![]() DIFFERENTIAL TEMPERATURE SENSOR CIRCUIT WITH GREAT DYNAMIC MARGIN SELECTION OR HIGH SENSITIVITY (Mac
专利摘要:
The present invention describes a sensor circuit to be integrated into a microelectronic circuit that allows measuring the temperature difference between two points on the surface of the semiconductor crystal. The sensor circuit has two modes of operation: mode 1, where it works with a high sensitivity and a small linear margin and mode 2, where it has a lower sensitivity but a large linear dynamic range. Figure 1 shows an example of transistor-level schematic of the differential sensor with high dynamic range selection or high sensitivity. (Machine-translation by Google Translate, not legally binding) 公开号:ES2694036A1 申请号:ES201730806 申请日:2017-06-16 公开日:2018-12-17 发明作者:Josep Altet Sanahujes;Xavier ARAGONÉS CERVERA;Diego Mateo Peña;Eva María VIDAL LÓPEZ;Sergio RUÍZ GONZÁLEZ 申请人:Universitat Politecnica de Catalunya UPC; IPC主号:
专利说明:
DIFFERENTIAL TEMPERATURE SENSOR CIRCUIT WITH GREAT SELECTIONDYNAMIC MARGIN OR HIGH SENSITIVITY SECTOR OF THE TECHNIQUE The present invention describes a sensor circuit to be integrated in a microelectronic circuit that allows to measure the temperature difference between two points of the semiconductor crystal surface. The technical sector is microelectronic design, specifically microelectronic design of differential temperature sensors. BACKGROUND OF THE INVENTION The temperature sensor circuits are integrated into microelectronic circuits together with other digital and / or analog circuits. The knowledge of the temperature in certain points of the circuit allows to manage which blocks of this one are working and to verify if they do it within the correct parameters of reliability. This information also allows choosing if it is necessary to activate cooling strategies and / or low consumption operation. Another application is the characterization and testing of subcircuits (also called blocks) integrated into the same semiconductor die (for example, amplifiers) [1,2]. For example, the patent [4] shows a method for measuring the efficiency of Class A linear amplifiers from temperature measurements. For this last application, article [3] proposes to use differential temperature sensors to perform the characterization. [1] P 200002735 Structural verification procedure for analog integrated circuits based on internal and concurrent temperature observation. [2] P200501512 Procedure to determine the electrical characteristics of integrated analog circuits. [3] J. Altet, A. Rubio, E. Shcaub, S. Dilhaire, W. Claeys, "Thermal coupling in integrated circuits: application to thermal testing". IEEE Journal of Solid State Circuits, 2001, Vol. 36, no. 1, pp. 81-91. [4] P201230234 Procedure for measuring the efficiency of class A linear integrated power amplifiers by means of continuous temperature measurements. EXPLANATION OF THE INVENTION Differential temperature sensors measure the temperature difference between two points on the surface of the integrated circuit. The sensor described in article [3] is composed of two temperature transducers, both based on a bipolar transistor. If the sensor is applied to perform the characterization of a circuit that is integrated in the same semiconductor crystal as the temperature sensor, one of the two transistors is located near the circuit to be characterized and detects the variations in ambient temperature and those caused by the power dissipation of the circuit to be characterized. The other transistor, located far from the circuit to be characterized, only detects the variations of the ambient temperature (if the circuit to be characterized is not active, the two transducers will be at the same temperature). These two transistors are connected in common emitter and constitute the differential pair of the input stage of a transconductance amplifier operating in open loop, in such a way that the output voltage is immune to the variations of the ambient temperature. This sensor circuit offers a high sensitivity to temperature variations caused by the circuit under measurement. This high sensitivity causes a saturation of the output voltage when there is an undesired unbalance between the two differential branches that make up the sensor. Undesired unbalance is understood to be that which is not caused by the power dissipation of the circuit to be characterized. They can be classified into: systematic by design (for example, in [3] it is shown how a structural imbalance is introduced in the differential pair when introducing the polarization circuitry of the base of the transistors that act as transducers), or random, by example, variations of the manufacturing process (known as "mismatching") or by the existence of a temperature gradient on the surface of the semiconductor crystal, caused by other subcircuits than the one to be characterized. Additionally, designing a differential sensor with high sensitivity has the detriment of a linear dynamic range of small input. This feature restricts the use of differential sensors, since in certain applications (for example the procedure described in the patent [4]) a sensor is required that has both characteristics: high sensitivity to perform some measurements and a high linear dynamic range to