in temperature. This means that a photodiode with 1 nanoamp of dark current at 25 degrees Celsius will have 2 nanoamps at 35 degrees Celsius and 4 nanoamps at 45 degrees Celsius. |
This temperature sensitivity must be considered in dark current cancellation circuit design. A circuit that provides perfect cancellation at one temperature may be inaccurate at another temperature. |
|
Chapter 25: The IDAC's Role in Dark Current Compensation |
The IDAC (current digital-to-analog converter) is a key component in many dark current compensation circuits. It provides a precise, digitally controlled current that can be used to cancel dark current. |
The patent describes the IDAC's function: 'The IDAC can be configured to produce a compensation current in response to the digital signal, thereby removing the residual dark current from the signal current.' The compensation current is subtracted from the signal current, removing the dark current component. |
The IDAC's output current is proportional to a digital input value. By changing the digital input, the compensation current can be adjusted to match the dark current. This allows the system to be calibrated digitally. |
The patent notes that 'the IDAC can be configured to receive a feedback signal from a feedback circuit to determine the amount of compensation to remove from the signal current.' This feedback loop allows the IDAC to track changes in dark current due to temperature or aging. |
Chapter 26: The Dark Current Transient Problem |
When a photodiode is exposed to a bright light source, it can become saturated. When the light is turned off, the photodiode takes time to recover. This transient period can cause errors in dark current cancellation. |
A patent addresses this problem with a bypass switch that shorts the photodiode during the light-on period. This prevents the photodiode from saturating, eliminating the transient period. The patent shows a diagram 'depicting a short settling time of a light sensor implemented in the presence of an 80 lux level of ambient light, where the light sensor output settles substantially before the end of the light source off time period.' |
The patent also describes a technique for reducing the transient period in low-light scenarios. By introducing a reference current and reversing the subtraction operation, the settling time can be reduced. The patent explains that 'the light sensor settling time is proportional to (ref minus photo). Such a signal has a longer settling time in scenarios having high ambient light detection levels and shorter settling times in low light detection scenarios.' |
|
Chapter 27: Current Mirrors for Dark Current Compensation |
Current mirrors are a fundamental building block for dark current compensation circuits. A current mirror copies the dark current from the shielded photodiode and provides it to the signal path for cancellation. |
The patent describes a current mirror circuit 'configured to mirror less than the dark current from the dark photodiode.' This ensures that the mirror does not over-compensate and remove some of the photocurrent. |
The current mirror's transfer ratio is set by the dimensions of the transistors. The patent explains that 'the current mirror circuit can be configured with a transfer ratio that is dependent on the dimensions of the channels of the transistors.' By choosing appropriate channel widths and lengths, the transfer ratio can be set to any desired value. |
The patent provides a specific example: 'For example, the dark current from the dark photodiode is 8 milliamps and the dark current from the signal photodiode is 7 milliamps. The transfer ratio must be less than or equal to 0.875 so that the current mirror only compensates for the dark current from the signal photodiode.' |
Chapter 28: The Use of Multiple Reference Sensors |
Some barcode readers use multiple reference sensors to improve dark current cancellation. By averaging the dark current from multiple sensors, the system can reduce the effect of sensor-to-sensor variations. |
A patent describes a reader with first and second reference sensors: 'The first and second reference sensors are connected to the housing and may be positioned inside or outside of the housing. The first and second reference sensors are positioned with an offset from the light sensor to keep away from the scanning beam reflection.' |
The reference sensors 'collect ambient light reflected from the nearby barcode.' The signals from the reference sensors may be combined to obtain a single ambient light signal. The patent explains that 'the ambient light signals may be averaged or combined to obtain a single signal representing a total ambient light signal.' |
This approach reduces the noise in the dark current estimate. By averaging multiple measurements, the random noise component is reduced, improving the accuracy of the cancellation. |
|
Chapter 29: Phase-Locked Dark Current Cancellation |
A patent describes a sophisticated dark current cancellation technique for laser-based barcode readers. The technique uses phase-locked sampling to separate the laser reflection from the ambient light and dark current. |
The patent explains: 'During the off-time of the laser, a dark field signal is captured and stored in a delay line circuit. This dark field signal represents the ambient light noise. During the on-time of the laser, a combined signal is captured. The dark field signal is then subtracted from the combined signal to extract the laser reflection signal.' |
The delay line circuit matches the timing of the dark field signal to the combined signal. The patent notes that 'the phase shift circuit shifts the phase of the signal to match the delayed signal.' This ensures that the subtraction is performed on signals with the same phase. |
The patent describes an alternative implementation where 'the delay line circuit shown in FIG. 12 may be replaced with an analog-to-digital converter and digital memory device for storing the dark field signal.' This digital implementation eliminates the need for a phase shift circuit. |
