Method and device for use in DC parametric tests

Abstract

A system and device suitable for use in performing a DC parametric test on an external load is provided. The device may be configured to apply a desired voltage or current to the external load. The circuit device receives a forcing parameter signal at an input and releases at an output a signal approximating the forcing parameter signal to the external load. The circuit device includes a first circuit segment between the input and the output having a search unit, an intermediate voltage point and an internal load between the intermediate voltage point and the output. A second circuit segment connected in a feedback arrangement with the first circuit segment provides the search unit with the voltage at the output. The search unit is adapted for generating a second voltage signal on the basis of the forcing parameter signal and the first voltage signal received and to apply the second voltage signal to the intermediate voltage point. The application of second voltage signal to the intermediate voltage point causes a change in either one of the voltage signal or the current signal at the output such that a signal approximating the forcing parameter signal is caused at the output. A current measuring circuit adapted for providing a measure of the current at the external load suitable for use with a voltage generating circuit device is also provided.

Description

FIELD OF THE INVENTION

The present invention relates generally to electronic chips and devices, and more particularly, to a method and device suitable for use in performing DC parametric tests.

BACKGROUND OF THE INVENTION

Recent years have seen a rapidly increasing demand for highly integrated mixed-signal integrated circuits (IC's). This demand is mostly driven by the ever-expanding communications industry. However, as the level of integration increases, more and more mixed-signal components are becoming buried deep inside large amounts of digital circuitry without any external I/O access. This creates a difficult problem for initial device and circuit characterisation and diagnosis, as well as during a production test. For example, to measure the bias current for a high precision ADC circuit requires some form of external access. However, the access mechanism, such as a test bus, can introduce additional noise from off-chip sources.

Typically circuit characterisation includes the determination of the electrical characteristics of a circuit such as for example measuring the input/output impedance of an amplifier circuit, or finding the voltage transfer characteristics of an amplifier circuit or transistor device amongst others.

One particular area of IC testing that is being affected is the DC parametric tests. These tests are typically conducted to characterise a wide variety of mixed-signal circuits such as Analog-to-Digital Converters (ADCs), PLLs and bias networks. Also, these tests are used in digital test applications such as pad current leakage and I_DDQ tests. For example, the pad current leakage test and the IDDQ test are common test techniques for detecting faults in digital ICs.

DC parametric tests are generally classified as one of two types. The first type of DC parametric test involves forcing a voltage at a circuit node while measuring the current that flows into the node. Commonly used method for on-chip current measurements include using device having either a transimpedance amplifier, as shown in FIG. 1a), an integrating network as shows in FIG. 1b) or a shunt resistance, as shown in FIG. 1c). For additional information regarding the above mentioned methods, the reader is invited to refer to the following documents:

• • 1. Teradyne, Inc., “Low Current Ammeter Channel Card”, Advanced Mixed-Signal Instrumentation Manual, 1996.
  • 2. C. D. Thompson, S. R. Bernadas, “A Digitally-Corrected 20b Delta-Sigma Modulator”, Proc. IEEE International Solid-State Circuits Conference, pp. 194-195, 1994.
  • 3. U.S. Pat. No. 5,274,375 issued to Charles D. Thompson Dec. 28, 1993;
  • 4. M. Breten, T. Lehmann, E. Bruun, “Integrating Data Converters for Picoampere Currents from Electrochemical Transducers”, Proc. IEEE International Symposium on Circuits and Systems, Vol. 5, pp. 709-712, May 2000.
  • 5. C. B. Wang, J. Todsen, T. Kalthoff, “A Dual Channel 20 Bit Current-Input A/D Converter for Photo-Sensor Applications”, Proc. Southwest Symposium on Mixed-Signal Design, pp. 57-60, 2000.
  • 6. Burr-Brown Product #DDC 112
  • 7. J. Kotowski, B. McIntyre, J. Parry, “Current Sensor IC Provides 9 bit+Sign Result without External Sense Resistor”, Proc. IEEE Custom Integrated Circuits Conference, pp. 35-38, 1998;
  • 8. U.S. Pat. No. 5,867,054 issued to Jeffrey P. Kotowski Feb. 2, 1999;
  • 9. National Semiconductor Product # LM3814The contents of the above documents are hereby incorporated by reference.

A deficiency of devices of the type described above is that they involves the use of elaborate Analog-to-Digital Converters (ADCs) with trimmed components, which makes these devices expensive and relatively non-scalable for on-chip implementation. Another deficiency of devices of the type described above is that they make use of op-amps (operational amplifiers) which also makes them relatively non-scalable for on-chip implementation. Generally, the size of the op-amp circuit does not shrink to the same extent as the size of logic circuits do as IC technology advances.

The second type of DC parametric test involves forcing a known current into a circuit node while measuring the voltage at the node.

A deficiency of commonly used on-chip current sources is that they generally suffer from low output resistance and shifts in current levels due to process variation. Such current sources are described in W. Sansen et al., “A CMOS Temperature-Compensated Current Reference”, IEEE Journal of Solid-State Circuits, Vol. 23, pp. 821-824, June 1988 and in H. J. Oguey et al., “CMOS Current Reference Without Resistance”, IEEE Journal of Solid-State Circuits, Vol. 32, pp. 1132-1135, July 1997 whose contents are herein incorporated by reference. Other on-chip current source implementations, of the type described in Burr-Brown Corporation, “Dual Current Source/Current Sink”, REF200 (Datasheet), October 1993 and in U.S. Pat. No. 4,792,748 issued to David M. Thomas et al. in Dec. 20, 1998, can generally achieve good current accuracy but require laser-trimmed on-chip resistors, which is costly when multiple measurement units are required on a single chip. The contents of the above documents are hereby incorporated by reference.

In the context of the above, there is a need in the industry to provide a method and device for use in performing DC parametric tests that alleviates at least in part problems associated with the existing devices and methods.

SUMMARY OF THE INVENTION

In one embodiment, the present invention is directed to a method of performing a DC parametric test on a first load. The method comprises electrically coupling an output of a second load to the first load and applying a first electrical signal to the second load so as to cause a second electrical signal at the output. The second signal is fed back to a search entity. A forcing parameter signal is input into the search entity, and the first electrical signal is determined as a function of the second electrical signal and the forcing parameter signal.

In another embodiment, the present invention is directed to a method of performing a DC parametric test on a first load. The method comprises the step of applying a first electrical signal and receiving a forcing parameter signal. The first electrical signal is caused to approximate the forcing parameter signal as a function of a feedback signal of the first electrical signal and the forcing parameter signal. A DC parametric signal of the first load is measured. The DC parametric signal is a response of the first load to the first electrical signal.