perform others Figure 1 shows an example of implementation of the sensor circuit proposed in the present patent. The circuit consists of three parts: Part (1) is a differential sensor with a topology similar to that reported in the article [3]. In the circuit of figure 1 the two devices that act as temperature transducers are two NMOS transistors (4), which are coupled by spout. These transistors have the same dimensions and the same voltage between door and spout (in this circuit, externally fixed through the terminal (9)). Consequently, at equal temperature and without considering variations due to the manufacturing process, the drain current of both transistors is the same. This current is disconnected when the two transistors work at different temperatures (what happens when the circuit to be characterized is activated). When this occurs, the difference in drain current between both transistors can be described by the equation: డ ூ ವ ሺ ర ሻ Equation 1 ቃ ሺ ସ ሻ ܶ ൌΔ ሺ ସ ሻ_ ் ܫ Δ డ ் ூ ವ ሺ ర ሻ ୀ ூ ವೂ ሺ ర ሻ where: ID (4) _T is the difference between the drain currents of the transistors (4) caused by the difference in their working temperature, T (4) is the difference in working temperature between the transistors (4) , ID (4) is the drain current of the transistors (4) and IDQ (4) is the polarization current that the transistors (4) have when they are working at the same temperature, that is, before the circuit that you want to characterize by means of temperature measurements, dissipate power and modify the thermal equilibrium between the transistors. In bipolar technologies, BiCMOS or CMOS with triple well, the devices (4) can be replaced by a differential pair of bipolar transistors coupled by common emitter. The current mirrors (5), (6) and (7), together with the stage of high output impedance formed by the transistors (6) converts the drain current difference between the transistors (4) into voltage variations of the terminal (8) (as long as the sensor remains within the linear range of operation). The objective of part (2) is to generate a reference current, which may be insensitive to temperature variations. If the temperature variations of this generated reference current are not a limitation for the final application of the temperature sensor, this part (2) can be ignored and one of the two branches of the differential circuit of part (1) can be used. (for example the transistors (4), (7) and (16) of the right branch of the differential structure) to generate the gate voltage that polarizes the transistors (13). Each of the parts (3) is a source of current controlled by digital code. The transistors (12), (13), (14) and (15) form current mirrors. The gate length of the transistors (12), (13), (14) and (15) is the same. However, the widths differ. For example, widths can verify (although other relationships are possible): W (15) = 2 · W (14) = 4 · W (13), W (12) = W (13) where W (i) represents the door width of the transistor (i). The transistors (17) and (18) function as switches: the switch / transistor will be closed and will allow current circulation when the voltage applied to its door is 0 V. Alternatively, it will be open (ie, the transistor will not conduct current) when the voltage applied to your door is equal to the supply voltage of the circuit. The transistors (16) are always driving: they have been added to the circuit to minimize the mismatch between the current mirrors that exist throughout the circuit. The objective of part 3 of the design is to introduce current to the differential sensor through branches (19) and (20). Assuming that the voltages applied to (10) and (9) are the same, the value of the current that is added to the differential sensor by branches (19) and (20) can be described as: IN N · K (4) _ (11) _ (13) · ID (4) Q Equation 2 where IN is the current flowing from part (3) to (1) through branches (19) or (20), N is the decimal number coded in natural binary -base 2- by vectors A = a2a1a0 (for the branch (19)) or B = b2b1b0 (for the branch (20)), K (4) _ (11) _ (13) is a constant that indicates the relationship between the dimensions of the doors of the transistors (4), (11) and (13), and IDQ (4) is the polarization current that the transistors (4) have when they are working at the same temperature, that is, before the desired circuit characterize by means of temperature measurements, dissipate power and modify the thermal equilibrium between the transistors (4) (already described in Equation 1). Other relations between the widths of the gates of the transistors (13), (14) and (15) different than powers of 2, would imply another different coding of natural binary for the number N present in Equation 2. In this example we have limited to 3 the number of components of each of the digital vectors, but may be increased depending on the resolution and dynamic range desired for each of the current sources (3). If we consider the two previous equations, the current mirrors formed by the NMOS transistors (5) and (6) can have a variation of drain current caused either by a variation of the temperature of the transistors (4) and / or by the binary code present in the door of the transistors (17) and (18). Assuming that the aspect ratio between transistors (5) and (6) is unitary - NMOS of (5) equal to NMOS of (6), PMOS of (5) equal to PMOS of (6) - the current variation of drain that have the transistors (5) (or (6)) is: ID (5) IN K (7) _ (5) · ID (4) _T Equation 3 with K (7) _ (5) being the aspect ratio between the gate sizes of the transistors (7) and (5). This variation of current is converted into variation of the terminal voltage (8) thanks to the high output resistance of the transistors (6). BRIEF DESCRIPTION OF THE DRAWINGS To complement the description that is being made and in order to help abetter understanding of the characteristics of the invention, it is accompanied as partmember of said description, a set of drawings where illustrative and notlimiting, the following has been represented: Figure 1.