ON-CHIP CALIBRATION CIRCUIT AND HALF-STEP RESOLUTION METHOD
The on-chip calibration circuit with a binary search process and post-calibration adjustments addresses the inefficiencies of current methods, providing enhanced resolution and reduced time and cost for chip-to-chip parameter variations.
Patent Information
- Application Number
- DE102019213253
- Authority / Receiving Office
- DE · DE
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2018-09-06
- Filing Date
- 2019-09-03
- Publication Date
- 2024-11-14
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Current on-chip calibration techniques for reference parameters, such as reference voltage, are time-consuming and costly due to chip-to-chip variations, and existing self-calibration circuits often compromise between speed and resolution.
An on-chip calibration circuit employing a binary search process with a digital-to-analog converter (DAC) and a comparator in a feedback loop, allowing for half-step adjustments and post-calibration processing to enhance resolution without increasing calibration time.
The proposed method achieves improved calibration resolution with reduced time and cost by statistically optimizing the DAC step selection, ensuring the calibrated parameter is closer to the target value.
Smart Images

Figure 00000000_0000_ABST
Abstract
Description
BACKGROUNDField of the invention
[0001] The present invention relates to on-chip parameter calibration and, more particularly, to a circuit and method for on-chip calibration with improved resolution. Description of the state of the art
[0002] Integrated circuits (ICs) often require the use of a reference parameter generated internally by an on-chip reference circuit (e.g., a reference voltage generated internally by a bandgap voltage reference circuit, a reference current generated by a current reference circuit, etc.). By design, this reference parameter can be set to a specific value for optimal IC performance. However, due to process variations, temperature-related variations, aging-related variations, etc., chip-to-chip variations in this reference parameter can occur. These chip-to-chip variations in the reference parameter can be relatively large as devices scale (e.g., to the 14nm technology node, 7nm technology node, and beyond). For example, chip-to-chip variations in a reference voltage can be as high as 100 millivolts (mV) or even greater.Therefore, calibration of the reference parameter is often required. Calibration refers to the process of adjusting a reference parameter so that the resulting calibrated parameter falls within an acceptable range. However, currently used calibration techniques can be time-consuming and therefore costly. Designers often use a programmable resistor array for at least one of the resistors in the bandgap voltage reference circuit, for example, to compensate for chip-to-chip variations in a reference voltage. In this case, a reference voltage is measured at the output of the bandgap voltage reference circuit, and depending on the value of this reference voltage, one or more of the programmable resistors can be programmed to adjust the value of the reference voltage at the output so that the resulting calibrated voltage falls within an acceptable voltage range.Measuring the reference voltage and programming resistors to generate a calibrated voltage can be time-consuming and thus costly as the resolution increases (more bits). Recently, self-calibration circuits have been developed to reduce calibration time. However, such self-calibration circuits generally require a significant trade-off between circuit features that enable fast calibration and those that enable high-resolution calibration.
[0003] The document US 2010 / 0 079 327 A1 shows a method comprising performing a partial asymmetric approximation of a negative input of an analog-to-digital converter using a first DAC and a first comparator to generate a negative result, performing a full single-ended approximation of an analog-to-digital converter plus input using a second DAC and a second comparator to generate a positive result, and generating a conversion result by combining the negative result and the positive result. SUMMARY
[0004] In view of the above, embodiments of an on-chip calibration circuit configured to perform a binary search calibration process with respect to a reference parameter and also provide additional calibration resolution (e.g., an additional half DAC step in resolution) using post-calibration processing are described herein. By using post-calibration processing, improved calibration resolution can be achieved without correspondingly increasing calibration time. Embodiments of a corresponding calibration method are also described herein.
[0005] More specifically, embodiments of an on-chip calibration circuit for calibrating a reference parameter are disclosed herein. This calibration circuit may include: an N-bit digital-to-analog converter (DAC) that outputs an analog parameter and includes an output parameter adjustment circuit; a comparator that receives a reference parameter and the analog parameter; and a control circuit operatively connected to the comparator and the DAC in a feedback loop. The calibration circuit also includes DAC step selection logic that is either incorporated into or in communication with the control circuit.
[0006] The calibration circuit can be selectively operated in either a calibration mode or an operating mode. In calibration mode, the output parameter adjustment circuit automatically adjusts the magnitude of the analog parameter by approximately half a DAC step (i.e., half the LSB (least significant bit) DAC step) in a first direction (e.g., automatically increases the magnitude of the analog parameter by half a DAC step). Additionally, the control circuit uses the feedback loop to perform a calibration process to identify a preliminary DAC step. During operating mode, the output parameter adjustment circuit automatically adjusts the magnitude of the analog parameter by half a DAC step in a second direction opposite to the first direction (e.g., automatically decreases the magnitude of the analog parameter by half a DAC step).Additionally, the DAC step selection logic selects either the preliminary DAC step previously identified during the calibration process or a neighboring DAC step (e.g., a next higher DAC step) as the final DAC step for operational use after calibration. The control circuit then outputs a final DAC code corresponding to the final DAC step, and the DAC generates a calibrated parameter in response to the final DAC code. The final DAC step selected by the DAC step selection logic is the DAC step statistically most likely to result in a calibrated parameter whose value is closest to a target voltage, given the half-DAC step decrease in the magnitude of the analog voltage in the operating mode.
[0007] A particular embodiment of the calibration circuit can be used for voltage calibration. In particular, this calibration circuit can include an N-bit digital-to-analog converter (DAC) that outputs an analog voltage and has an output voltage adjustment circuit; a voltage comparator that receives a reference voltage and the analog voltage; and a control circuit operatively connected to the voltage comparator and the DAC in a feedback loop. The calibration circuit can also include DAC step selection logic that is either integrated into the control circuit or communicates with it.
[0008] This calibration circuit can be operated either in a calibration mode or an operating mode. In calibration mode, the output voltage adjustment circuit automatically increases the magnitude of the analog voltage by approximately half a DAC step (i.e., half the DAC step of the least significant bit (LSB)). The control circuit then uses the feedback loop to perform a binary search calibration process to identify a preliminary DAC step. In operating mode, the output voltage adjustment circuit automatically decreases the magnitude of the analog voltage by approximately half a DAC step. Additionally, the DAC step selection logic selects either the preliminary DAC step previously identified during the calibration process or a next higher DAC step as the final DAC step for use in operation after calibration.The control circuit then outputs a final DAC code corresponding to the final DAC step, and the DAC generates a calibrated voltage in response to the final DAC code. The final DAC step selected by the DAC step selection logic is the DAC step statistically most likely to result in a calibrated voltage whose value is closest to a target voltage, given the half-DAC step decrease in the magnitude of the analog voltage in the operating mode.
[0009] Embodiments of a corresponding calibration method are also described herein. The method may include providing a calibration circuit as described above. That is, the provided calibration circuit may include: an N-bit digital-to-analog converter (DAC) that outputs an analog parameter and has an output parameter adjustment circuit; a comparator that receives a reference parameter and the analog parameter; and a control circuit operatively connected to the comparator and the DAC in a feedback loop. The calibration circuit also includes DAC step selection logic that is either incorporated into or in communication with the control circuit.
[0010] The method may further comprise selectively operating the calibration circuit in either a calibration mode or an operation mode. Operating the calibration circuit in the calibration mode comprises automatically adjusting a magnitude of the analog parameter by the output parameter adjustment circuit by approximately half a DAC step (i.e., half of the LSB (least significant bit) DAC step) in a first direction (e.g., automatically incrementing the analog parameter by half a DAC step). Operating the calibration circuit in the calibration mode further comprises using the feedback loop, by the control circuit, to perform a binary search calibration process to identify a preliminary DAC step.Operating the calibration circuit in the operating mode includes automatically adjusting the magnitude of the analog parameter by approximately one-half of a DAC step in a second direction opposite the first direction (e.g., automatically decreasing the magnitude of the analog parameter by one-half of a DAC step). Operating the calibration circuit in the operating mode also includes selecting, by the DAC step selection logic, either the preliminary DAC step or an adjacent DAC step (e.g., the next higher DAC step) as the final DAC step for use in post-calibration operation; outputting, by the control circuit, a final DAC code corresponding to the final DAC step; and generating, by the DAC, a calibrated parameter in response to the final DAC code.The final DAC step selected by the DAC step selection logic is the DAC step that is statistically most likely to result in a calibrated parameter whose value is closest to a target voltage given the decrease in the magnitude of the analog voltage by half a DAC step in the operating mode.