BRIEF DESCRIPTION OF THE DRAWINGS

In the accompanying drawings:

FIGS. 1 a, 1b and 1c show examples of prior art circuit devices for on-chip current measurements;

FIGS. 2 a and 2b show circuit devices suitable for use in performing a DC parametric test on an external load in accordance with non-limiting examples of implementation of the invention;

FIG. 3 is a graph showing the DC characteristics of internal load of the circuit device shown in FIG. 2b in accordance with a non-limiting example of implementation of the invention;

FIGS. 4 a and 4b show specific example of implementation of the internal load of the circuit device shown in FIG. 2b in accordance with a non-limiting example of implementation of the invention;

FIG. 5 is a graph showing the DC characteristics of external load R₂ in accordance with a non-limiting example of implementation of the invention;

FIG. 6 is a graph showing on a same graph the load lines of internal load R₁ and the load line of internal load R₂ depicted in FIGS. 3 and 5 when the loads R₁ and R₂ are connected in series;

FIG. 7 is a graph showing the location of the desired voltage to apply to the intermediate voltage point V_DAC shown in FIG. 2b of the drawings for a desired forcing parameter voltage signal V_force in accordance with a specific example of implementation of the invention;

FIG. 8 shows the circuit device of FIG. 2b adapted for searching for a voltage to apply to intermediate voltage point V_DAC for a desired forcing parameter voltage signal V_force in accordance with a non-limiting example of implementation of the invention;

FIG. 9 is a graph showing the DC characteristics of internal load R₁ of the circuit device shown in FIG. 2b and showing the location of the current measurements in accordance with a non-limiting example of implementation of the invention;

FIG. 10 is a graph illustrating a function mapping a forcing parameter signal V_force and the voltage to apply to intermediate voltage point V_DAC to a current value at the output of the circuit device shown in FIG. 8 in accordance with a non-limiting example of implementation of the invention;

FIG. 11 shows the circuit device of FIG. 8 adapted for further providing a current measurement value in accordance with a non-limiting example of implementation of the invention;

FIG. 12 shows a calibration circuit for calibrating the circuit device shown in FIG. 8 in accordance with a non-limiting example of implementation of the invention;

FIG. 13 is a graph showing a plot of k*_DAC vs. I_out in accordance with a non-limiting example of implementation of the invention;

FIG. 14 shows a test circuit for calibrating the circuit device shown in FIG. 8 including a test current reference I_test in accordance with a non-limiting example of implementation of the invention;

FIG. 15 shows an alternate calibration circuit for calibrating the circuit device shown in FIG. 8 in accordance with a non-limiting example of implementation of the invention;

FIG. 16 shows the circuit device of FIG. 2b adapted for searching for a voltage to apply to intermediate voltage point V_DAC for a desired forcing parameter current signal I_force in accordance with a non-limiting example of implementation of the invention;

FIG. 17 is a graph showing the location of a desired voltage to apply to intermediate voltage point V_ADC for a desired forcing parameter current signal I_force in accordance with a non-limiting example of implementation of the invention;

FIG. 18 shows the circuit device of FIG. 2b adapted for forcing a voltage at output V_ADC for a desired forcing parameter voltage signal I_force in accordance with a non-limiting example of implementation of the invention;

FIG. 19 is a graph showing the location of a desired voltage to apply to output V_ADC for a desired forcing parameter current signal I_out=I_force in accordance with a non-limiting example of implementation of the invention;

FIG. 20 is a graph showing the DC characteristics of internal load R₁ of the circuit device shown in FIG. 2b and showing the location of the voltage measurements for I_out=I_force in accordance with a non-limiting example of implementation of the invention;

FIG. 21 is a graph illustrating a function “H” mapping a forcing parameter signal I_force and the voltage to apply to intermediate voltage point V_DAC to a voltage value at the output of the circuit device shown in FIG. 16 in accordance with a non-limiting example of implementation of the invention;

FIG. 22 is a graph illustrating a function “H⁻¹” mapping forcing parameter signal I_force and to a voltage value at the output of the circuit device shown in FIG. 16 to a voltage to apply to intermediate voltage point V_DAC in accordance with a non-limiting example of implementation of the invention;

FIGS. 23 a), b), c) and d) show four alternate specific examples of implementation of the circuit device of FIG. 2b adapted for searching for a voltage to apply to intermediate voltage point V_DAC or V_ADC for a desired forcing parameter voltage signal I_force in accordance with non-limiting examples of implementation of the invention;

FIGS. 24 and 25 are graphs showing on a same plot the load lines of internal load R₁ and the load line of internal load R₂ when the loads R₁ and R₂ are connected in series in accordance with a non-limiting example of implementation of the invention;

FIG. 26 shows an alternate specific example of implementation of the circuit device of FIG. 2b adapted for searching for a voltage to apply to intermediate voltage point V_DAC for a desired forcing parameter voltage signal I_force in accordance with a non-limiting example of implementation of the invention;

FIGS. 27 and 28 show portions of calibrating circuits for calibrating the circuit device shown in FIG. 23 in accordance with a non-limiting example of implementation of the invention;

FIG. 29 shows the circuit device of FIG. 2b adapted for searching for a voltage to apply to intermediate voltage point V_DAC for a desired forcing parameter voltage signal V_force in accordance with a non-limiting example of implementation of the invention where the internal load R₁ is an inverting load;

FIG. 30 shows a modified version of the calibration circuit of FIG. 15 adapted for an inverting internal load R₁ in accordance with a non-limiting example of implementation of the invention;

FIG. 31 shows a modified version of the calibration circuit of FIG. 18 adapted for an inverting internal load R₁ in accordance with a non-limiting example of implementation of the invention;

FIG. 32 shows a modified version of the calibration circuit of FIG. 23a) adapted for an inverting internal load R₁ in accordance with a non-limiting example of implementation of the invention;

FIG. 33 shows a modified version of the calibration circuit of FIG. 23b) adapted for an inverting internal load R₁ in accordance with a non-limiting example of implementation of the invention;

FIG. 34 shows a modified version of the calibration circuit of FIG. 23c) adapted for an inverting internal load R₁ in accordance with a non-limiting example of implementation of the invention;

FIG. 35 shows the circuit device of FIG. 2b adapted for forcing voltage V_DAC on the basis of a forcing parameter including a linear resistive internal load R₁ in accordance with a non-limiting example of implementation of the invention;

FIG. 36 shows the circuit device of FIG. 2b adapted for forcing voltage V_ADC on the basis of a forcing parameter including a linear resistive internal load R₁ in accordance with another non-limiting example of implementation of the invention;

FIG. 37 shows the circuit device of FIG. 11 where internal load R₁ is a linear resistive load in accordance with a non-limiting example of implementation of the invention;

FIGS. 38 and 39 show calibration circuits for a linear resistive internal load R₁ suitable for use in calibrating the circuit devices of FIGS. 36 and 37 in accordance with non-limiting examples of implementation of the invention;

FIGS. 40-62 show various embodiments of the circuit device of FIG. 2b in accordance with non-limiting examples of implementation of the invention;

FIGS. 63-66 show an alternative configuration of a circuit device suitable for use in performing a DC parametric test on an external load in accordance with a non-limiting example of implementation of the invention;

FIG. 67 shows another alternative configuration of a circuit device suitable for use in performing a DC parametric test on an external load in accordance with another non-limiting example of implementation of the invention.

