RU2679272C1 - Device for registration of infrared radiation based on matrix of bolometric detectors with nonuniformity compensation scheme - Google Patents

Abstract

FIELD: physics.SUBSTANCE: invention relates to the field of detecting electromagnetic radiation, in particular infrared, based on bolometric detectors. Device contains a bolometric detector array consisting of bolometers sensitive to incident infrared radiation, called “active”, and infrared-insensitive bolometers, called “thermal shorted”, formed on a semiconductor substrate containing a readout circuit consisting of sets of pairs of transistors of different types of conductivity, connected by sources to “active” and “thermal shorted” bolometers, respectively, receiving some bias voltage, with combined drains, which are connected to the integrator inputs, to which with the help of a set of keys controlled by a digital code for compensation of digit capacity n, one of the digits of which determines the sign of the compensation current, sources of positive and negative current compensation are also connected, representing the current mirrors with given current multiplication factors, which are multiplied for each column and multiplied with a given coefficient, the current formed by the means of forming the reference compensation current based on two additional “thermal shorted” bolometers, located outside the matrix field, receiving offsets from transistors in exactly identical to the transistors offset “active” and “thermal shorted” bolometers and offset by exactly the same bias voltage.EFFECT: technical result consists in compensating the technological variation of the resistance values of matrix bolometers in a wide temperature range without using thermal stabilizing elements, in devices for detecting infrared radiation.1 cl, 3 dwg

Description

FIELD OF THE INVENTION

The invention relates to the field of infrared technology and can be used in the manufacture of devices that detect radiation in the infrared range. State of the art

From the existing level of technology according to the patent US 5286976 A, 02.15.1994 "Microstructure design for high IR sensitivity" it is known to use devices called detectors, characterized by a change in the value of a certain physical quantity when the temperature of the detector is absorbed due to absorption of infrared radiation, in particular for a bolometric detector of this the physical quantity is its electrical resistance. Devices for obtaining one- and two-dimensional images use the appropriate matrix of such detectors. Typically, the arrays of bolometric detectors are made on a substrate, in the general case of silicon, containing integral means for measuring the values of the required physical quantity of each detector, followed by the derivation of the obtained values as electrical signals. The measurement of the values of the required physical quantity of the matrix detectors can occur sequentially, simultaneously, and it is also possible to simultaneously measure the values of the detectors included in the group, united by some attribute, for example, by belonging to one row, with a sequential measurement of such groups.

In the case of two-dimensional arrays of bolometric detectors, line-by-line measurement of electrical resistance values has spread, while in a given time interval, the resistance values of all detectors of a given line are simultaneously measured, amplified, and subsequently transmitted to the output with a given gain. In the next similar time interval, a measurement and sequential output of the resistance values of all the detectors included in the next line is performed. This process continues until the values of the resistances of all rows of the bolometric matrix are alternately measured and transmitted to the output. These values form an image frame.

The difficulty associated with recording infrared images is the relatively small change in the temperature of the bolometer due to the absorption of infrared radiation against the background of a change in the temperature of the substrate as a whole, associated with self-heating from flowing currents, fluctuations in ambient temperature and other factors. To solve this problem, in bolometric matrices, compensation structures are used that contain the so-called “thermally shorted” bolometers. “Thermally shorted” bolometers are structurally designed so that their temperature is as close as possible to the temperature of the substrate, and their temperature coefficient of resistance corresponds to the temperature coefficient of resistance of “active” bolometers. When reading the resistance values of “active” bolometers line by line, the required number of compensation structures is equal to the number of matrix columns. A typical reading scheme is described, for example, in the following articles:

VGA IRFPA with 25 μm pixel-pitch for high volume applications», C. Minassian, J.L. Tissot, M. Vilain, O. Legras; Infrared Technology and applications XXXIV, SPIE vol. 6940, 2008;- “Uncooled amorphous silicon TEC-less

VGA IRFPA with 25 μm pixel-pitch for high volume applications ”, C. Minassian, JL Tissot, M. Vilain, O. Legras; Infrared Technology and applications XXXIV, SPIE vol. 6940, 2008;

- “320 × 240 uncooled microbolometer 2D array for radiometric and process control applications”, B. Fieque et al; Optical Systems Design Conference, SPIE 5251, 2003.

In FIG. 1 shows a typical electrical readout circuit for a single matrix column. CMOS transistors 16 and 17, combined by drains in node 23, operate in saturation mode and set the rated currents flowing through the “active” 15 and “thermally shorted” 18 bolometers, respectively. The magnitude of the current equal to the algebraic difference of the currents flowing through the "active" 15 and "thermally shorted" 18 bolometers is fed to the input of the current integrator 5, made on the basis of the operational amplifier 20 and connected to it in the feedback capacitor, also called the integration capacitance. This scheme is present in each column of the matrix.