Chapter 30: Dark Current Compensation in Ambient Light Sensors |
Ambient light sensors, which are used in barcode readers to adjust the illumination brightness, also require dark current compensation. The ISL29009 from Renesas is an example of an ambient light sensor with integrated dark current compensation. |
The ISL29009 datasheet explains: 'A dark current compensation circuit aids the photodiode to minimize temperature dependent leakage currents in the absence of light, improving the light sensitivity at low lux levels.' This ensures that the ambient light sensor provides accurate readings even in dim conditions. |
The ISL29009's photodiode has a human-eye-like spectral response, with peak sensitivity at 550 nanometers. The integrated non-linear current amplifier extends the dynamic range while maintaining sensitivity at low light levels. The device consumes minimal power because the current consumption ramps in a square root fashion. |
|
Chapter 31: The TXS0108E Bidirectional Level Shifter |
Texas Instruments offers the TXS0108E, a bidirectional level shifter that can be used in barcode reader interfaces. While not directly related to dark current cancellation, this device illustrates the breadth of TI's interface portfolio for barcode readers. |
The TXS0108E is an 8-bit non-inverting level translator that operates from 1.2 to 3.6 volts on the A-side and 1.65 to 5.5 volts on the B-side. It allows communication between devices with different voltage levels, such as a 3.3-volt microcontroller and a 5-volt interface. |
In a barcode reader, the TXS0108E might be used to interface between the imager module and the main processor. The device's bidirectional capability allows data to flow in both directions, simplifying the interface design. |
Chapter 32: The LSF0108 Level Shifter |
Texas Instruments also offers the LSF0108, an 8-bit bidirectional multi-voltage level translator. Like the TXS0108E, this device facilitates communication between circuits with different voltage levels. |
The LSF0108 is designed for open-drain and push-pull applications, making it suitable for a wide range of interfaces. It operates with no direction pin, simplifying the design and reducing pin count. |
In a barcode reader, the LSF0108 might be used to interface between the analog front end and the digital processing circuitry. The level shifter ensures that signals are properly translated between different voltage domains. |
|
Chapter 33: The TIDA-050000 Reference Design |
Texas Instruments provides the TIDA-050000 reference design, which uses the TPS61376 boost converter for barcode scanner applications. This design illustrates the complete system architecture for a barcode reader. |
The TPS61376 is a 3A boost converter for barcode scanners that incorporates the features needed for barcode reader illumination. The reference design includes power management, LED drive, and sensor interface circuitry. |
The TIDA-050000 is an example of how TI provides complete system solutions for barcode reader designers. By integrating the various functional blocks, the reference design simplifies the overall system design. |
Chapter 34: The Problem of Leakage Currents |
Leakage currents can be a significant source of DC offset in barcode reader circuits. Unlike dark current, which comes from the photodiode, leakage currents come from PCB contamination, moisture, and other environmental factors. |
The OPT101 datasheet notes that the integrated design 'eliminates the problems commonly encountered in discrete designs, such as leakage current errors, noise pick-up, and gain peaking as a result of stray capacitance.' By integrating the photodiode and amplifier on a single chip, the OPT101 minimizes the exposed high-impedance nodes that can pick up leakage currents. |
Guard rings are another technique for reducing leakage currents. A guard ring is a trace that surrounds a high-impedance node and is driven to the same voltage as the node. This prevents leakage currents from flowing into the node. |
The Datalogic patent on dark frame subtraction also addresses leakage currents, as they contribute to the fixed pattern noise that the dark image captures and removes. |
|
Chapter 35: Temperature Compensation in the Digital Domain |
Digital signal processing enables sophisticated temperature compensation for dark current. By using temperature information and calibration data stored in memory, the system can adjust the dark current estimate for temperature variations. |
A patent describes a system where 'a first one-time programmable memory is provided which holds a compensatory factor. A second one-time programmable memory provides two compensation coefficients.' These coefficients are used to calibrate the temperature compensation for the specific photodiodes in the system. |
The patent explains that 'different diodes will have different dark currents at the same temperature. Likewise those different diodes will have different behavior from one another as the temperature is increased.' The calibration coefficients account for these differences. |
The compensation circuit 'is further configured to provide dark current compensation for the output of the photosensitive device by processing temperature information and temperature dependent calibration information to compensate for dark current.' This allows the system to provide accurate dark current cancellation across a wide temperature range. |
Chapter 36: Dark Current in Image Sensors |
Image sensors, such as CCDs and CMOS imagers, have dark current characteristics similar to discrete photodiodes. However, the dark current in an imager is pixel-dependent, creating fixed pattern noise. |