[0011] It should be noted that in the above-described embodiments of the calibration circuit and method, when the calibration circuit is operated in calibration mode and the magnitude of the analog parameter output by the DAC (e.g., the analog voltage) is adjusted by approximately half a DAC step, the DAC would ideally be centered around the target, so that the binary search calibration process would ideally calibrate to the midpoint. That is, since calibration mode moves the voltage to be calibrated by half a DAC step, the middle DAC step (i.e., the midpoint voltage) should be centered on the reference voltage to calibrate with the highest degree of accuracy. BRIEF DESCRIPTION OF THE DIFFERENT VIEWS OF THE DRAWINGS
[0012] The present invention will become more apparent from the following detailed description with reference to the drawings, which are not necessarily drawn to scale and in which: Fig. 1 is a schematic diagram illustrating an on-chip calibration circuit; Fig. 2 is a schematic diagram illustrating one embodiment of an on-chip calibration circuit for voltage calibration; Fig. 3A and Fig. 3B, respectively, depicts a diagram and a table illustrating the steps of a 4-bit digital-to-analog converter (DAC); Fig. 4 is a schematic diagram illustrating an exemplary 4-bit DAC incorporated into the on-chip calibration circuit of Fig. 2 can be installed; Fig. Figure 5 is a graph illustrating a linear search tree and a binary search tree for a 4-bit DAC; Fig. Figure 6 is a graph illustrating an exemplary binary search tree and algorithm that can be used by a 4-bit DAC incorporated into the calibration circuit of Fig. 2 is installed and operating in calibration mode; Fig. Figure 7 is a graph illustrating the voltage drop associated with the 4-bit DAC incorporated in the calibration circuit of Fig. 2 is installed and operating in calibration mode; Fig. 8 is a flowchart illustrating an on-chip calibration method; and Fig. 9 is a flowchart illustrating one embodiment of an on-chip calibration method for voltage calibration. Detailed description
[0013] As mentioned above, integrated circuits (ICs) often require the use of a reference parameter generated internally by an on-chip reference circuit (e.g., a reference voltage generated internally by a bandgap voltage reference circuit, a reference current generated by a current reference circuit, etc.). By design, this reference parameter can be set to a specific value to achieve optimal IC performance. However, chip-to-chip variations in this reference parameter can occur due to process variations, temperature-related variations, age-related variations, etc. As device sizes scale (e.g., to the 14nm technology node, to the 7nm technology node, and beyond), these chip-to-chip variations in the reference parameter can be relatively large.For example, chip-to-chip variations in a reference voltage can be as high as 100 mV or even greater. Therefore, calibration of the reference parameter is often required. Calibration refers to a process of adjusting a reference parameter so that the resulting calibrated parameter falls within an acceptable range, and in particular, within an acceptable distance from a target. However, currently used calibration techniques can be time-consuming and therefore costly. For example, to compensate for chip-to-chip variations in a reference voltage, designers often use a programmable resistor array for at least one of the resistors in the bandgap voltage reference circuit.In this case, a reference voltage is measured at the output of the bandgap voltage reference circuit, and depending on the value of this reference voltage, one or more of the programmable resistors can be programmed to adjust the value of the reference voltage at the output so that the resulting calibrated voltage is closer to the target voltage for this reference voltage. Measuring the reference voltage and programming resistors to generate a calibrated voltage closer to the target voltage can be time-consuming and thus costly as the resolution is increased (more bits). Recently, self-calibration circuits have been developed to reduce calibration time.However, in such self-calibration circuits, there is generally a significant trade-off between the circuit features that enable fast calibration and the circuit features that enable high-resolution calibration.
[0014] With reference to Fig. 1, embodiments of an on-chip calibration circuit 100 are generally disclosed herein for calibrating a reference parameter 101, generated, for example, by an on-chip reference circuit (not shown), such that the resulting calibrated parameter at an output node 105 is within a half step of a digital-to-analog converter (DAC) of a target value for the reference parameter. The reference parameter 101 could be any reference parameter (e.g., voltage, current, temperature, pressure, frequency, clock, etc.) generated by a suitable reference circuit.
[0015] In any of the embodiments, calibration circuitry 100 may include a digital-to-analog converter (DAC) 130 that outputs an analog parameter 102 and has an output parameter adjustment circuitry 135; a comparator 110 that receives the reference parameter 101 and the analog parameter 102; and a control circuitry 120 operatively connected to DAC 130 and comparator 110 in a feedback loop 140. Calibration circuitry 100 may also include DAC step selection logic 150, either incorporated into control circuitry 120 (as shown) or in communication with control circuitry 120.
[0016] This calibration circuit 100 can be operated either in a calibration mode or an operating mode.
[0017] In the calibration mode, the output parameter adjustment circuit 135 of the DAC 130 can be selectively enabled by the control circuit 120 to automatically adjust the magnitude of the analog parameter 102 output by the DAC 130 by approximately (and ideally exactly) half a DAC step (i.e., half the DAC step for the least significant bit (LSB)) in a first direction. For example, in the embodiment described and illustrated in the figures, the magnitude of the analog parameter 102 can be automatically increased by approximately (and ideally exactly) half a DAC step in the calibration mode. Alternatively, the magnitude of the analog parameter 102 could be automatically decreased by approximately (and ideally exactly) half a DAC step in the calibration mode.
[0018] The control circuit 120 may then use the feedback loop 140 to perform a calibration process to determine whether the parameter to be calibrated (particularly, to be adjusted to closely mirror the reference parameter) is within the calibration range of the calibration circuit 100, and to further identify a preliminary DAC step that is closest to the target (particularly, the reference parameter 101) given the DAC half-step adjustment (e.g., increase) in the analog parameter 102 during the calibration mode. This calibration process may, for example, be a binary search calibration process in which the reference parameter 101 is compared by the comparator 110 to a series of analog parameters output by the DAC 130 in response to a series of different calibration DAC codes.Based on the results of these comparisons, control circuit 120 can determine whether the parameter to be calibrated falls within the calibration range of calibration circuit 100 and, if so, can identify a preliminary DAC step. An analog parameter 102 generated by DAC 130 in calibration mode using a DAC code corresponding to this preliminary DAC step falls within a single DAC step of the target (i.e., the value of reference parameter 101). In particular, the binary search calibration process performed in calibration mode provides a calibration resolution of one DAC step.
[0019] The embodiments disclosed herein further include recalibration processing to improve calibration resolution, and more specifically, to provide a calibration resolution of 1 / 2 a DAC step (i.e., half the DAC step of the least significant bit). In particular, the output parameter adjustment circuit 135 of the DAC 130 can be selectively disabled by the control circuit 120 in the operating mode to adjust the magnitude of the analog parameter 102 output by the DAC 130 by approximately (and ideally exactly) half a DAC step in a second direction opposite the first direction.
[0020] For example, in the embodiment described and illustrated in the figures, the magnitude of the analog parameter 102 may be automatically decreased by approximately (and ideally exactly) half a DAC step in the operating mode. Alternatively, if the magnitude of the analog parameter 102 is automatically decreased by half a DAC step in the calibration mode, it could be automatically increased by half a DAC step in the operating mode.
[0021] Given the deactivation of output parameter adjustment circuitry 135, DAC step selection logic 150 can select either the preliminary DAC step previously identified during the calibration process or an adjacent DAC step as the final DAC step for use in post-calibration operation. It should be understood that if the analog parameter magnitude is increased by half a DAC step in calibration mode and decreased by half a DAC step in operation mode, DAC step selection logic 150 selects either the preliminary DAC step or the next higher DAC step as the final DAC step for use in post-calibration operation.That is, knowing that, in response to the preliminary DAC code, the analog parameter 102 output by the DAC 130 (and thereby the calibrated parameter at output node 105 of the calibration circuit 100) will be one-half DAC step lower in the run mode than that from the calibration mode, a determination is made as to whether the last DAC step is the preliminary DAC step identified during the calibration process or represents the next higher DAC step. However, it should be understood that if the magnitude of the analog parameter is decreased by one-half DAC step in the calibration mode and increased by one-half DAC step in the run mode, the DAC step selection logic 150 selects either the preliminary DAC step or the next lower DAC step as the last DAC step for use after calibration.