Since transistors 16 and 17 operate in saturation mode, their currents are given by the expressions:

where V _det1 - voltage on the "active"bolometer; V _cm1 - bias voltage of the transistor 16; V _zi16 - the voltage of the gate-source of the transistor 16; V _det2 - voltage on the " _{thermally shorted} " bolometer 18; V _cm2 is the bias voltage; V _zi17 - the voltage of the gate-source of the transistor 17.

Voltages V _cm1 and V _{cm2 are} selected in such a way that in the absence of external radiation the equality holds:

where I _int is the current at the input of the integrator.

Voltages V _cm1 and V _cm2 require high accuracy (at the current level of technology, these voltages are generated by digital-to-analog converters with a resolution of at least 14) and are set the same for all “active” and “thermally shorted” bolometers, respectively. When the crystal temperature changes, voltage correction V _cm1 and V _{cm2 is} required to maintain the nominal values of currents I ₁₆ , I ₁₇ . Thus, for a known operating temperature range of an infrared photodetector, a table of sets of voltage values V _cm1 and V _{cm2 is} required for each temperature interval into which a given operating temperature range is divided. The smaller the temperature interval in which fixed voltages V _cm1 and V _{cm2 are used} , the smaller the spread of currents I ₁₆ , I ₁₇ when the temperature changes. In modern technology, it is customary to choose similar temperature ranges from 5 ° C to 15 ° C.

An alternative method is to use thermal stabilization based on Peltier elements. In this case, the crystal temperature is maintained at a constant level and does not depend on the ambient temperature. A known disadvantage of this method is the large power consumption of the device as a whole.

A known complication associated with the registration of infrared images is the scatter of the actual values of the resistance of the bolometers along the field of the matrix, caused by fluctuations in the process. This variation in resistance values significantly exceeds the change in resistance caused by a change in temperature due to absorption of infrared radiation. Thus, the measurement error associated with the technological spread of the resistance of the bolometers significantly exceeds the measured value — the change in resistance caused by heating of the bolometer due to absorption of infrared radiation.

In addition, the currents specified by transistors 16 and 17 can be written as functions of the current equation in injection MOS transistors according to the following expressions:

where μ ₁ is the mobility of minority charge carriers; With _oh - the specific capacity of the gate dielectric; W ₁₆ is the channel width of the transistor 16; L _por16 is the channel length of the transistor 16; V _{por 16} - threshold voltage of the transistor 16.

where μ ₂ is the mobility of minority charge carriers; With _oh - the specific capacity of the gate dielectric; W ₁₇ - channel width of the transistor 17; L ₁₇ is the channel length of the transistor 17; V _{pore 17} - threshold voltage of the transistor 17.

The values of many parameters in equations (4) and (5) have a range of values within a single crystal, associated with the technological features of microelectronic production.

The problems described above in practice lead to the fact that when reading bolometers having a different electrical resistance than the nominal value, a certain current I _int is observed in the absence of an external signal, and this current can be either flowing into the integrator or leaking.

This problem can be solved by external processing of the output signal by the well-known one-point and two-point correction methods described in the article "Long Wavelength Infrared 128 × 128 Alx Gal-x As / GaAs Quantum Well Infrared Camera and Imaging System", C.G. Bethea, IEEE Transactions On Electron Devices, Vol. 40, No. 11, November 1993, pp. 1957-1963. The disadvantage of this solution, used in its pure form, is that the electrical range of the output signal is usually limited by the supply voltage of the device, and therefore there is a need to limit the gain of the circuit so that the obtained values of the electrical signal do not go out of the range of possible values of the output signal. In this regard, it is advisable to compensate for the zero level displacement directly in the matrix signal reader, and before integrating the differential current received from the matrix, because the range of possible values at the output of the integrator is also limited by the supply voltage.

The closest technical solution - the prototype, is the device described in patent US 6028309 A, February 22, 2000, "Methods and circuitry for correcting temperature-induced errors in microbolometer focal plane array" (Methods and circuits for correcting errors caused by temperature changes in microbolometric photodetector). The prototype proposed the use of digital-to-analog converters (DACs) in each reading channel. When reading the DAC signal individually for each "active" bolometer, the supply voltage of the "active" bolometer 15 (Fig. 2a), or the bias voltage of the current-sensing transistor 16 (Fig. 2b) is adjusted to compensate for the deviation of the resistance values of both the "active" bolometer and the corresponding him a “thermally shorted” bolometer by correcting the current flowing through the “active” bolometer.