- Shows an example of a schematic at the transistor level of the differential sensorwith high dynamic range or high sensitivity selection.Figure 2.- Shows the typical output function of the temperature sensor when workingin high sensitivity mode.Figure 3.- Shows the typical output function of the temperature sensor when workingin high dynamic range mode. PREFERRED EMBODIMENT OF THE INVENTION Figure 1 shows an example of implementation of the sensor circuit proposed in the present patent. The circuit has two modes of operation: Mode 1. High sensitivity between voltage variation at the output and differential temperature of the differential input pair. During the polarization phase of the temperature sensor circuit, the value of the vectors a2a1a0 or b2b1b0 is set to ensure that the sensor is working in its linear zone. Once the digital vectors are set, the circuit to be characterized is activated and the temperature measurement is carried out. Figure 2 shows an example of sensor input-output characteristic. In this example, the sensor has been powered at a voltage of 3.3 V. The horizontal axis (21) represents the temperature difference (T, in degrees Celsius ° C) that exists between the transistors (4), the axis vertical (22) shows the output voltage (in volts V) in the terminal (8) of the sensor (V (8)). A total of 15 different input-output characteristic curves are shown, based on the value of the vectors a2a1a0 and b2b1b0. For example, the function (33) corresponds to the digital codes a2a1a0 = 000 and b2b1b0 = 000, the function (29) corresponds to the digital codes a2a1a0 = 100 and b2b1b0 = 000 and the function (30) corresponds to the digital codes a2a1a0 = 101 and b2b1b0 = 000. The linear input margin is defined as the set of values in which the voltage of the output terminal of the sensor (8) has a behavior that can be defined by the equation: V (8) STD · T Equation 4 Figure 2 shows the linear input margin of the sensor (23) when the digital codes A = a2a1a0 and B = b2b1b0 applied to the gates of the transistors (17) and (18) respectively are 0 (33). This linear input margin is delimited by its minimum value (24) and maximum (25). Depending on the applied digital code, the minimum (24) and maximum (25) values that define the linear input margin can be modified. This feature makes it possible to compensate any unwanted imbalance between the two differential branches that make up the sensor and that can saturate the output voltage thereof, allowing high-sensitivity temperature measurements even in the presence of pre-existing thermal gradients in the integrated circuit, such as the temperature measurements that are required when applying the procedure described in the patent [4]. Mode 2: Large dynamic range. We define C as: C = Value represented by A, when B is 0 C = - Value represented by B, when A is 0. The condition is imposed that only one of the two, A or B, can have a different value than 0. We define the reference voltage VREF (26), this voltage being one of the possible values of sensor output voltage. We define C�T as the digital code that must be applied to the part (3) of the circuit so that the output voltage of the sensor is the reference when the temperature difference between the transistors (4) is T. For example, observing Figure 2, (26), (27) and (29) indicate that if we consider VREF = 1.6 V, then C6 ° C = 4 (a2a1a0 = 100, b2b1b0 = 000). Observing (26), (28) and (30), C8 ° C = 5 (a2a1a0 = 101, b2b1b0 = 000), considering the same reference voltage. Figure 3 shows C�T (vertical axis (32), without units) as a function of T (horizontal axis (31), expressed in ° C) for VREF = 1.6 V. We define T0 as the difference in temperature existing between the transistors (4) before activating the circuit to be characterized. We define T1 as the temperature difference between the transistors (4) after activating the circuit to be characterized. We associate the codes C�T0 to T0 and C�T1 to T1. The temperature variation (T1-T0) can be obtained by the equation: (T1-T0) = SCD · (C�T1 - C�T0) Equation 5 Where SCD is the inverse of the slope of the function represented in figure 3. If you compare figures 2 and 3, while in the figure 2 the linear margin of input to the sensor is approximately about 5 ° C, for figure 3 the linear margin of input is about 30 ° C. In this operating mode, the sensor output is directly digital. The quantization and linearity errors are related to the resolution and linearity of the current sources implemented in part (3) of the circuit of Figure 1.