The Datalogic patent describes this phenomenon: 'Fixed pattern noise comprises threshold mismatch dark signal non-uniformity and dark current dark signal non-uniformity.' Threshold mismatch is caused by variations in the pixel readout circuitry, while dark current non-uniformity is caused by variations in the photodiodes. |
The patent describes a method for separating these two components. By capturing dark images with different exposure times, the system can estimate both components and compensate for them separately. The threshold mismatch is estimated from the first dark image, and the dark current component is estimated by subtracting the first dark image from the second. |
|
Chapter 37: The IDAC and Digital Calibration |
Digital calibration is a powerful technique for dark current cancellation. By using an IDAC and a digital feedback loop, the system can provide accurate cancellation without manual trimming. |
The patent explains that 'the IDAC can be configured to convert a digital input, e.g., a binary number, into an analog output.' The IDAC is controlled by the feedback circuit, which senses the residual offset at the transimpedance amplifier output. |
The calibration can be performed during manufacturing, with the calibration coefficients stored in memory. The patent notes that 'a first one-time programmable memory is provided which holds a compensatory factor.' This factor is used to calibrate the IDAC for the specific photodiode in the system. |
Digital calibration offers several advantages over manual trimming. It is more precise, more repeatable, and can be performed automatically. It also allows the system to adjust the calibration dynamically in response to changing conditions. |
Chapter 38: The Dark Current Voltage Offset |
In systems that use transimpedance amplifiers, dark current appears as a voltage offset at the amplifier output. The offset voltage is equal to the dark current multiplied by the feedback resistor. |
A patent describes this offset: 'The transimpedance amplifier converts the dark photocurrent from the photodetector into an output voltage that provides the offset voltage.' This offset voltage is then subtracted from the signal voltage. |
The offset voltage is the same for the signal and dark channels, as they use identical transimpedance amplifiers. By subtracting the dark channel output from the signal channel output, the offset voltage is canceled. The patent explains that 'the value of the final output voltage is equal to the difference between the color channel output and the dark channel output.' |
|
Chapter 39: The Need for Matching |
For dark current cancellation to be effective, the signal and dark channels must be closely matched. This includes matching of the photodiodes, the transimpedance amplifiers, and the PCB layout. |
A patent states: 'Photodiodes operating at different biasing points can also have dark currents. Various embodiments of device can be designed such that the dark photodiode and the signal photodiode can operate at approximately the same bias point.' This ensures that the dark currents are matched. |
The dark photodiode 'can be designed to match the signal photodiode, i.e. both photodiodes having similar design parameters so that they will emit the same amount of dark current.' This minimizes the mismatch and improves cancellation accuracy. |
If the photodiodes are not perfectly matched, the current mirror transfer ratio can be adjusted to compensate. The patent notes that 'the current mirror can be configured with a transfer ratio that results in the mirrored current that is less than or equal to the dark current when taking into consideration the manufacturing tolerances.' |
Chapter 40: The Auto-Calibration Technique |
A patent describes an auto-calibration technique for dark current cancellation. The technique uses the signal itself to adjust the compensation, eliminating the need for a separate dark photodiode. |
The technique works by measuring the DC level of the signal during periods when no barcode is present. This DC level includes the dark current and ambient light. The compensation is then adjusted to null the DC level. |
The patent explains that this technique is 'manual-adjustment-free, making the circuit easy to manufacture and calibrate.' The auto-calibration eliminates the need for factory calibration, reducing manufacturing costs. |
The auto-calibration is implemented with a feedback loop that has a slow time constant. The slow time constant ensures that the compensation does not track the barcode signal, only the DC offset. The patent notes that 'the time constant is about one second.' |
|
Chapter 41: The Capacitive Transient Problem |
When a light source is turned off, the photodiode output takes time to settle. This capacitive transient can cause errors in dark current measurements, especially in pulsed illumination systems. |
A patent addresses this problem with a discharge switch that shortens the transient period. The switch is closed during the light-on period, discharging the photodiode capacitance. When the light is turned off, the switch opens, and the photodiode output settles quickly. |
The patent explains that 'the saturation of a light sensor positioned close to a light source when the light source is on can be detrimental to detection of an ambient light level during the short time periods when the light source is off. This is based on a transient period of the light sensor being longer than the off time period of the light source.' |
The discharge switch improves the sensor's performance by ensuring that the output settles before the next measurement. The patent shows a diagram 'depicting a short settling time of a light sensor ... where the light sensor output settles substantially before the end of the light source off time period.' |
Chapter 42: The Dark Frame Storage |
In digital dark frame subtraction, the dark image must be stored for later subtraction. The Datalogic patent describes using an array for this purpose: 'Digitize and store pixels from dark linear image into array D(i) = ADC(i), where i = 0 to number of pixels in linear array.' |