In the methods proposed in the prototype, the correction is carried out by selecting the voltage V _det1 - V _cm1 - V _z1616 individually for each matrix bolometer in such a way that equality (3) is fulfilled for each of them up to the _{least significant} compensation digit. Coefficients are selected once during factory calibration of the matrix. The selection of voltage is carried out by corrective DACs. From equations (1) and (4) we express V _zi16 :

Where

When compensating with the proposed method, the following equality must be fulfilled:

According to the expression (1) we obtain:

where R ₁₅ is the nominal resistance of the bolometer, ΔR ₁₅ is the deviation of the resistance from the nominal.

To fulfill condition (3) at a certain temperature T ₀ , called the calibration temperature, a voltage is required, defined as follows:

Or, taking into account (1), we obtain:

The dependence of the electrical resistance of a bolometric material on temperature is determined by the expression:

- сопротивление болометра при температуре (Т₀+ΔТ), Т₀ - температура калибровки, ΔТ - отклонение температуры от температуры калибровки,

- сопротивление при температуре калибровки, α - температурный коэффициент сопротивления.Where

- the resistance of the bolometer at a temperature (T ₀ + ΔТ), T ₀ - calibration temperature, ΔТ - temperature deviation from the calibration temperature,

- resistance at calibration temperature, α - temperature coefficient of resistance.

When the temperature changes, according to (11) and (8), the current through the active bolometer will be determined by the expression:

- напряжение затвор-исток транзистора 16 при температуре (Т₀+ΔТ), которое может быть выражено с учетом (6) следующим образом:Where

- the voltage of the gate-source of the transistor 16 at a temperature (T ₀ + ΔT), which can be expressed taking into account (6) as follows:

The integrator current can be written as:

In view of (3), (10), (11), and (14) we obtain:

It is easy to see from expressions (13) and (15) that at a temperature different from the calibration temperature (ΔТ ≠ 0) and a bolometer resistance other than the nominal value (ΔR ≠ 0) there is a non-zero current I _int , which violates the equality ( 3). This causes the appearance of a non-uniform background image in the output signal, the non-uniformity of which increases with the difference between the actual temperature of the bolometer and the temperature of the bolometer during calibration, which significantly complicates the current task of refusing to use thermal stabilization of the matrix.

Disclosure of the invention

The objective of the invention is the construction of a device for recording infrared radiation with a readout circuit that allows you to compensate for the technological spread of the resistance values of the matrix bolometers in a wide temperature range without the use of thermally stabilizing elements.

The technical result is achieved by introducing a reference current source characterized by the exact same dependence of the output current on the temperature and bias voltage of the bolometers, as well as on the current flowing through the "active" bolometer. The proposed reference current source contains two additional “thermally shorted” bolometers located outside the matrix field with a given ratio of the nominal resistance values, the current through which is set by transistors that exactly repeat the transistors that specify the current through the “active” and “thermally shorted” matrix bolometers. Moreover, transistors that specify currents through bolometers in the reference current source are biased by exactly the same voltages as transistors that specify currents through “active” and “thermally shorted” matrix bolometers.

The implementation of the invention

To solve this problem, a device is proposed for recording infrared radiation based on a matrix of bolometric detectors (Fig. 3), consisting of bolometers 15 sensitive to incident infrared radiation, called "active", and insensitive to infrared radiation, bolometers 18, called "thermally shorted", formed on a semiconductor substrate containing a readout circuit consisting of a plurality of transistors 16 and 17 connected by sources to “active” and “thermally shorted” bolometers, respectively of having different conductivity type and receiving bias voltages V and V _{cm-1 cm 2,} respectively, with the combined effluents connected to the input of the integrator 5. In this aspect of the invention is a means for generating a reference current compensating 4 on the basis of two additional "termozakorochennyh" bolometers 6 and 9, located outside the field of the matrix, receiving bias from transistors 7 and 8, exactly identical to transistors 16 and 17, respectively, receiving bias with exactly the same voltages V _cm1 and V _cm2, respectively about with combined drains. The magnitude of the current equal to the difference between the currents flowing through the bolometers 6 and 9 is supplied to the means of copying the compensation current for each column of the matrix 3, which is a current mirror, the outputs of which are connected to the inputs of the respective sources of positive 12 and negative 14 compensation current, representing current mirrors with the specified multiplication factors of the current, which are connected to the inputs of the integrators 5 using the keys 21 and 22, controlled by a digital code of the compensation coefficient k of the capacity n, one of the times the rows of which selects the sign of the compensation current. The total number of compensation sources n is selected depending on the required accuracy of the installation of the video output signal and the maximum possible dispersion of the resistance values of the bolometers for this technology. The area of the transistors in the compensation current sources is selected so that each subsequent transistor has a W / L ratio 2 times larger than the previous one. In this case, n compensation sources will provide 2 ⁿ possible values of compensation currents.