[0022] In any case, the control circuit 120 then outputs a final DAC code 114 corresponding to the final DAC step, and in response to the final DAC code 114, the DAC 130 generates a final analog parameter, and specifically, a calibrated parameter. Note that the final DAC step selected by the DAC step selection logic 150 is the DAC step statistically most likely to result in a final analog parameter (i.e., a calibrated parameter) at the output node 105 whose value is closest to the target, knowing that the final DAC code 114 during the operational mode generates a final analog parameter adjusted by half a DAC step (e.g., half a DAC step lower than what would have been achieved using the same DAC code in the calibration mode).
[0023] The reference parameter 101 can, for example, be a known fixed reference voltage. Such a reference voltage can, for example, be generated by a bandgap voltage reference circuit or by a resistive divider network from a known supply voltage. Alternatively, the reference parameter 101 could be any known fixed reference parameter (e.g., a known fixed current, a known fixed temperature, a known fixed pressure, a known fixed frequency, a known fixed clock speed, etc.) generated by a suitable reference circuit.
[0024] For illustration, a voltage calibration circuit 200 specifically configured to calibrate a voltage output by a digital-to-analog converter (DAC) 230 is described in more detail below and in Fig. 2. For illustrative purposes, this particular embodiment of a voltage calibration circuit provides an increase in voltage by half a DAC step in calibration mode and a decrease in voltage by half a DAC step in operation mode. However, it should be understood that embodiments of a voltage calibration circuit are also anticipated that provide a decrease in voltage by half a DAC step in calibration mode and an increase in voltage by half a DAC step in operation mode.
[0025] In any case, this calibration circuit 200 may include a digital-to-analog converter (DAC) 230 that outputs an analog voltage 202 and has an output voltage adjustment circuit 235. The calibration circuit 200 may further include a voltage comparator 210 that receives a known fixed reference voltage 201 (e.g., 600 mV) and the analog voltage 202. The calibration circuit 200 may further include a control circuit 220, which may include, for example, a tester 225 (e.g., a built-in self-test machine with test pins), a binary adder / subtractor 221, and a shift register 222 that is in communication with both the tester and the binary adder / subtractor 221 and operably connected to the voltage comparator 210 in a feedback loop 240.The calibration circuit 200 may also include a DAC step selection logic 250 that is either incorporated into or in communication with the control circuit 220 (as shown).
[0026] In calibration mode, the output voltage adjustment circuit 235 of the DAC 230 can be selectively enabled by the control circuit 220 to automatically increase the magnitude of the analog voltage 202 output by the DAC 230 by approximately (and ideally by exactly) half a DAC step (i.e., half the DAC step with the least significant bit (LSB)). The control circuit 220 can then use the feedback loop 240 to perform a calibration process, and in particular, a binary search calibration process to determine whether the voltage to be calibrated is within a calibration range of the calibration circuit 200, and further to identify a preliminary DAC step that is closest to the target (i.e., the reference voltage 201) given the ½ DAC step increase in the analog voltage 202 in calibration mode. In this binary search calibration process, the reference voltage 201 (e.g.,600 mV) from comparator 210 is compared with a series of analog voltages 202 output by DAC 230 in response to a series of different calibration DAC codes 204 from control circuit 220. Based on the results of these comparisons, control circuit 220 can determine whether the voltage to be calibrated is within the calibration range of calibration circuit 200 and, if so, can further identify the preliminary DAC step. An analog voltage 202 generated by DAC 230 in calibration mode using a DAC code corresponding to this preliminary DAC step falls within a single DAC step of the target (i.e., reference voltage 201). That is, the binary search calibration process performed in calibration mode provides a calibration resolution of one DAC step.
[0027] The embodiments disclosed herein further include post-calibration processing to improve calibration resolution, and more specifically, to provide a calibration resolution of half a DAC step. In particular, the output voltage adjustment circuit 235 of the DAC 230 can be selectively disabled in the operational mode by the control circuit 220 to automatically decrease the magnitude of the analog voltage 202 output by the DAC 230 by approximately (and ideally exactly) half a DAC step (i.e., half the least significant bit (LSB) DAC step). In light of the disabling of the output voltage adjustment circuit 235, the DAC step selection logic 250 selects either the preliminary DAC step previously identified in the calibration process or the next higher DAC step as the final DAC step for operational use after calibration.That is, knowing that, in response to the preliminary DAC code, the analog voltage 202 output by DAC 230 (and thus the calibrated voltage at output node 205 of calibration circuit 200) is half a DAC step lower in the operating mode than in the calibration mode, a determination is made as to whether the final DAC step should be the preliminary DAC step identified in the last cycle of the binary search calibration process or the next higher DAC step. Then, control circuit 220 outputs a final DAC code 214 corresponding to the final DAC step, and in response to the final DAC code 204, DAC 230 generates a final analog voltage, and specifically, a calibrated voltage (Vo), at the output. It should be noted that the last DAC step selected by the DAC step selection logic 250 is the DAC step that is statistically most likely to result in a last analog voltage (ie,a calibrated voltage (Vo) at output node 205) whose value is closest to a target voltage when the final DAC code 214 in the operating mode is known to produce a final analog voltage that is half a DAC step lower than what would have been achieved using the same DAC code in the calibration mode.
[0028] In particular, in the calibration circuit 200, the reference voltage 201 can be generated, for example, by an on-chip bandgap reference voltage circuit or a resistor divider network from a known supply voltage (not shown), which is optionally configured to generate multiple bandgap reference voltages with different values (e.g., V REF1-REFN). Such bandgap reference voltage circuits are well known in the art, and therefore, the details have been omitted from this description to allow the reader to focus on the salient aspects of the disclosed embodiments.
[0029] The DAC 230 may be configured to output an analog voltage 202 to the comparator 210, and the magnitude of this analog voltage 202 may vary depending on the DAC code received by the DAC 230 from the control circuit 220 and further depending on whether the output voltage adjustment circuit 235 has been enabled or not. Those skilled in the art will appreciate that the magnitude of the analog voltage from a DAC generally lies within a predetermined voltage range (Vrange) from a minimum voltage (Vmin) to a maximum voltage (Vmax), with the midpoint being the target (i.e., the reference voltage, for example, 600 mV) when the calibration circuit 200 is in calibration mode and the output voltage adjustment circuit 235 is enabled (as explained in more detail below), and where the minimum and maximum voltages represent the expected maximum variation in voltage.Various possible magnitudes for the analog voltage are typically set to uniform DAC stages (also referred to herein as DAC levels) over this voltage range, with each DAC stage corresponding to a different DAC code (i.e., with each DAC stage being output by the DAC in response to a different DAC code).
[0030] Those skilled in the art will recognize that the number (N) of bits in the DAC code defines the number of different DAC steps achievable by the DAC. In particular, an N-bit DAC code yields 2 (N) DAC steps. For example, a 4-bit DAC, as shown in the graph or table of the Fig. 3A and Fig. 3B shown, 2 4 (or 16) DAC steps and is thereby configured to output 16 different analog output voltages 202, which are stepped over a specific voltage range (Vrange) (e.g., in steps of 5 mV). To use one of the Fig. 3A-3B, a 4-bit DAC code would be required. Eight or fewer DAC steps would require a fewer number of bits, and more than sixteen DAC steps would require a greater number of bits than four. Consequently, the magnitude of a DAC's output voltage at DAC step 1 in response to DAC code "0001" is one full DAC step (e.g., 5 mV) less than the magnitude of the DAC's output voltage at DAC step 2 in response to DAC code "0010"; the magnitude of the DAC's output voltage at DAC step 2 in response to DAC code "0010" is one full DAC step (e.g., 5 mV) less than the magnitude of the DAC 130's output voltage at DAC step 3 in response to DAC code "0011"; and so on.
[0031] Before entering calibration mode, the output voltage adjustment circuit 235 is not activated, and Vrange is set to some value between Vmin and Vmax. When the output voltage adjustment circuit 235 is activated in calibration mode, Vrange shifts upward by half of a DAC step, and consequently, the various analog voltages at the various DAC steps also shift upward. For example, Vrange can be 557.5 mV to 632.5 mV (with a midpoint of 597.5 mV), and the DAC steps can all be 5 mV above Vrange. When the output adjustment circuit 235 is enabled, the V range can shift up by 2.5 mV to 560 mV - 635 mV (with a center point of 600 mV, which is equal to the target) and each DAC step within this V range will also be shifted up (e.g., increased by 2.5 mV or ½ DAC step).If the output adjustment circuit 235 is subsequently disabled in the operating mode, the Vrange can be shifted down by 2.5 mV (e.g., Vrange goes back to 557.5 mV - 632.5 mV) and each DAC step within the Vrange can also be shifted down (e.g., decreased by 2.5 mV).