Other aspects and features of the present invention will become apparent to those ordinarily skilled in the art upon review of the following description of specific embodiments of the invention in conjunction with the accompanying figures.

DETAILED DESCRIPTION

With reference to FIGS. 2a and 2b, there is shown a general configuration of a system 200 suitable for use in performing a DC parametric test and an unknown external load 206 representing a circuit under test.

The system 200 includes an input 202 for receiving a forcing parameter signal, an output 204 suitable for releasing to an external load 206 a signal approximating the forcing parameter signal. The system also includes a first circuit segment between the input 202 and the output 204 and a second circuit segment connected in a feedback arrangement with the first circuit segment.

As shown in FIG. 2a, the first circuit segment includes a search entity 199, an intermediate voltage point 214 and an internal load 216 between the intermediate voltage point and the output 204. The second circuit segment provides the search entity 199 in the first circuit segment with a signal indicative of the voltage at the output 204. The search entity 199 is adapted for generating a second voltage signal on the basis of the forcing parameter signal received at input 202 and the first voltage signal received from the second segment and to apply the second voltage signal to the intermediate voltage point 214. The application of the second voltage signal to the intermediate voltage point 214 causes a change in either one of the voltage signal or the current signal at the output 204 such that a signal approximating the forcing parameter signal is caused at the output 204.

It will be appreciated to the person skilled in the art that the search entity 199 may be adapted for processing and generating signals in either digital format or analog format or a combination of both without detracting from the spirit of the invention. For example, the forcing parameter signal applied at input 202 and the second voltage signal released by the search entity 199 may also be in either digital or analog format.

In a specific configuration shown in FIG. 2b, the search entity 199 is shown including a search unit 208, an analog-to-digital converter module (ADC) 212 and a digital-to-analog converter module (DAC) 210. In this configuration, the second circuit segment includes the analog-to-digital converter module (ADC) 212 for processing the voltage at the output 214 to generate a digital approximation 85 of the voltage signal at the output 204. The first circuit segment includes the search unit 208, the intermediate voltage point 214, the internal load 216 between the intermediate voltage point 214 and the output 204 and the digital-to-analog converter module (DAC) 210. The search unit 208 is for generating a digital target voltage 87 on the basis the digital approximation of the voltage signal at the output 204 and the forcing parameter signal at input 202. The digital-to-analog converter module 210 is between the search unit 208 and the intermediate voltage point 214 and is adapted for processing the digital target voltage 87 to generate a second voltage signal and applying the second voltage signal to the intermediate voltage point 214. The application of the second voltage signal to the intermediate voltage point 214 causes a change in either one of the voltage signal and the current signal at the output 204 such that, at equilibrium, a signal approximating the forcing parameter signal is caused at the output 204.

In yet another configuration (not shown in the figures), the first path includes a search unit and a circuit module having digital-to-analog conversion functionality and load functionality. The circuit module is connected between the output of the search unit and the circuit output. A non-limiting example of a circuit module having digital-to-analog conversion functionality and load functionality is shown in FIGS. 46A and 46B which will be described in greater detail further on in the specification. Another example of a circuit module having digital-to-analog conversion functionality and load functionality is a circuit module including digital-to-analog converter module (DAC) 210 and internal load 216 connected as depicted in FIG. 2b.

With reference to FIG. 2b, when the system 200 is in use, a first terminal of external load R₂ 206 is connected to output 204. For the purpose of simplicity, it is assumed that the other terminal of external load R₂ 206 is connected to a fixed voltage 220. The resistance of external load R₂ 206 can be derived from the voltage V_ADC at output 204, from the voltage at 200 and by the current I_out flowing at output 204.

The input 202 allows receiving either a forcing voltage value or a forcing current value. In embodiments where the input 202 is for receiving a forcing voltage, the set of functional elements allows forcing the voltage V_ADC at output 204 to a required voltage value at the external load R₂ 206 using a search algorithm implemented by search unit 208. The search unit 208 controls the voltage at intermediate voltage point 214, which is input to the Digital-to-Analog-Converter (DAC) 210, such that the desired voltage V_ADC at output 204 is set as desired. In embodiments where the input 202 is for receiving a forcing current, the set of functional elements allows forcing the current I_out at output 204 to a required current value at the external load R₂ 206.

Depending on the forcing parameter (the voltage V_ADC or I_out at output 204), different searching algorithms are implemented by the search unit 208. The algorithms implemented by search unit 208 are described in greater detail herein below.

The resistance of the internal load R₁ 216 depends on both terminal voltage V_DAC at point 214 and voltage V_ADC at output 204. The Analog-to-Digital-Converter (ADC) 212 has an input coupled to output 204 and an output coupled to the search unit 208. The ADC 212 has essentially an infinite resistance in the operating range of the system 200. As such the ADC 212 does not draw current and consequently, the current at output 204 flowing in external load 206 also flows into internal load R₁ 216.

Internal Load R₁ 216

In a specific implementation, the internal load R₁ 216 is a non-linear resistor device whose resistance depends on both terminal voltages V_DAC and V_ADC. As a result, I_out is dependent on both V_DAC and V_ADC. A non-limiting example of the DC characteristics of internal load R₁ 216 is shown in FIG. 3 of the drawings.

Positive Resistance Property

From FIG. 3, it can be observed that for any particular voltage V_DAC at intermediate voltage point, current I_out at 204 increases when the voltage V_ADC at 204 increases. Mathematically, this can be expressed by the following: $\begin{array}{cc}\frac{\partial V_{\mathrm{ADC}}}{\partial I_{\mathrm{out}}}>0\mathrm{for}\mathrm{all}\mathrm{voltage}\mathrm{values}{V}_{\mathrm{DAC}}& \mathrm{Equation}1\end{array}$An internal load R₁ 216 is said to have a “positive resistance” if the condition in equation 1 is satisfied. Two non-limiting examples of implementations of such a load element are shown in FIGS. 4(a) and 4(b). FIG. 4(a) shows a linear resistor and FIG. 4(b) shows a CMOS inverter. For either one of these elements, the current I_out at 204 increases with voltage V_ADC if voltage V_DAC is fixed at any voltage. Hence the examples of load elements shown in FIGS. 4(a) and 4(b) follow the relationship in equation 1.