The dark image is stored in the array, and then it is subtracted from the illuminated image pixel-by-pixel. The patent provides pseudocode for this process: 'Capture a dark linear image using exposure time T1. Digitize and store pixels from dark linear image into array. Capture an illuminated linear image using exposure time T2. Digitize and store a fixed pattern noise compensated image.' |
The storage of the dark image is typically done in memory that is accessible to the processor. The patent notes that 'the computer-readable storage medium is to store the dark image.' The memory may be volatile or non-volatile, depending on the system requirements. |
|
Chapter 43: Dark Image Reacquisition |
The dark image must be reacquired periodically to account for changes in dark current. The Datalogic patent describes several conditions that trigger reacquisition: 'An updated dark image may be acquired periodically during use, or may be captured as required by conditions, such as changes in temperature.' |
The patent also describes automatic reacquisition: 'In certain embodiments using a trigger pull scanner, a new dark image may be obtained each time the trigger is pulled.' This ensures that the dark image is always up to date. |
The reacquisition process is transparent to the user. The user pulls the trigger, and the scanner captures a dark image and then a light image. The dark image is subtracted from the light image, and the result is decoded. |
Chapter 44: The System-Level Integration |
Modern barcode readers integrate dark current cancellation at the system level. The imager, the processor, and the memory work together to provide accurate compensation. |
The Datalogic patent describes a system that includes 'an image sensor to capture a dark image and an illuminated image, a computer-readable storage medium to store the dark image, and a fixed pattern noise compensation subsystem to subtract the dark image from the illuminated image.' |
The compensation subsystem may be implemented in hardware, software, or a combination of both. The patent notes that 'the fixed pattern noise compensation subsystem comprises a processor, and the computer-readable storage medium further stores a fixed pattern noise compensation module executable on the processor.' |
The system may also include a temperature sensor. The patent explains that 'based upon the detection of a change of temperature measured by the temperature sensor, the image sensor recaptures the dark image.' This provides temperature-aware dark current cancellation. |
|
Chapter 45: Summary --- Dark Current Cancellation in Perspective |
Dark current cancellation is an essential technique for accurate barcode reading. The dark current from the photodiode adds a DC offset to the signal, shifting the baseline and potentially causing decoding errors. The dark current also varies with temperature, making it a moving target for cancellation circuits. |
We have examined how different companies and technologies have approached the challenges of dark current cancellation: |
Texas Instruments offers the OPT101, a monolithic photodiode with on-chip transimpedance amplifier that minimizes leakage current errors by integrating the photodiode and amplifier on a single chip. The device is suitable for barcode scanners and other optical sensing applications. |
Renesas offers the ISL29009, a light-to-current optical sensor with integrated dark current compensation. The device includes a dark current compensation circuit that minimizes temperature-dependent leakage currents, improving light sensitivity at low lux levels. |
The patent literature reveals a wealth of techniques: shielded reference photodiodes, current mirror-based compensation, IDAC fine-tuning, digital dark frame subtraction, switch-based transient reduction, and temperature-aware compensation. |
|
The key lessons from our exploration are: |
Dark current is a real and significant problem. The photodiode's dark current adds a DC offset to the signal, and this offset varies with temperature. Without cancellation, the dark current can corrupt the barcode signal. |
The shielded reference photodiode is the standard solution. By using a matched photodiode that is shielded from light, the system can measure the dark current and subtract it from the signal. |
Subtraction can be performed in the analog or digital domain. Analog subtraction is typically faster but requires matched components. Digital subtraction is more flexible but requires an ADC and processor. |
Current mirrors provide precision cancellation. A current mirror can copy the dark current from the reference photodiode and inject a compensation current into the signal path. The transfer ratio can be adjusted to account for mismatched photodiodes. |
Temperature compensation is essential. Dark current varies with temperature, and a cancellation circuit that does not account for temperature will be inaccurate. Temperature sensors and calibration data are used to compensate for temperature variations. |
Digital dark frame subtraction is effective for imagers. By capturing a dark image and subtracting it from the illuminated image, the system can remove fixed pattern noise from the image. |
Integration is the trend. Companies are integrating dark current cancellation on a single chip, reducing cost and improving reliability. |
In the end, dark current cancellation is a testament to the importance of analog design in barcode readers. A well-designed cancellation circuit removes the dark current from the signal, providing a clean, stable signal that can be digitized and decoded. A poorly designed cancellation circuit leaves residual dark current, reducing the signal-to-noise ratio and degrading decoding performance. The art of dark current cancellation lies in the careful balance of circuit design, component selection, and system-level integration. |
cs@easiersoft.com