[0032] Fig. Figure 4 is a schematic diagram illustrating an exemplary 4-bit DAC 230 configured to provide different output voltages at the Fig. 3A - Fig. 3B, and is further configured so that the output voltage adjustment circuit 235 automatically increases or decreases the DAC output by 1 / 2 a DAC step depending on the mode. Specifically, this exemplary DAC 230 includes the following components: four switches 401 0-3in the form of pairs of N-type field-effect transistors (NFET pairs) connected to a weighted resistor ladder network (see e.g. resistors 402 0-3 ) and a positive voltage. The positive voltage 405 could, for example, be generated by a bandgap reference voltage circuit such that it is fixed and known (e.g., in the same way as the reference voltage 201) or can be an internally generated voltage that varies from chip to chip and consequently requires calibration as described above. Each switch 401 0-3can receive a corresponding bit of the bits from the received DAC code, turning the switch on in the case of a bit value of "1" and turning the switch off in the case of a bit value of "0." Depending on the value of the various bits, from the most significant bit (MSB) to the least significant bit (LSB), the magnitude of the analog voltage 202 output by the DAC 230 can be selectively varied (e.g., by DAC steps of 5 mV over the full voltage range).
[0033] The output voltage adjustment circuit 235 may include an inverter 415 that receives the mode selection signal 236 (CALSTARTP) (e.g., from the tester 225). The output adjustment circuit 235 may further include an additional resistor 412 in the resistor ladder network and an additional switch 411 coupled to the additional resistor 412, with the switches 401. 0-3connected in series and controlled by an on / off signal output by inverter 415. In this configuration, a low mode selection signal 236 enables the output voltage adjustment circuit 235 (in the operating mode) by causing the on / off signal 414 to go high and the additional switch 411 to turn on, thereby pulling down the analog voltage 202 (i.e., the calibrated voltage (Vo) at the output node 205). A high mode selection signal 236 deactivates the output adjustment circuit 235 (in the calibration mode) by causing the on / off signal 414 to go low and the additional switch 411 to turn off, thereby pulling down the analog voltage 202 (i.e., the calibrated voltage (Vo) at the output node 205). It should be noted that the additional resistor 412 is relatively high relative to the resistors 402 0-3can be dimensioned such that when the additional switch 411 is off, the calibrated voltage (Vo) at the output node 205 is increased by approximately (and ideally exactly) one-half of a DAC step, and vice versa. One skilled in the art would recognize that if a fixed current is injected into a resistor ladder with resistors that are binary weights of a resistance R, it produces a linear step as the DAC increases from minimum to maximum. For example, if resistor 4023 is controlled by the DAC MSB and set to resistance R, resistor 4022 should be set to R / 2, resistor 4021 to R / 4, and LSB resistor 4020 to R / 8. Based on this, resistor 412, which is sized to add or subtract ½ a DAC step, should be sized to R / 16, where R represents the resistance associated with MSB DAC resistor 4023.
[0034] The control circuit 220, in calibration mode (after the output adjustment circuit 235 has been enabled and Vrange has been shifted upward), as mentioned above, uses the feedback loop 240 to perform a binary search calibration process to determine if the voltage to be calibrated is within the calibration range of the calibration circuit 200 and further, if so, to identify a preliminary DAC step and, in particular, to preliminary identify a particular DAC step (and corresponding DAC code) that can be used by the DAC 230 to generate an analog voltage 202 whose value is closest to a target voltage.
[0035] Those skilled in the art will recognize that a binary search calibration process significantly reduces the maximum possible number of cycles that may be required to identify an optimal DAC step. Specifically, in a conventional linear search calibration process, the analog voltage at each DAC step is sequentially compared (e.g., up or down) with the reference voltage to identify the DAC step that produces the calibration voltage closest in magnitude to the reference voltage. In this case, the maximum possible number of cycles (comparisons by a comparator) is equal to the maximum number of DAC steps. Thus, as in the Fig. 5, 16 cycles may be required to complete the calibration if a DAC has 0-15 DAC steps. However, in a binary search calibration process, the reference voltage is first compared with a midpoint calibration voltage (Vmid) within the voltage range (Vrange). Based on the results of this comparison, half of the range in which the reference voltage cannot lie is excluded from further consideration, and the search continues on the remaining half until it is successful. Thus, the number of cycles required to complete the calibration is reduced to log2(X), where X is the number of DAC steps. As shown in the Fig. 5, the number of cycles required to complete the calibration process for a 4-bit DAC with 0-15 steps is log2(16) (i.e., 4 cycles).
[0036] In particular, in the Fig. 2, the outputs of the tester 225 may include, without limitation: the mode selection signal 236 indicating the calibration mode or the operating mode; the adjusted DAC level signal 215 (DAC<N: 0> ) (e.g., for a 4-bit DAC 230) and the HOLD signal 216, which is used to sample / hold the calibration result at each cycle in the calibration process after a certain amount of time / delay. The inputs to the tester 225 may include, without limitation: the DAC code 204 / 214 output by the binary adder / subtractor 221 at each cycle of the calibration process during the calibration mode and during the run mode; and the pass / fail flag 203, which is output at each cycle in the calibration process during the calibration mode and indicates the calibration result at each cycle.
[0037] To initiate the calibration mode, the control circuit 220, and in particular the tester 225, switches the mode selection signal 236 from low to high to activate the output voltage adjustment circuit 235 of the DAC 230, as explained above.
[0038] The binary search calibration process can then be continued as described below with reference to the Fig. 6 is discussed in detail in the exemplary binary search calibration algorithm (decision tree) 602 shown. As in Fig. 6, this binary search algorithm represents the level of the DAC step in the current cycle as the level of the DAC step in the previous cycle plus K*(X / (2 Count )). In this equation, X is the total number of DAC steps possible with the DAC in question and is equal to 2 Nwhere N is the number of bits in the DAC in question (e.g., X = 16 for a 4-bit DAC, as shown). Also, K in this equation is either +1 (if there was an error in the previous cycle) or -1 (if there was a pass in the previous cycle). Finally, in this equation, Count represents the number in the sequence for the binary number of the search operation in the current cycle (i.e., search cycle 1, search cycle 2, through search cycle N for an N-bit DAC).
[0039] The control circuit 220 optionally determines whether the voltage to be calibrated is outside the calibration range of the calibration circuit 220.
[0040] To achieve this, tester 225 outputs a set DAC level signal 215 to trigger a preliminary cycle at the lowest possible DAC level, causing adder / subtractor 221 to output a "0000" code (DAC step 0). DAC 230 converts this "0000" code into an analog voltage 202. Comparator 210 compares analog voltage 202 with reference voltage 201 and outputs a pass / fail flag 203 with a value of "1," indicating that analog voltage 202 is higher than reference voltage 201, and with a value of "0," indicating that it is lower. Passing this preliminary cycle indicates that the voltage to be calibrated is outside the calibration range of the calibration circuit. This concludes the calibration process. An error in this preliminary cycle indicates that the voltage being calibrated is above the minimum voltage of the range.Thus, the binary search calibration process can continue.
[0041] Next, the tester 225 outputs a set DAC level signal 215 to set the DAC level for the first cycle in the middle DAC step. For example, using the algorithm mentioned above, this DAC step in the first cycle may be DAC step 8 (e.g., S0 = 0 + 1(16 / (2 1)). In response, the binary adder / subtractor 221 outputs a DAC code 204 corresponding to this middle DAC step, and more specifically, a DAC code with an MSB at "1" and all remaining bits at "0" (i.e., "1000"). The DAC 230 converts this DAC code "1000" into an analog voltage 202 at the middle voltage (Vmid) (e.g., 600 mV) within the switched-up V range (e.g., 560-635 mV), assuming a 5 mV / DAC step. Comparator 210 compares analog voltage 202 to reference voltage 201 at this first DAC step in the cycle and outputs a pass / fail flag 203 with a value of "1" indicating a pass if analog voltage 202 is greater than reference voltage 201, and with a value of "0" indicating a fail if it is lower. Pass / fail flag 203 is fed back to tester 225.After a certain delay, the tester 225 sets the HOLD signal 216 and checks whether the Pass / Fail flag 203 is a "1" (ie, a pass) or a "0" (ie, a fail).