权利要求:
Claims (1) [1] 1. Differential temperature sensor circuit with high dynamic range or high sensitivity selection, characterized by: 5 I) It consists of three parts: Part 1 is a differential transconductance amplifier that contains a differential pair, called input torque, and current mirrors. The currents that circulate through each of the devices of the differential pair are proportional to temperature variations in said 10 devices. Part 2 is a block that generates a reference current, which can be independent of the temperature. Part 3 is two blocks that, from the reference current generated by part 2, generate a current that can be controlled by a digital code. II) The currents generated in part 3 of the circuit are added to two branches of the 15 part 1 of the circuit, so that in each of these branches the current can be expressed as the weighted sum of: current generated by one of the sources of part 3 plus the current flowing through one of the devices that make up the differential input pair and that is sensitive to the temperature of the device. Figure 1: Figure 2: Figure 3:
类似技术:
公开号 | 公开日 | 专利标题 US9170308B2|2015-10-27|Circuit and method for biasing a plate-shaped sensor element of semiconductor material Oshita et al.2015|Compact BJT-based thermal sensor for processor applications in a 14 nm tri-gate CMOS process RU2624565C1|2017-07-04|Instrument amplifier for work at low temperatures KR100674974B1|2007-01-29|Semiconductor temperature sensor capable of adjusting sensing temperature CN106249804B|2018-03-09|Reference voltage circuit ES2694036B2|2019-05-16|DIFFERENTIAL TEMPERATURE SENSOR CIRCUIT WITH GREAT DYNAMIC MARGIN SELECTION OR HIGH SENSITIVITY US20160252923A1|2016-09-01|Bandgap reference circuit US9213351B2|2015-12-15|Bi-directional current sensor Chan et al.2007|A CMOS ISFET interface circuit with dynamic current temperature compensation technique Wang et al.2012|Temperature sensor using BJT-MOSFET pair Malits et al.2017|Temperature sensing circuits in CMOS-SOI technology TWI425236B|2014-02-01|Threshold voltage measurement device US20160041571A1|2016-02-11|Current generator circuit and method of calibration thereof Liu et al.2008|Experimental validation of PTAT for in situ temperature sensor and voltage reference Bratt et al.1993|Aspects of current reference generation and distribution for IDDx pass/fail current threshold determination WO2013093159A1|2013-06-27|Method for estimating electrical characteristics of an analog circuit by means of continuous temperature measurement ES2294888A1|2008-04-01|Analog integrated power circuits for semiconductor crystal, has semiconductor crystal that contain different analog circuits like amplifier, and operation of analog circuit causes power dissipation and increase in temperature Ocaya2006|An experiment to profile the voltage, current and temperature behaviour of a P–N diode US9074943B2|2015-07-07|Production-test die temperature measurement ES2425007B1|2014-07-30|PROCEDURE FOR MEASURING THE EFFICIENCY OF LINEAR INTEGRATED POWER AMPLIFIERS CLASS A USING CONTINUOUS TEMPERATURE MEASUREMENTS Altet et al.2014|Electro-thermal characterization of a differential temperature sensor in a 65 nm CMOS IC: Applications to gain monitoring in RF amplifiers US9470582B2|2016-10-18|Temperature sensing circuit and method for sensing temperature SU697937A1|1979-11-15|Device for measuring threshold voltages of mos-transistors Nateghi et al.2015|A self-healing technique using ZTC biasing for PVT variations compensation in 65nm CMOS technology JP2004502133A|2004-01-22|Amplifying device for sensor and physical quantity measuring system using this device
同族专利:
公开号 | 公开日 ES2694036B2|2019-05-16|
引用文献:
公开号 | 申请日 | 公开日 | 申请人 | 专利标题
法律状态:
2018-12-17| BA2A| Patent application published|Ref document number: 2694036 Country of ref document: ES Kind code of ref document: A1 Effective date: 20181217 | 2019-05-16| FG2A| Definitive protection|Ref document number: 2694036 Country of ref document: ES Kind code of ref document: B2 Effective date: 20190516 |
优先权:
[返回顶部]
申请号 | 申请日 | 专利标题 ES201730806A|ES2694036B2|2017-06-16|2017-06-16|DIFFERENTIAL TEMPERATURE SENSOR CIRCUIT WITH GREAT DYNAMIC MARGIN SELECTION OR HIGH SENSITIVITY|ES201730806A| ES2694036B2|2017-06-16|2017-06-16|DIFFERENTIAL TEMPERATURE SENSOR CIRCUIT WITH GREAT DYNAMIC MARGIN SELECTION OR HIGH SENSITIVITY| 相关专利
Sulfonates, polymers, resist compositions and patterning process
Washing machine
Washing machine
Device for fixture finishing and tension adjusting of membrane
Structure for Equipping Band in a Plane Cathode Ray Tube
Process for preparation of 7 alpha-carboxyl 9, 11-epoxy steroids and intermediates useful therein an
国家/地区
|