For the purpose of simplicity, only elements with a positive resistance will be used for internal load R₁ 216. Therefore, in all the analysis that follows, R₁ will be assumed to have a positive resistance. It will be readily apparent to the person skilled in the art in light of this description that an implementation where R₁ has a non-positive resistance can be implemented without detracting from the spirit of the invention.

Non-Inverting Property

If the voltages at the intermediate voltage point 214 correspond to V_DAC voltages in FIG. 3 as V_DAC3>V_DAC2>V_DAC1, then voltage V_ADC will increase with voltage V_DAC if I_out is held at any constant value. Mathematically, this can be expressed by the following: $\begin{array}{cc}\frac{\partial V_{\mathrm{ADC}}}{\partial V_{\mathrm{DAC}}}>0\mathrm{for}\mathrm{all}{I}_{\mathrm{out}}& \mathrm{Equation}2\end{array}$A load element R₁ 216 is said to be “non-inverting” if the condition in equation 2 is satisfied. An example of a non-inverting load R₁ 216 is shown in FIG. 4(a). It will be readily apparent to the person skilled in the art that if current I_out at output 204 is fixed at any value, there will be a constant potential difference between V_ADC and V_DAC. Therefore, V_ADC increases with V_DAC. The condition in equation 2 is satisfied.

Alternatively, load element R₁ 216 may be “inverting” if it satisfies the following condition. If the voltages shown in FIG. 3 at the intermediate voltage point 214 as V_DAC3<V_DAC2<V_DAC1, then V_ADC will increase while V_DAC decreases (and vice versa) if I_out is held constant. Mathematically, this can be expressed by the following: $\begin{array}{cc}\frac{\partial V_{\mathrm{ADC}}}{\partial V_{\mathrm{DAC}}}<0\mathrm{for}\mathrm{all}{I}_{\mathrm{out}}& \mathrm{Equation}3\end{array}$

A load element R₁ 216 is said to be “inverting” if the condition in equation 3 is satisfied. An example of an inverting R₁ is shown in FIG. 4(b). For a constant I_out, the V_ADC increases while V_DAC decreases, and vice versa.

Depending on whether an inverting or a non-inverting internal load R₁ 216 is implemented in the system 200 (shown in FIGS. 2a and 2b), slightly different search algorithms have to be implemented by the search unit 208. The search algorithms implemented by the search unit 208 will first be described with reference to an non-inverting internal load R₁ 216. The algorithm modifications for an inverting-type load R₁ 216 will be described later on in the specification.

External Load R₂ 206

With reference to FIGS. 2a and 2b of the drawings, the current I_out flowing from external load R₂ at output 204 is dependent on the voltage V_ADC at the output 204. In most microelectronic circuits, whenever a current is pumped into a node, the node voltage increases. In a specific example of implementation, the system 200 is designed with the assumption that current I_out decreases as V_ADC increases (i.e., V_ADC increases when a positive current is pumped, into R₂). Mathematically, this constraint can be expressed as $\begin{array}{cc}\frac{\partial I_{\mathrm{out}}}{\partial V_{\mathrm{ADC}}}<0& \mathrm{Equation}4\end{array}$An example DC characteristic of external load R₂ 206 that follows equation 4 is shown in FIG. 5. For the purpose of simplicity, R₂ is assumed to follow the property defined by equation 4. It will be readily apparent to the person skilled in the art in light of this description that an implementation where R₂ does not follow the property defined by equation 4 can be implemented without detracting from the spirit of the invention.Series-connected Loads

When the system 200 is in operation, the internal load R₁ 216 is connected in series with external load R₂ 206 as shown in FIGS. 2a and 2b. As a result, the load currents (I_out's) through the load R₁ 216 and load R₂ 206 are equal. The DC bias points can be found by overlapping the DC plots of FIG. 3 and FIG. 5 to form a load-line plot shown in FIG. 6.

It can be seen from FIG. 6 that for a voltage V_DAC at intermediate voltage point 214, there is only one valid V_ADC and I_out at output 204 that can satisfy the equilibrium condition. Thus V_ADC and I_out are functions of voltage V_DAC. Mathematically, this relationship can be expressed as two simultaneous equations:V_ADC=f₁(V_DAC)  Equation 5I_out=f₂(V_DAC)  Equation 6The relationships in equations 5 and 6 imply that it is possible to force either a voltage (V_ADC) or a current (I_out) at output 204 by establishing a corresponding voltage V_DAC at the intermediate voltage point 214.Force-Voltage-Measure-Current Algorithm

In a first example of implementation of the invention, the input 202 (shown in FIGS. 2a and 2b) is adapted for receiving a forcing parameter signal indicative of a forced voltage signal V_force. In such an implementation, the system is a voltage generating circuit and the search unit 208 is said to implement a force voltage algorithm.

The Search Algorithm

An objective of the force-voltage algorithm is to vary the voltage V_DAC at the intermediate voltage point 214 such that the voltage at the output (V_ADC) 204 will be set to approximate the desired forcing voltage V_force. We will refer to this desired V_DAC voltage as V*_DAC.

FIG. 7 shows a graphical view of the search process used to obtain V*_DAC. For a particular V_force applied at input 202, the operating current I_out at output 204 corresponds to the intersection of a vertical line representing V_ADC=V_force and the load line for R₂. This point of intersection is labelled as point A 71 in FIG. 7. The search algorithm adjusts V_DAC at the intermediate voltage point 214 such that the load line for load R₁ 216 intersects point A 71. The corresponding V_DAC is then V*_DAC.

FIG. 8 shows the circuit device of FIG. 2b adapted for searching for a voltage to apply to intermediate voltage point V_DAC for a desired forcing parameter voltage signal V_force in accordance with a non-limiting example of implementation of the invention.

As shown in FIG. 8, a forced voltage signal V_force is applied to input 202. The search unit 208 includes an ADC 80, a digital comparator 82, a digital integrator module 83 and an output for releasing a digital target voltage k_DAC 87. The ADC 80 processes the forced voltage signal V_force to generated a digital forced voltage signal k_FORCE 84. The digital comparator 82 is for generating a digital difference voltage signal dependent on the difference between the digital forced voltage signal 84 and the digital approximation 85 of the voltage signal at the output 204 of the circuit device. The digital integrator module 83 is adapted for processing the digital difference voltage signal to derive the digital target voltage k_DAC 87. The digital target voltage k_DAC 87 is released at the output of the search unit 208 for processing by the digital-to-analog converter module 210. Where the forcing parameter signal is a digital signal, the ADC 80 may be omitted from the system shown FIG. 8.