[0042] In the case of a pass in the first cycle, all DAC steps above the middle DAC step are excluded from further consideration. In the case of a fail in the first cycle, all DAC steps below the middle DAC step are excluded from further consideration.
[0043] Given the pass / fail flag 203 from the first cycle, the control circuit 220 (using the tester 225, the shift register 222, and the binary adder / subtractor 221) outputs a new DAC code 204 for the second cycle. In the DAC code 204 for the second cycle, the MSB will be a "1" if the pass / fail flag 203 from the first cycle indicates a pass, and will be a "0" if the pass / fail flag 203 from the first cycle indicates a fail. The second MSB will be a "1," and all remaining bits will be a "0." Specifically, tester 225 maintains DAC <3:0> at "1000," and shift register 222 shifts the "1" in the MSB right by the count, inserting a "0" into the MSB each time. For example, shift register 222 takes "1000" and shifts right once to obtain "0100" as offset value 217. This offset value 217 is combined with DAC <3:0> at "1000" in the binary adder / subtractor.Whether the offset value 217 is added to or subtracted from the DAC <3:0> at "1000" is determined by the value of the pass / fail flag 203 from the previous cycle. Thus, in the second cycle, if the pass / fail flag 203 from the first cycle was a "1," the offset value 217 is subtracted, and the DAC code output by the adder / subtractor 221 becomes "0100" (8-4 = 4). However, if the pass / fail flag 203 from the first cycle was a "0," the offset value 217 is added, and the DAC code output by the adder / subtractor 221 becomes "1100." In other words, if the Pass / Fail flag 203 from the first cycle was a pass, then the DAC code for the second cycle is “0100”, which corresponds to DAC step 4 (i.e., S1P = S0-1(16 / (2. 2 )) However, if the Pass / Fail flag 203 from the first cycle was a fail, the DAC code for the second cycle is “1100”, which corresponds to DAC step 12 (ie S1F = S0+1(16) / (2 2)). One skilled in the art would understand that the right-shift operator can be used to add or subtract binary weights (powers of 2) to the previous DAC result. Any other suitable technique for generating the offset value 217 could be used.
[0044] The DAC 230 then converts the DAC code for the second cycle into a new analog voltage 202. The comparator 210 compares the new analog voltage 202 in this DAC step of the second cycle with the reference voltage 201 and outputs a new pass / fail flag 203 with a value of "1" indicating a pass if the new analog voltage 202 is higher than the reference voltage 201, and with a value of "0" indicating a fail if it is lower.
[0045] Given the pass / fail flag from the first and second cycles, the control circuit 220 (using the tester 225, the shift register 222, and the binary adder / subtractor 221, as described above) outputs a new DAC code for the third cycle. In the DAC code for the third cycle, the MSB is the same as that used in the second cycle; the second MSB is "0" in the case of a second cycle pass and "1" in the case of a second cycle fail; the third MSB becomes a "1," and the least significant bit (LSB) becomes a "0." Consequently, the DAC code for the third cycle could be one of the following: "0010" corresponding to DAC step 2 if the pass / fail flag from the first and second cycles were passed (i.e., S2PP = S1P-1 (16 / (2 3)); "0110" corresponding to DAC step 6 if the pass / fail flag from the first cycle was a pass and the pass / fail flag from the second cycle was a fail (ie S2PF = S1P+1(16) / (2 3 ))); "1010" corresponding to DAC step 10 if the pass / fail flag from the first cycle was a fail and the pass / fail flag from the second cycle was a fail (ie S2FP = S1F-1(16 / (2 3 ))) and “1110” corresponding to DAC step 14 if the Pass / Fail flag from the first cycle and the second cycle was a fail (ie S2FF = S1F+1(16 / (2 3 )). The comparator 210 may compare the new analog voltage 202 in this DAC step in the third cycle with the reference voltage 201 and output a new pass / fail flag 203 with a value of "1" indicating a pass if the analog voltage 202 is higher than the reference voltage 201 and with a value of "0" indicating a fail if it is lower.
[0046] Given the pass / fail flags from the first, second, and third cycles, the control circuit 220 (which in turn uses the tester 225, the shift register 222, and the binary adder / subtractor 221, as described above) can output a new DAC code for the fourth cycle. In the new DAC code for the fourth cycle, the MSB is the same as that used in the second cycle, the second MSB is the same as that used in the third cycle, the third MSB is "0" in the case of a pass in the third cycle and "1" in the case of a fail in the third cycle, and the LSB becomes "1." Consequently, the new DAC code for the fourth cycle could be one of: "0001" corresponding to DAC step 1 if the pass / fail flags from the first, second, and third cycles resulted in a pass (i.e., S3PPP = S2PP-1(16 / (2 4))); "0011" corresponding to DAC step 3, if the pass / fail flags from the first and second cycle became a pass and the pass / fail flag from the third cycle was a fail (ie S3PPF = S2PP+1(16 / (2 4 )) ); "0101" corresponding to DAC step 5, if the pass / fail flags from the first and third cycles became a pass and the pass / fail flag from the second cycle is a fail (ie S3PFP = S2PF-1(16 / (2 4 ))); "0111" corresponding to DAC step 7 if the pass / fail flag from the first cycle was a pass and the pass / fail flags from the second and third cycles were a fail (ie S3PFF = S2PF+1 (16 / (2 4 ))); "1001" corresponding to DAC step 9 if the pass / fail flag from the first cycle was a fail and the pass / fail flags from the second and third cycles were a pass (ie S3FPP = S2FP-1(16 / (2 4)); "1011" corresponding to DAC step 11 if the pass / fail flags from the first and third cycles were a fail and the pass / fail flag from the second cycle was a pass (ie S3FPF = S2FP+1(16 / (2 4 )); "1101" corresponding to DAC step 13 if the pass / fail flags from the first and second cycles were fails and the pass / fail flag from the third cycle was a pass (ie S3FFP = S2FF-1(16 / (2 4 ))) and “1111” corresponding to DAC step 15 if the Pass / Fail flags from the first, second and third cycles are fails (ie S3FFF = S2FF+1(16 / (2 4 )).
[0047] In any case, the DAC code output by control circuit 220 for the final cycle during calibration mode (in this example, the DAC code selected for use in the fourth cycle of the binary search calibration process) is the solution of the binary search calibration process and, more specifically, will correspond to the preliminary DAC step identified by the binary search calibration process. Note that to determine the final DAC step used in operation mode, as discussed below, a final comparison is made during calibration mode between reference voltage 201 and the analog voltage 202 output by DAC 230 in the preliminary DAC step (i.e., the DAC step identified by the binary search operation) when output voltage adjustment circuit 235 is still enabled to determine the pass / fail status.As explained in more detail below with respect to the operating mode, when the result is a pass "1", the optimal DAC becomes the preliminary DAC step + 1, since the DAC half-step reduces the final voltage by ½ DAC step; whereas when the result is a fail, "0", the optimal DAC becomes the preliminary DAC step.
[0048] Once the control circuit 220 has completed the binary search calibration process, the operation of the calibration circuit 200 switches from the calibration mode to the operation mode (e.g., the tester 225 may toggle the value of the mode selection signal 236 from low to high), which disables the output voltage adjustment circuit 235 of the DAC 230 so that the magnitude of an analog voltage output by the DAC 230 in response to any of the DAC codes "0000" through "1111" is automatically readjusted, and in particular, reduced by approximately (and ideally exactly) half the DAC step (e.g., by 2.5 mV) from what it was during the calibration mode. Fig. Figure 7 is a graph comparing the analog voltage output by the DAC 230 at each of the DAC steps 0-15 when the output voltage adjustment circuit 235 is enabled during the calibration mode (as is also shown in the binary search tree of Fig. 6) and when the output voltage adjustment circuit 235 is disabled during the operating mode. In this example, Vrange shifts downward to 557.5 mV (Vmin) to 632.5 mV (Vmax), with Vmid being 597.5 mV.
[0049] In the operational mode, the DAC selection logic 250 may perform post-calibration processing to provide an additional half DAC step of calibration resolution. That is, the DAC step selection logic 250 may determine the optimal final DAC step to be used to generate a final analog voltage (i.e., the calibrated voltage) for operational post-calibration use, given the disabling of the output voltage adjustment circuit 235 in the operational mode (i.e., assuming that the final DAC code in the operational mode causes the DAC 230 to generate a final analog voltage that is 1 / 2 a DAC step, e.g., 2.5 mV less than would have been the case in the calibration mode).