In the factory calibration, for each bolometer individually, by the selection method, the value of the compensation current coefficient is found at which the equality is fulfilled:

- ток компенсации младшего разряда.where k is the compensation factor,

- current compensation low order.

Expression (16) should be valid when the temperature of the substrate changes. From (1), (2), (11) it follows that the currents flowing through the “active” and “thermally shorted” bolometers, when the temperature changes, are determined by the expressions:

The temperature of the “active” bolometer depends not only on the substrate temperature, but also on the infrared radiation energy absorbed by the bolometer. The current active bolometer 15 when exposed to infrared radiation can be written in the form:

where ΔТ _ik - temperature change caused by the absorption of infrared radiation.

Consider a reference current source 4 (Fig. 3), consisting of two additional "thermally shorted" bolometers 6 and 9, current-setting transistors 7 and 8, and transistor 10, which is a copy of the current. Bolometers 6 and 9 have exactly the same temperature coefficient of resistance as the “active” and “thermally shorted” bolometers, and the inequality holds for the values of their resistance:

Transistors 7 and 8 are similar in design and size to transistors 16 and 17, respectively. Therefore, transistors 7 and 8 also work in saturation mode, and their currents are given by the expressions:

where V _zi7 is the gate-source voltage of the transistor 7, V _zi8 is the gate-source voltage of the transistor 8.

When the temperature changes, expressions (21) and (22), taking into account (11), can be rewritten in the form:

The reference compensation current when the temperature changes will be determined by the expression:

The integrator current, taking into account (17), (18), (25), will be determined by the following expression:

Obviously, equality (26) will hold if equality (16) holds for any ΔT.

In the presence of infrared radiation, the read current, taking into account equalities (19) and (26), will be determined by the equality:

When reading the signal using the key sets 21 and 22, compensation sources are connected that form the required value of the compensation current and its sign. Then, the "Reset" key of the integrator 5 is turned on. At the same time, a voltage is formed at the output, determined by the expression:

Where V _op - the reference voltage of the integrator.

Напряжение, сформированное на выходе интегратора, будет определяться следующим образом:After a time sufficient to generate a voltage defined by expression (28) at the output, the Reset switch opens and switches 1 and 2 close. From this moment, the capacitor 19 starts charging with current I _int until the switch 2 is turned off. The time for which the switch is turned on 2, denote by

The voltage generated at the output of the integrator will be determined as follows:

where T _int is the integration time, C _int is the capacitance of the capacitor 19.

In the general case, taking into account expression (27), the output voltage of the integrator will be determined by the expression:

Brief Description of the Drawings

In FIG. Figure 1 shows a typical scheme for reading the resistance of a bolometer without compensating for matrix heterogeneity.

In FIG. 2 shows reading schemes with compensation for matrix heterogeneity proposed in the prototype:

a) compensation due to voltage correction V _det1 ;

b) compensation due to voltage correction V _cm1 .

In FIG. 3 shows the proposed readout scheme with compensation for matrix inhomogeneity.

Claims (1)

A device for detecting infrared radiation based on bolometric detectors, comprising a matrix of bolometric detectors, consisting of bolometers sensitive to incident infrared radiation, called “active”, and infrared insensitive bolometers, called “thermo-shortened”, formed on a semiconductor substrate containing a readout circuit consisting of sets of pairs of transistors of various types of conductivity connected by sources to “active” and “thermally shorted” bol to the meters, respectively, receiving some bias voltages, with combined drains that are connected to the inputs of the integrators, characterized in that the readout circuit contains: means for generating a reference compensation current based on two additional “thermally shorted” bolometers located outside the matrix field, receiving bias from the transistors, exactly identical to the bias transistors of the "active" and "thermally shorted" bolometers and biased by exactly the same bias voltages, the combined drains of which are connected to the means of copying the compensation current for each column of the matrix, which are current mirrors, the outputs of which are connected to the inputs of the corresponding sources of positive and negative compensation current, which are current mirrors with topologically specified current multiplication coefficients, which are connected to the integrator inputs using a set of keys, controlled by a digital code for compensation of bit depth n, one of the bits of which determines the sign of the compensation current.

Images