Mathematically, k_force 84 can be expressed as follows: $\begin{array}{cc}{k}_{\mathrm{force}}=\frac{Q\left[V_{\mathrm{force}}\right]}{V_{\mathrm{LSB}-\mathrm{ADC}}}=\frac{{\hat{V}}_{\mathrm{force}}}{V_{\mathrm{LSB}-\mathrm{ADC}}}& \mathrm{Equation}7\end{array}$where Q[x] is the quantizer function of the ADC 80, V_LSB-ADC is the LSB voltage of the ADC 80 and {circumflex over (V)}_force is the quantized V_force aplied at 202. Also, the digitized value of V_DAC at intermediate point 214 is denoted by k_DAC at point 87 in FIG. 8.

It will be readily apparent that for any voltage x and its quantized value Q[x], the difference is no bigger than half the LSB voltage of the ADC 80 (V_LSB-ADC). This voltage difference is negligible when ADC 80 has a sufficiently small quantization step. Therefore, for the purpose of simplicity of the description, the following assumption have been made:x=Q[x]=k×V_LSB-ADC  Equation 8where k is the digital representation of voltage x at the output of the ADC. From equation 8, it can be seen that:V_force={circumflex over (V)}_force  Equation 9Therefore, equation 7 becomes: $\begin{array}{cc}{k}_{\mathrm{force}}=\frac{V_{\mathrm{force}}}{V_{\mathrm{LSB}-\mathrm{ADC}}}& \mathrm{Equation}10\end{array}$Similarly, for the ADC 212 at the feedback path of FIG. 8, $\begin{array}{cc}{k}_{\mathrm{track}}=\frac{V_{\mathrm{ADC}}}{V_{\mathrm{LSB}-\mathrm{ADC}}}& \mathrm{Equation}11\end{array}$where V_ADC is the voltage at output 204 and k_track is the digital approximation of the voltage signal V_ADC at node 85 in FIG. 8.

Upon equilibrium in the system, V_ADC=V_force and V_DAC=V*_DAC.

Current Measurement

Once the voltage V_force has been applied, the current I_out at output 204 needs to be obtained in order to obtain the DC characteristics of external load R₂ 206.

With reference to FIG. 9, the load lines of internal load R₁ 216 are shown with reference to current I_out at output 204. It can be seen from this figure that I_out is dependent on both the voltage V_DAC at the intermediate voltage point 214 and on the voltage V_ADC at the output 204. Mathematically, this can be expressed as follows:I_out=G(V_DAC, V_ADC)  Equation 12If the voltage V_ADC at output 204 is kept constant at a voltage which approximates V_force, current I_out at output 204 becomes a one variable function of the voltage V_DAC at the intermediate voltage point 214 as follows:I_out=G(V_DAC, V_force)  Equation 13This is illustrated in FIG. 9, where the vertical line representing V_ADC=V_force have been superimposed on the load lines. As shown in FIGS. 9 and 10, the intersection points uniquely relate the current I_out at output 204 and the voltage V_DAC at the intermediate voltage point 214.

With reference to FIG. 8, by forcing a voltage V_ADC at output 204 approximating a particular V_force applied at input 202, the voltage V_DAC at the intermediate voltage point 214 will settle at a corresponding V*_DAC. Suppose the relationship G in equation 13 is known, then the current I_out at output 204 can be deduced by simply substituting V_DAC=V*_DAC into equation 13 to obtain:I_out=G(V*_DAC, V_force)  Equation 14Using the relationship in equation 14, a current measurement mechanism can be derived as follows. First, it can be seen in FIG. 8 that the value of the voltage V_DAC at the intermediate voltage point 214 is reconstructed from the digital value k_DAC at point 87 via a DAC 210. If we denote the LSB voltage of the DAC 210 by V_LSB-DAC, then mathematically V_DAC can be expressed as follows:V_DAC=k_DAC×V_LSB-DAC  Equation 15

When the circuit in FIG. 8 is in equilibrium, V_DAC=V*_DAC. According to equation 15, the value of k_DAC will be given by: $\begin{array}{cc}{K}_{\mathrm{DAC}}=k_{\mathrm{DAC}}^{*}=\frac{V_{\mathrm{DAC}}^{*}}{V_{\mathrm{LSB}-\mathrm{DAC}}}& \mathrm{Equation}16\end{array}$

By substituting equations 7 and 16 into equation 14, the value of the current I_out at output 204 can be deduced as follows:I_out=G(V*_DAC, V_force)=G(k*_DAC×V_LSB-DAC,k_force×V_LSB-ADC)  Equation 17As V_LSB-DAC and V_LSB-ADC in equation 17 are constant scale factors, equation 17 can be simplified as follows: I_out=G(k*_DAC×V_LSB-DAC,k_force×V_LSB-ADC)=Ĝ(k*_DAC,k_force)  Equation 18The relationship described in equation 18 can be implemented by the circuit shown in FIG. 11.

The system depicted in FIG. 11 includes the same system as that is depicted in FIG. 8 in addition to a current measuring circuit. The current measurement circuit includes search logic 90. Search logic 90 receives a signal conveying a forced voltage signal and a signal conveying the target voltage signal at intermediate point 214 (or alternatively at point 87) and processes these signals to derive a measurement of the current flowing at the output 204 of the voltage generating circuit. In the specific example depicted in FIG. 11, the current measurement circuit includes a first input for receiving the digital representation of the second voltage signal k_DAC 87, a second input for receiving a digital approximation of the forced voltage signal k_FORCE 84, search logic 90 and an output 92. The search logic 90 derives a certain current measurement on the basis of the digital representation of the second voltage signal k_DAC 87 and the digital approximation of the forced voltage signal k_FORCE 84. The certain current measurement is released at the output 92 of the current measurement circuit. When the system is in equilibrium, the certain current measurement is indicative of an approximation of measurement of the current flowing at the output of the voltage generating circuit when the forced voltage is applied to the output 204. It will be appreciated that although the search logic 90 has been described as receiving signals in digital format, search logic 90 may be implemented such that it may receive signals in analog format without detracting from the spirit of the invention.

In accordance with a specific implementation, the search logic of the current measurement circuit includes a data structure in the form of a lookup table 90 containing Ĝ. The data structure may be stored on any suitable memory unit such as a RAM, ROM, PROM, EPROM and EEPROM. In a specific non-limiting implementation, the data structure is stored on a RAM device. The values k*_DAC 87 and k_force 84 are fed into lookup table 90 to derive I_out. When the system is in equilibrium, the value of I_out can be computed using equation 18.

Calibration Techniques

In a specific implementation, to perform a force-voltage current measurement, a lookup table containing Ĝ 90 is used. In this section, two different examples of methods of the calibration procedure for generating the required lookup-table entries for Ĝ are described. It will be readily apparent that methods other than those described herein below may be used for generating the lookup table containing Ĝ 90 without detracting from the spirit of the invention.