[0050] The selection of the last DAC step is based on whether the signal output by comparator 210 in the last cycle in the binary search calibration process (e.g., during the fourth cycle in the given example) resulted in a pass / fail flag 203 with a value of "1" (i.e., a pass, indicating that the reference voltage 201 was below the fourth cycle analog voltage output by DAC 230) or a pass / fail flag 203 with a value of "0" (i.e., a fail, indicating that the reference voltage 201 was above the fourth cycle analog voltage output by DAC 230). Given the value of the last pass / fail flag asserted during calibration mode, DAC step selection logic 250 selects either the preliminary DAC step previously identified during the binary search calibration process or the next higher DAC step as the final DAC step for operational use after calibration.In particular, if the last pass / fail flag asserted in calibration mode is a pass, DAC selection logic 250 selects the preliminary DAC step as the final DAC step because the voltage shift of half the DAC step downward in run mode moves the calibrated voltage (i.e., the final analog voltage from DAC 230 at output node 205) statistically closer to reference voltage 201, which is now known to be below the preliminary DAC step and above the next lower DAC step. In contrast, if the final pass / fail flag output in calibration mode is a fail, the DAC selection logic 250 selects the next higher DAC step above the preliminary DAC step as the final DAC step because it is now known that the reference voltage 201 is between the preliminary DAC step and the next higher DAC step.In this case, shifting the half-DAC step voltage in run mode moves the calibrated voltage (i.e., the final analog voltage from DAC 230 at output node 205) closer to reference voltage 201 because the next higher DAC step (i.e., the preliminary DAC step + 1) is selected; if the preliminary DAC step were to be selected instead, the half-DAC step voltage in run mode would actually shift the calibrated voltage further down.
[0051] Referring to the graph of Fig. 7, for example, consider the case where the preliminary DAC step identified in the fourth cycle of a binary search calibration process was DAC step 7. When the code "0111" corresponding to DAC step 7 is passed to DAC 230, an analog voltage 202 having a magnitude S3PFF is generated and output to comparator 210. Comparator 210 compares the fourth cycle analog voltage 202 to the reference voltage 201. If a pass is indicated, DAC step 7 is selected by DAC step selection logic 250 as the last DAC step in the recalibration processing because the voltage shift of half the DAC step in the run mode brings the calibrated voltage at output node 205 closer to the reference voltage 201, now known to be between DAC steps 6 and 7.When an error is indicated, DAC step 8 is selected by DAC step selection logic 250 as the last DAC step because reference voltage 201 is now known to be between DAC steps 7 and 8. A shift in voltage in the operating mode of half a DAC step downward moves the calibrated voltage at output node 205 closer to reference voltage 201. In other words, the last DAC step selected by DAC step selection logic 250 is the DAC step that is statistically most likely to be closest to the target voltage, given the additional resolution achievable due to output voltage adjustment circuitry 235.
[0052] Referring to the flowchart of Fig. 8, embodiments of a calibration method are also disclosed herein. The method may include providing a calibration circuit (see calibration circuits 100, 200 and process 801 described in detail above). In particular, calibration circuit 100, 200 may include: an N-bit digital-to-analog converter (DAC) 130, 230 that outputs an analog parameter 102, 202 and that has an output parameter adjustment circuit 135, 235; a comparator 110, 210 that receives a reference parameter 101, 201 and the analog parameter 102, 202; and a control circuit 120, 220 operably connected to comparator 110, 210 and DAC 130, 230 in a feedback loop 140, 240. The calibration circuit 100, 200 may also include a DAC step selection logic 150, 250 that is either built into the control circuit or connected to the control circuit.As mentioned above, the reference parameter and the analog parameter may be a reference voltage and an analog voltage, respectively (as specifically described in the calibration circuit 200 of FIG. Fig. 2 and described in detail above). Alternatively, the reference parameter and the analog parameter may be any other parameter that requires calibration (e.g., current, temperature, pressure, frequency, timing, etc.).
[0053] The method may further comprise selectively operating the calibration circuit 100, 200 in either a calibration mode (see process step 810) or an operation mode (see process step 820).
[0054] Operating the calibration circuit 100, 200 in calibration mode in process step 810 may include enabling the output parameter adjustment circuit 135, 235 by the control circuit 120, 220 to automatically adjust the magnitude of the analog parameter 102, 202 output by the DAC 130, 230 approximately (and ideally exactly) by half the DAC step (see process step 814).
[0055] For example, in the embodiment described and illustrated in the figures, the magnitude of the analog parameter 102, 202 may be automatically increased by approximately (and ideally by exactly) half a DAC step in calibration mode. In this case, the magnitude of the analog parameter 102, 202 is automatically decreased by approximately (and ideally by exactly) half a DAC step in operation mode (see process step 830, discussed further below). Alternatively, the magnitude of the analog parameter 102, 202 could be automatically decreased by approximately (and ideally by exactly) half a DAC step in calibration mode in process 814 and automatically increased by approximately (and ideally by exactly) half a DAC step in operation mode in process 830.
[0056] Operating the calibration circuit 100, 200 in calibration mode may further include using the feedback loop 140, 240 by the control circuit 120, 220 to perform a binary search calibration process to determine whether the parameter to be calibrated is within the calibration range of the calibration circuit 100, 200 and, if so, to identify a preliminary DAC step and, in particular, to preliminary identify a particular DAC step (and corresponding DAC code) that can be used by the DAC 130, 230 to generate an analog parameter 102, 202 whose value is closest to the reference parameter 101, 201 (see process 818 and the exemplary binary search calibration process described in detail above with reference to the Fig. 3A - 6 was discussed).
[0057] Once the preliminary DAC step is identified in process step 818, the operation of the calibration circuit 100, 200 may be selectively switched from the calibration mode to the operational mode (see process step 820). Operating the calibration circuit 100, 200 in the operational mode may also include selecting either the preliminary DAC step or an adjacent DAC step as the final DAC step for operational use after calibration by the DAC step selection logic 150, 250 (see process steps 824-826).
[0058] For example, in an embodiment where an increase of the analog parameter by half was provided in calibration mode at process 814 and a decrease of the analog parameter by half was provided in calibration mode at process 830, either the preliminary DAC step or the next higher DAC step will be selected as the final DAC step.
[0059] Specifically, in this case, to select the final DAC step, it is first determined whether a comparison performed by the comparator 110, 210 in the final cycle of the binary search calibration process in process step 818 results in a pass, indicating that the reference parameter 101, 201 was below the analog parameter output by the DAC 130, 230 of the last cycle, or a fail, indicating that the reference parameter 101, 201 was above the analog parameter output by the DAC 130, 230 of the last cycle (see process step 824). If it was a pass, the preliminary DAC step is selected as the final DAC step (see process step 825). If it was a fail, the next higher DAC step above the preliminary DAC step is selected as the final DAC step (see process step 826).It should be noted, however, that in an embodiment where a decrease in the analog parameter in calibration mode by half a DAC step is provided at process 814 and an increase in the analog parameter in operation mode by half a DAC step is provided at process 830, either the preliminary DAC step or the next lower DAC step is selected as the final DAC step.
[0060] In any case, operating the calibration circuit 100, 200 in the operating mode further comprises: outputting, by the control circuit 120, 220, a final DAC code 114, 214 corresponding to the final DAC step (see process step 828); deactivating the output parameter adjustment circuit 135, 235 by the control circuit 120, 220 to automatically decrease the magnitude of the analog parameter 102, 202 output by the DAC 130, 230 by approximately (and ideally exactly) half a DAC step (see process step 830) and then generating a final analog parameter, and in particular a calibrated parameter (e.g., a calibrated voltage) at the output node by the DAC 130, 230 in response to the final DAC code 114, 214, 105, 205 (see process step 832).
[0061] Particular attention is drawn to the flow chart of Fig. 9 in combination with the voltage calibration circuit 200 of Fig. 2. For illustrative purposes, this particular embodiment of a voltage calibration method provides for a voltage increase of half a DAC step in calibration mode and a voltage decrease of half a DAC step in operation mode. However, it should be understood that embodiments of a voltage calibration method are also contemplated that provide for a voltage decrease of half a DAC step in calibration mode and an increase of half a DAC step in operation mode.