Method 1

In a first specific example, if k_ADC is forced to a constant value approximating k_force, equation 18 can be written as follows:I_out=Ĝ(k*_DAC,k_ADC)|_kADC=kforce  Equation 19Equation 19 indicate that I_out would become a one-variable function of k*_ADC. Alternatively, we can write the inverse relationship of equation 19 as follows:k*_DAC={circumflex over (=)}Ĝ⁻¹(I_out,k_ADC)|_kADC=kforce  Equation 20where k*_DAC is a function of I_out, and Ĝ⁻¹ is the inverse function of G. The relationship in equation 20 can be obtained by the circuit shown in FIG. 12. In the circuit, voltage V_ADC at output 204 is held constant at the desired forcing voltage dictated by k_force 84. The inverse function Ĝ⁻¹ is found by sweeping an external current reference I_sweep 94 over a range of current values, followed by recording k*_DAC 87 at each current step. The values of k*_DAC may be recorded in any suitable readable memory device. In a non-limiting implementation, the values of k*_DAC 87 are recorded on a RAM. In other words, for a given combination of k_force 84 and I_sweep 94, a corresponding value of k*_DAC is recorded.

Once the inverse mapping Ĝ⁻¹ is known, the input and output variables can be interchanged to obtain the function in equation 19. The resultant mapping Ĝ can then be used in the current measurement circuit in FIG. 11. Note that the lookup table Ĝ 90 needs to be re-calibrated using this procedure should a different forcing voltage V_force/k_force be needed.

Method 2

A drawback of a force-voltage current measurement with calibration method 1 described above is that a large number of calibration points is required before an actual measurement can be done. For example, an n-bit current measurement will require a calibration of 2ⁿ points for the function Ĝ. If the number of actual current measurements is small, a significant amount of test time will be wasted to generate the unused entries of the lookup table Ĝ 90.

This section describes a second specific example of a calibration scheme that can avoid unnecessary calibrations. To achieve this, the current measurement system is calibrated after an actual measurement is made. The details of this algorithm are described below.

A plot of Ĝ⁻¹ defined in equation 20 is shown in FIG. 13. It can be seen that for the same forcing voltage given by a constant k_ADC=k_force 84, voltage k*_DAC 87 is a function of I_out at the output 204. Suppose I_out=I_x when the voltage at the external load R₂ 206 in FIG. 8 is forced to approximate V_force. According to FIG. 13, the corresponding k*_DAC will be given by k*_DACx. Mathematically, this can be expressed as follows:

 k*_DACx=Ĝ⁻¹(I_x,k_ADC)|_kADC=kforce  Equation 21

Now, suppose the output node 204 is connected to a current source at a value I_test 96 as shown in FIG. 14. It can be seen from FIG. 13 that when I_out is not equal to I_x (e.g., when I_out=I_test1 or I_out=I_test2), the corresponding k*_DAC will not equal k*_DACx. The principles of this calibration algorithm is to use the setup in FIG. 8 to first measure k*_DACx when I_out is equal to the unknown current I_x. Then, a current reference is applied to the system in a manner shown in FIG. 14. A search can be conducted to find the required I_test value that generates k*_DAC=k*_DACx. The resultant I_test value from the search will be the value of the unknown current I_x.

It can be seen from FIG. 13 that when I_test is set to I_test1<I_x, the corresponding k*_DAC1 is bigger than k_DACx. From this, we can conclude that:k*_DAC>k*_DACx when I_test<I_x  Equation 22

Similarly, by observing the intersection point at I_out=I_test2>I_x, it can be said thatk*_DAC<k*_DACx when I_test>I_x  Equation 23

The relationships in equations 22 and 23 provide the basis for the calibration search algorithm. In each iteration, I_test is set to a value and the corresponding k*_DAC will be compared with k*_DACx, the comparison result can then be used to increment/decrement I_test in the next iteration. The detail of this algorithm is summarised in Table 1.

TABLE 1
Calibration Search Algorithm
		Implied	Adjustment of Itest in the
	Comparison Result	Observation	next iteration
	k*DAC > k*DACx	Itest < Ix	increase Itest;
			otherwise decrease Itest

The calibration algorithm described in Table 1 can be implemented by the calibration circuit 102 shown in FIG. 15. The calibration circuit 102 includes a current DAC (IDAC) 104, a digital comparator module 108 and a digital integrator 106. In the figure, IDAC 104 is a current DAC whose output current I_test can be controlled digitally. Upon equilibrium, k*_DAC (87)=k*_DACx (110), the digital value of I_test 112 would be equal to the digitised value of the unknown current I_x.

The advantage of this calibration algorithm is that for an n-bit current measurement, a calibration of the 2n points for function Ĝ is not required. For example, when the search algorithm in Table 1 is implemented using the step search circuit in FIG. 15, on average only 2ⁿ⁻¹ calibration points are required. If the search algorithm in Table 1 is implemented by a binary search algorithm, only n calibration points are required for each measurement.

Force-Current-Measure-Voltage Algorithms

In a second example of implementation of the invention, the input 202 (shown in FIGS. 2a and 2b) is adapted for receiving a forcing parameter signal indicative of a forced current signal I_force.

The Search Algorithms

The objective of the force-current algorithms is to control the voltage V_DAC at intermediate voltage point 214 or voltage V_ADC at output 204 using the system in FIGS. 2a and 2b such that the current I_out at output 204 will be set to approximate the desired forcing current I_force applied at input 202.

Search Variables

From equation 6, it can be seen that by varying voltage V_DAC intermediate voltage point 214, the value of I_out can be set to approximate a desired value, I_force. A search of the voltage V_DAC can be implemented using the system shown in FIG. 16. The desired V_DAC voltage is defined as V*_DAC. Mathematically, this can be expressed as follows:I_force=f₂(V*_DAC)  Equation 24

FIG. 17 depicts a graphical view of the search process used to obtain V*_DAC. For a particular I_force applied at input 202, the operating V_ADC (labelled V*_ADC) corresponds to the intersection of a horizontal line representing I_out≈I_force and the load line for R₂ labelled point B. The search algorithm adjusts the voltage V_DAC such that the load line for internal load R₁ 216 intersects point B. The corresponding V_DAC is then V*_DAC.

V_ADC may also be viewed as a search variable. According to equations 5 and 6, we can write:I_out=f₂(f₁⁻¹(V_ADC))=f₃(V_ADC)  Equation 25Equation 25 indicates that if voltage V_ADC an output 204 can be controlled, I_out at output 204 can be set to a desired value by searching for the corresponding V_ADC. A search of voltage V_ADC at output 204 can be implemented using the system shown in FIG. 18. FIG. 18 shows the same system depicted in FIG. 2b but with the search unit 208 designed such that voltage V_ADC at output 204 is set via a feedback network. In this non-limiting implementation, the search unit 208 includes a “V_ADC Search Logic” circuit 304, a digital comparator module 302 and a digital integrator module 300. The “V_ADC Search Logic” circuit 304 implements a search algorithm where voltage V_ADC is a search variable. When the I_out at output 204 is equal to I_force applied at input 202, the desired voltage V_ADC value is given by V*_ADC in:I_force=f₃(V*_ADC)  Equation 26

FIG. 19 provides a graphical view of the search process used to obtain V*_ADC. For a particular current I_force applied at input 202, the operating V_DAC (labelled V*_DAC) corresponds to the intersection of a horizontal line representing I_out≈I_force and the load line for R₂, labelled point C in the graph of FIG. 19. The search algorithm adjusts V_ADC such that the vertical line representing V_ADC intersects point C. The corresponding V_ADC is then V*_ADC.

Convergence Conditions

From equations 24 and 26, it can be seen that when I_out=I_force, the following two conditions will be satisfied:V_DAC=V*_DAC  Equation 27V_ADC=V*_ADC  Equation 28Therefore, a search algorithm can determine if I_out=Iforce by observing either V_DAC or V_ADC. This means that either equation 27 or 28 can be used as a convergence condition of the search.The Four Search Algorithms

From the above discussions, it has been shown that a force-current search algorithm can be implemented by using either V_DAC or V_ADC as the search variable. Also, the convergence condition can be determined by observing either V_DAC or V_ADC. Therefore, there are at least four possible variations for the force-current algorithm, as shown in Table 2.

TABLE 2
Four Force-Current Search Algorithms
	Search	Convergence
Algorithm	Variable	Condition	Algorithm Name
1	VDAC	VADC = V*ADC	vary-VDAC-compare-VADC
2	VADC	VADC = V*ADC	vary-VADC-compare-VADC
3	VADC	VDAC = V*DAC	vary-VADC-compare-VDAC
4	VDAC	VDAC = V*DAC	vary-VDAC-compare-VDAC

The descriptions for four force-current algorithms will be presented in the following sections.

Voltage Measurement

For the force-current-measure-voltage operation, the voltage V_ADC at output 204 must be measured after the force-current algorithm is applied. From the general architecture of the System in FIG. 2b, it can be seen that the value of the voltage V_ADC at output 204 can be readily measured by reading the value k_ADC from the ADC 212 at the end of the force-current searching process.

Convergence Criteria

In the search control logic implemented by search unit 208 of a force-current system, the convergence conditions are observed by evaluating a function (or its inverse) that relates the quantities k_DAC, k_ADC and I_out. That function will be described in this section.

Consider the load lines of internal load R₁ 216 in FIG. 20. It can be seen from this figure that voltage V_ADC at output 204 is dependent on both the voltage V_DAC at the intermediate voltage point 214 and current I_out at output 202. Mathematically, this can be expressed as follows:V_ADC=H(V_DAC, I_out)  Equation 29If current I_out at output 204 is kept constant at a value that approximates I_force, voltage V_ADC at output 204 becomes a one variable function of V_DAC as follows:V_ADC=H(V_DAC, I_force)  Equation 30This is illustrated in FIG. 20, where the horizontal line representing I_out=I_force is shown in the load lines for internal load R₁ 216. As can be seen, the intersection points uniquely relates voltage V_ADC and voltage V_DAC in a manner such as that shown in FIG. 21.

If we denote the LSB voltage of the DAC 210 used in the system 200 by V_LSB-DAC, then V_DAC can be represented by a digital number k_DAC defined in equation 15. Similarly, if the LSB voltage of the ADC 212 is defined as V_LSB-ADC, V_ADC can be represented by a digital value k_ADC as follows:V_ADC=k_ADC×V_LSB-DAC  Equation 31Substituting equation 15 and 31 into equation 30, the following relationship is obtained:k_ADC×V_LSB-ADC=H(k_DAC×V_LSB-DAC, I_force)  Equation 32As V_LSB-DAC and V_LSB-ADC in equation 32 are constant scale factors, equation 32 can be simplified as follows: $\begin{array}{cc}{k}_{\mathrm{ADC}}=\frac{H\left(k_{\mathrm{DAC}}\times V_{\mathrm{LSB}-\mathrm{DAC}},I_{\mathrm{force}}\right)}{V_{\mathrm{LSB}-\mathrm{ADC}}}=\hat{H}\left(k_{\mathrm{DAC},}{I}_{\mathrm{force}}\right)& \mathrm{Equation}33\end{array}$On the other hand, it can be seen from equation 30 that if current I_out is kept constant at I_force, voltage V_ADC is a one variable function of V_DAC. From this, the reverse relationship can be expressed as:V_DAC=H⁻¹(V_ADC, I_out)|_Iout=Iforce  Equation 34

The relationship in equation 34 is illustrated in FIG. 22, where the horizontal line representing I_out=I_force is superimposed on top of vertical lines representing different values of V_ADC. The intersection points are circled in the figure. The intersection point at V_ADC=V_ADC1, labelled point D, will be taken as an example. For V_ADC=V_ADC1, the corresponding value of V_DAC according to equation 34 is given by the load line of internal load R₁ 216 that intersects with Point D. Therefore the value of V_DAC at point D is given by the following:V_DAC=H⁻¹(V_ADC1, I_out)|_Iout=Iforce  Equation 35

If k_DAC and k_ADC in FIG. 18 are defined by equation 15 and 31, equation 34 becomes:k_DAC×V_LSB-DAC=H⁻¹(k_ADC×V_LSB-ADC, I_out)|_Iout=Iforce  Equation 36

As V_LSB-DAC and V_LSB-ADC in equation 36 are constant scale factors, equation 36 can be simplified as follows: $\begin{array}{cc}\begin{array}{c}{k}_{\mathrm{DAC}}=\frac{{H^{-1}\left(k_{\mathrm{ADC}}\times V_{\mathrm{LSB}-\mathrm{ADC}},I_{\mathrm{out}}\right)}_{I_{\mathrm{out}}=I_{\mathrm{force}}}}{V_{\mathrm{LSB}-\mathrm{DAC}}}\\ {={\hat{H}}^{-1}\left(k_{\mathrm{ADC}},I_{\mathrm{force}}\right)}_{I_{\mathrm{out}}=I_{\mathrm{force}}}\end{array}& \mathrm{Equation}37\end{array}$

The functions Ĥ and Ĥ⁻¹ defined in equations 33 and 37 are used in the four force-current search algorithms. A summary of the algorithms is presented in Table 3.