[0062] Calibration mode begins by activating the output voltage adjustment circuit 235 to increase the analog voltage 202 output by the DAC 230 by approximately (and ideally exactly) half a DAC step (e.g., by 2.5 mV) (see process step 901).
[0063] To confirm whether the voltage to be calibrated is within the calibration range of the calibration circuit 200, the tester 225 sets the DAC 230 to the lowest DAC level (i.e., DAC step 0 with the code "0000") (see process step 902). The analog voltage 202 output by the DAC 230 in response to the code "0000" is compared with the reference voltage 201 by the comparator 210, and the comparator 210 outputs a pass / fail flag 203 indicating the results of this comparison (see process step 903). If the pass / fail flag has a value of "1," indicating a pass and, in particular, indicating that the analog voltage 202 is above the reference voltage 201, it is determined that the voltage to be calibrated is outside the calibration range of the calibration circuit 200 (see process step 904). In this case, no further processing is performed (see process step 999) and the IC chip can be discarded.If the Pass / Fail flag 203 has a value of "0", which indicates a fail and, in particular, indicates that the analog voltage 202 is below the reference voltage 201, it is determined that the voltage to be calibrated is above the minimum voltage of the calibration range of the calibration circuit 200 and the binary search calibration process can continue.
[0064] Assuming a Pass / Fail flag 203 with a value of "0", the counter value is incremented by 1 (see process step 905). The count is related to the cycle number. The minimum count value is 0 in process step 902, and since a binary search calibration process is performed, the maximum count value is N (i.e., the maximum number of cycles in the binary search process is N, where N is the number of bits allocated to the N-bit DAC 230). Incrementing the count value in process step 905 triggers a change in the DAC step according to the binary search calibration algorithm (see, for example, the Fig. 6 and discussed in more detail above). Next, the value of the count is checked to determine if the maximum count (i.e., the maximum number of cycles) has been reached (see process step 906).
[0065] If yes, the binary search calibration process is complete. If not, additional comparison cycles are required. That is, as discussed in detail above with respect to the binary search calibration algorithm / decision tree 602, the binary search calibration process requires N comparisons for an N-bit DAC (which is a reduction in test time compared to a linear search). If the count is less than the limit of N, the analog voltage 202 generated by the DAC 230 in response to the current cycle's DAC code is compared to the reference voltage 201 (see process step 907).If the analog voltage 202 generated by the DAC 230 in response to the current cycle DAC code is above the reference voltage 201, a pass / fail flag 203 with a value of "1" (pass) and K = -1 is output, so that the DAC 230 is adjusted to a lower level in the next cycle according to the binary search calibration algorithm to bring the analog voltage closer to the reference voltage (see process step 908). However, if the analog voltage 202 generated by the DAC 230 in response to the current cycle DAC code is below the reference voltage 201, a Pass / Fail flag 203 with a value of "0" (Fail) is output and K = +1, so that in the next cycle, the DAC 230 is adjusted to a higher level according to the binary search calibration algorithm to bring the analog voltage closer to the reference voltage (see process step 909). The comparison processes continue with each new cycle until the maximum number (ieNumber = N) has been reached.
[0066] If it is determined in process step 906 that the maximum count has been reached (i.e., count = N), the binary search calibration process is complete. It should be noted that if, when count = N, the DAC is set to the highest DAC step (i.e., DAC = max), and the Pass / Fail flag 203 has a value of "0" (Fail), indicating that the reference voltage 201 is still higher than the analog voltage 202 (in the last cycle), the voltage to be calibrated is still too low compared to the reference voltage 201, and it can be determined that the voltage to be calibrated is outside the calibration range (see process steps 910 and 904). In this case, no further processing is performed (see process step 999), and the IC chip can be discarded.
[0067] Otherwise, the DAC step specified for the final cycle of the binary search calibration process is a preliminary DAC step, which may or may not be used to generate a calibrated voltage for post-calibration operation. As mentioned above, an analog voltage 202 generated by the DAC 230 in calibration mode using a DAC code corresponding to this preliminary DAC step identified in the final cycle of the binary search calibration process falls within a single DAC step of the reference voltage 201. That is, the binary search calibration process executed in calibration mode provides a calibration resolution of one DAC step. This preliminary DAC step is as close as possible to the fixed and known reference voltage in the number of bits provided.
[0068] Additional post-calibration processing may be performed to improve calibration resolution and, in particular, to produce a calibrated voltage that is statistically more likely to be close to the target. Specifically, knowing that, in response to the preliminary DAC code, the analog voltage 202 output by the DAC 230 (and thereby the calibrated voltage at the output node 205 of the calibration circuit 200) is half a DAC step lower in the operating mode than in the calibration mode, a determination is made as to whether or not the final DAC step should be the preliminary DAC step identified in the last cycle of the binary search calibration process or the next higher DAC step (see process steps 920-922). That is, a determination is made as to whether the pass / fail flag 203 from the last cycle of the binary search calibration process passed or failed.If the Pass / Fail flag 203 is set to "1" (Pass), indicating that the analog voltage 202 output by the DAC 235 in the last cycle was higher than the reference voltage 201, the preliminary DAC step (i.e., the DAC step identified by the binary search calibration process) will be the optimal DAC step and will be selected as the final DAC step for operational use after calibration (see process 922). In this case, a final DAC code 214 corresponding to the preliminary DAC step is passed to the DAC (in process step 990); the operational mode is initiated, during which the output voltage adjustment circuit 235 is disabled to adjust the analog voltage 202 output by the DAC 230 by approximately (and ideally exactly) half a DAC step (e.g.,by 2.5 mV) (see process step 992); and a final analog voltage, and specifically a calibrated voltage (Vo) at output node 205 is generated in response to the final DAC code 214 (at process 995). However, if the Pass / Fail flag 203 is "0" (Fail), indicating that the analog voltage 202 output by the DAC 235 in the last cycle was lower than the reference voltage 201, then the next higher DAC step above the DAC step is identified by the binary search calibration process as the optimal DAC step and is selected as the final DAC step (see process 921). In this case, the DAC step is incremented by 1, and a final DAC code 214 corresponding to this next higher DAC step is passed to the DAC (at process 990).The operating mode is then initiated in which the output voltage adjustment circuit 235 is deactivated to decrease the analog voltage 202 output by the DAC 230 by approximately (and ideally exactly) half a DAC step (e.g., by 2.5 mV) (see process step 992); and a final analog voltage, and more specifically, a calibrated voltage (Vo) at the output node 205 is generated in response to a final DAC code 214 (at process 995).
[0069] It is understood that the terminology used herein is for the purpose of describing the disclosed structures and methods and is not intended to be limiting. For example, as used herein, the singular forms "a," "an," and "the" are intended to include the plural forms unless the context clearly indicates otherwise. Additionally, as used herein, the terms "comprises," "comprising," and / or "having" specify the presence of stated features, integers, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof. Further, terms such as "right," "left," "vertical," "horizontal," "top," "bottom," "above," "below," "below," "below," "above," "above," "above," "parallel," "perpendicular," etc. are used herein.used to describe underlying and relative locations when aligned and shown in the drawings (unless otherwise noted), and terms such as "touching", "in direct contact with", "abutting", "directly adjacent to", "adjacent", etc. are intended to indicate that at least one element is physically touching another element (without other elements separating the described elements). The term "lateral" is used herein to describe the relative positions of elements, and in particular to indicate that one element is positioned to the side of another element as opposed to above or below the other element when those elements are aligned and shown in the drawings.For example, an element positioned laterally adjacent to another element is adjacent to the other element, an element positioned laterally immediately adjacent to another element is directly adjacent to the other element, and an element laterally surrounding another element is adjacent to and bordering the outer side walls of the other element. The corresponding structures, materials, acts, and equivalents of all means or steps plus function elements in the claims below are intended to include any structure, material, or act for performing the function in combination with other claimed elements as specifically claimed.
Claims
[1] Calibration circuit (100; 200) comprising: a digital-to-analog converter (130; 230) outputting an analog parameter (102; 202) and having an output parameter setting circuit (135; 235); a comparator (110; 210) receiving a reference parameter (101; 201) and the analog parameter (102; 202); a control circuit (120; 220) operatively connected to the comparator (110; 210) and the digital-to-analog converter (130; 230) in a feedback loop (140; 240); and a digital-to-analog converter step selection logic (150; 250), wherein the calibration circuit (100; 200) is selectively operable in a calibration mode and an operating mode, wherein in the calibration mode the output parameter adjustment circuit (135; 235) adjusts a magnitude of the analog parameter (102; 202) in a first direction by approximately one-half of a digital-to-analog converter step, and the control circuit (120; 220) uses the feedback loop (140; 240) to perform a calibration process to identify a preliminary digital-to-analog converter step, and wherein in the operating mode the output parameter adjustment circuit (135; 235) adjusts the magnitude of the analog parameter (102; 202) in one step by approximately one-half of a digital-to-analog converter step in the second direction, which is opposite to the first direction, Digital-to-analog converter step selection logic (150;250) selects one of the preliminary digital-to-analog converter step and an adjacent digital-to-analog converter step as the final digital-to-analog converter step, the control circuit (120; 220) outputs a final digital-to-analog converter code corresponding to the final digital-to-analog converter step, and the digital-to-analog converter (130; 230) outputs a calibrated parameter in response to the final digital-to-analog converter code. [2] Calibration circuit (100; 200) according to claim 1, wherein the digital-to-analog converter (130; 230) comprises an N-bit digital-to-analog converter (130; 230), wherein the calibration process is a binary search calibration process performed in N cycles to identify the preliminary digital-to-analog converter step by an N-bit digital-to-analog converter code, and wherein each of the N cycles determines a bit in the N-bit digital-to-analog converter code from a first cycle determining a most significant bit to an N-th cycle determining a least significant bit. [3] The calibration circuit (100; 200) of claim 2, wherein the N-bit digital-to-analog converter is a 4-bit digital-to-analog converter such that the binary search calibration process is performed in 4 cycles. [4] Calibration circuit (100; 200) according to claim 1, wherein the control circuit (120; 220) determines in the calibration mode whether the parameter to be calibrated lies within a calibration range of the calibration circuit (100; 200). [5] Calibration circuit (200) according to claim 1, where the calibration process is a binary search calibration process, and wherein the digital-to-analog converter step selection logic (250) selects the preliminary digital-to-analog converter step as the final digital-to-analog converter step if the preliminary digital-to-analog converter step results in a pass flag (203) from the comparator (210) in a final cycle of the binary search calibration process, and selects the adjacent digital-to-analog converter step if the preliminary digital-to-analog converter step results in a fail flag (203) from the comparator (210) in the final cycle of the binary search calibration process. [6] The calibration circuit (100; 200) of claim 1, wherein the calibration process is a binary search calibration process enabled by a built-in self-test engine. [7] Calibration circuit (100; 200) according to claim 1, wherein the reference parameter (101; 201) comprises a voltage, a current, a temperature, a pressure, a frequency or a clock. [8] Calibration circuit (100; 200) comprising: a digital-to-analog converter (130; 230) outputting an analog voltage and including an output voltage adjustment circuit; a comparator (110; 210) receiving a reference voltage and the analog voltage; a control circuit (120; 220) operatively connected to the comparator (110; 210) and the digital-to-analog converter (130; 230) is connected in a feedback loop (140; 240); and a digital-to-analog converter step selection logic (150; 250), wherein the calibration circuit (100; 200) is selectively operable in a calibration mode and an operating mode, wherein, in calibration mode, the output voltage adjustment circuit increases a magnitude of the analog voltage by approximately half a digital-to-analog converter step, and the control circuit (120; 220) uses the feedback loop (140; 240) to perform a binary search calibration process to identify a preliminary digital-to-analog converter step, and wherein the output voltage setting in the operating mode decreases the magnitude of the analog voltage by approximately one-half of a digital-to-analog converter step, the digital-to-analog converter step selection logic (150; 250) selects one of the preliminary digital-to-analog converter step and a next higher digital-to-analog converter step as the final digital-to-analog converter step, the control circuit (120; 220) outputs a final digital-to-analog converter code corresponding to the final digital-to-analog converter step, and the digital-to-analog converter (130; 230) generates a calibrated voltage in response to the final digital-to-analog converter code. [9] Calibration circuit (100; 200) according to claim 8, wherein the digital-to-analog converter (130; 230) comprises an N-bit digital-to-analog converter (130; 230), wherein the binary search calibration process is performed in N cycles to identify the preliminary digital-to-analog converter step by an N-bit digital-to-analog converter code, and wherein each of the N cycles determines a bit in the N-bit digital-to-analog converter code from a first cycle determining a most significant bit to an N-th cycle determining a least significant bit. [10] The calibration circuit (100; 200) of claim 9, wherein the N-bit digital-to-analog converter is a 4-bit digital-to-analog converter such that the binary search calibration process is performed in 4 cycles. [11] Calibration circuit (100; 200) according to claim 8, wherein the control circuit (120; 220) determines in the calibration mode whether a voltage to be calibrated is within a calibration range of the calibration circuit (100; 200). [12] The calibration circuit (200) of claim 8, wherein the digital-to-analog converter step selection logic (250) selects the preliminary digital-to-analog converter step as the last digital-to-analog converter step when the preliminary digital-to-analog converter step results in a pass flag (203) from the comparator (210) in a last cycle of the binary search calibration process, and selects the next higher digital-to-analog converter step when the preliminary digital-to-analog converter step results in a fail flag (203) from the comparator (210) in the last cycle of the binary search calibration process. [13] The calibration circuit (100; 200) of claim 8, wherein each digital-to-analog converter step is approximately 5 mV, wherein the output voltage adjustment circuit increases the magnitude of the analog voltage by approximately 2.5 mV in the calibration mode, and wherein the output voltage adjustment circuit decreases the magnitude of the analog voltage by approximately 2.5 mV in the operation mode. [14] The calibration circuit (100; 200) of claim 8, wherein the binary search calibration process is enabled by a built-in self-test engine. [15] Calibration procedures, including: operating a calibration circuit (100; 200) in a calibration mode, wherein operating the calibration circuit (100; 200) in the calibration mode comprises: adjusting a magnitude of an analog parameter (102; 202) output by a digital-to-analog converter (130; 230) by approximately half a digital-to-analog converter step in a first direction; and using a feedback loop (140; 240) between a control circuit (120; 220) and the digital-to-analog converter to perform a binary search calibration process to identify a preliminary digital-to-analog converter step; and operating the calibration circuit (100; 200) in an operating mode, wherein operating the calibration circuit (100; 200) in the operating mode comprises: adjusting the magnitude of the analog parameter (102; 202) output by the digital-to-analog converter (130; 230) by approximately half a digital-to-analog converter step in a second direction opposite to the first direction; selecting either the preliminary digital-to-analog converter step or an adjacent digital-to-analog converter step as the final digital-to-analog converter step; outputting a final digital-to-analog converter code corresponding to the final digital-to-analog converter step; and generating a calibrated parameter in response to the final digital-to-analog converter code. [16] Method according to claim 15, wherein the digital-to-analog converter (130; 230) comprises an N-bit digital-to-analog converter (130; 230), wherein the binary search calibration process is performed in N cycles to identify the preliminary digital-to-analog converter step by an N-bit digital-to-analog converter code, and wherein each of the N cycles determines a bit in the N-bit digital-to-analog converter code from a first cycle determining a most significant bit to an N-th cycle determining a least significant bit. [17] The method of claim 16, wherein the N-bit digital-to-analog converter (130; 230) is a 4-bit digital-to-analog converter (130; 230) such that the binary search calibration process is performed in four cycles. [18] The method of claim 15, wherein operating in calibration mode further comprises determining whether a parameter to be calibrated is outside a calibration range of the calibration circuit (100; 200). [19] Method according to claim 15, wherein the size of the analog parameter (202) is increased in calibration mode, wherein the size of the analog parameter (202) is reduced in the operating mode, and wherein the preliminary digital-to-analog converter step is selected as the last digital-to-analog converter step if the preliminary digital-to-analog converter step results in a pass flag (203) from the comparator (210) in the last cycle of the binary search calibration process, and a next higher digital-to-analog converter step is selected as the last digital-to-analog converter step if the preliminary digital-to-analog converter step results in a fail flag (203) from the comparator (210) in the last cycle of the binary search calibration process. [20] The method of claim 15, wherein the analog parameter (102; 202) comprises a voltage, a current, a temperature, a pressure, a frequency, or a clock rate.
Citation Information
Patent Citations
Data conversion circuitry and method therefor
US20100079327A1