TABLE 3
Force current search algorithms
			Adjustment
Algorithm		Implied	in the
Name	Comparison Result	Observation	next iteration
Vary-VDAC-	kADC > Ĥ(kDAC, Iforce)	kDAC < k*DAC	increase kDAC,
Compare-			otherwise
VADC			decrease
			kDAC
Vary-VADC-	kADC > Ĥ(kDAC, Iforce)	kADC < k*ADC	increase kADC,
Compare-			otherwise
VADC			decrease
			kADC
Vary-VADC-	kDAC < Ĥ−1	kADC < k*ADC	increase kADC,
Compare-	(kADC, Iout)|Iout = Iforce		otherwise
VDAC			decrease
			kADC
Vary-VDAC-	kDAC < Ĥ−1	kDAC < k*DAC	increase kADC,
Compare-	(kADC, Iout)|Iout = Iforce		otherwise
VDAC			decrease
			kDAC

Table 4 is an index to the figures corresponding to the Force-Current Search Algorithms described in the specification. It will be appreciated that binary searches can also be performed by replacing the integrators 352362372376 in FIG. 23(a) to (d) with Successive Approximation Registers (SARs) without detracting from the spirit of the invention.

TABLE 4
Force-Current Search Algorithms - Diagrams and Descriptions
Algorithm Name
(Algorithm Number)               Circuit Diagram(s)
vary-VDAC-compare-VADC (Algorithm 1) FIG. 23 (a)
vary-VADC-compare-VADC (Algorithm 2) FIG. 23 (b)
vary-VADC-compare-VDAC (Algorithm 3) FIG. 23 (c)
vary-VDAC-compare-VDAC (Algorithm 4) FIG. 23 (d),
                                 FIG. 26 (Simplified)

The Four Search Implementations

In this section, details of the four search algorithms will be described. Note that because of the similarities in the four algorithms, the reader should get the basic idea from any one of the four algorithm descriptions and may not need to read the other descriptions. In addition, the algorithms are described for signals represented in digital format. It will be appreciated that corresponding algorithms for signals represented in the analog domain may be used without detracting from the spirit of the invention. Such corresponding algorithms and will become apparent to a person skilled in the art in light of the present specification and as such will not be described further here.

Vary−V_DAC−Compare−V_ADC (Algorithm 1)

The plot in FIG. 6 can be superimposed with that in FIG. 20 to form a load line plot in FIG. 24. It can be seen that out of all the three V_DAC voltages, V_DAC2 will yield I_out=I_force, i.e., V*_DAC=V_DAC2. Also, at V_DAC2, the corresponding V_ADC voltage (V_ADC2) follows the following relationship:V_ADC2=H(V_DAC2, I_force)  Equation 38Therefore, it follows that: V_ADC=H(V_DAC, I_force) when V_DAC=V*_DAC  Equation 39Moreover, it can be seen that for V_DAC1<V_DAC2, the corresponding V_ADC1 is given by:V_ADC1>H(V_DAC1, I_force)  Equation 40Therefore,V_ADC>H(V_DAC, I_force) when V_DAC<V*_DAC  Equation 41Similarly, by observing V_DAC3, it can be shown that:V_ADC<H(V_DAC, I_force) when V_DAC>V*_DAC  Equation 42

The relationships in equations 39, 41 and 42 provides the search algorithm required to force I_out=I_force. In each iteration, voltage V_DAC is set to a value and the corresponding V_ADC will be compared with H(V_DAC, I_force). The comparison result can then be used to increment/decrement V_DAC in the next iteration. Using the definitions of k_DAC, k_ADC and Ĥ in equations 15, 31 and 33, the conditions in equations 41 and 42 can be summarised into the force current algorithm in Table 3. The corresponding circuit implementation is shown in FIG. 23a). As shown, the search unit 208 implemented in accordance with this first algorithm includes a lookup table 354 containing the function Ĥ, a digital comparator 350 and a digital integrator 352.

Vary−V_ADC−Compare−V_ADC (Algorithm 2)

An alternative force-current approach can be derived by interpreting the load line plot in FIG. 24 in another manner. Let voltage V_ADC be the search variable and the desired V_ADC that yield I_out=I_force be V*_ADC. It can be seen from FIG. 24 that out of all the three V_ADC voltages, V_ADC2 will yield I_out=I_force, i.e., V*_ADC=V_ADC2. Also, at V_ADC2 the corresponding V_DAC voltage (V_DAC2) follows the following relationship:V_ADC2=H(V_DAC2, I_force)  Equation 43Therefore, it follows that:V_ADC=H(V_DAC, I_force) when V_ADC=V*_ADC  Equation 44Moreover, by observing the intersection points at V_ADC=V_ADC1 and V_ADC=V_ADC3, it can be shown that:V_ADC>H(V_DAC, I_force) when V_ADC<V*_ADC  Equation 45V_ADC<H(V_DAC, I_force) when V_ADC>V*_ADC  Equation 46

Using the definitions of k_DAC, k_ADC and Ĥ in equations 15, 31 and 33, the conditions in equations 45 and 46 can be summarised into the force current algorithm in Table 3. The corresponding circuit implementation is shown in FIG. 23(b). As shown, the search unit 208 implemented in accordance with this second algorithm includes a lookup table 354 containing the function Ĥ, a first digital comparator 356, a first digital integrator 358, a second digital comparator 360 and a second digital integrator 362.

Vary−V_ADC−Compare−V_DAC (Algorithm 3)

The graph in FIG. 6 can be superimposed with that in FIG. 22 to form a load line plot in FIG. 25. It can be seen that out of all the three V_ADC voltages, V_ADC2 will yield I_out=I_force, i.e., V*_ADC=V_ADC2. Also, at V_ADC2 the corresponding V_DAC voltage (V_DAC2) follows the following relationship:

 V_DAC2=H⁻¹(V_ADC2, I_force)|_Iout=Iforce  Equation 47

Therefore, it follows that:V_DAC=H⁻¹(V_ADC, I_force)|_Iout=Iforce when V_ADC=V*_ADC  Equation 48Moreover, it can be seen that for V_ADC1<V_ADC2, the corresponding V_DAC1 is given by:V_DAC1<H⁻¹(V_ADC1, I_force)|_Iout=Iforce  Equation 49Therefore,V_DAC<H⁻¹(V_ADC, I_force)|_Iout=Iforce when V_ADC<V*_ADC  Equation 50Similarly, by observing V_ADC3, it can be shown that:V_DAC>H⁻¹(V_ADC, I_force)|_Iout=Iforce when V_ADC>V*_ADC  Equation 51

The relationships in equations 48, 50 and 51 provides the search algorithm required to force I_out=I_force. In each iteration, V_ADC is set to a value and the corresponding V_DAC will be compared with H⁻¹(V_ADC, I_force)|_Iout=Iforce, the comparison result can then be used to increment/decrement V_ADC in the next iteration. Using the definitions of k_DAC, k_ADC and H⁻¹ in equations 15, 31 and 37, the conditions in equations 50 and 51 can be summarised into the force current algorithm in Table 3. The corresponding circuit implementation is shown in FIG. 23c. As shown, the search unit 208 implemented in accordance with this third algorithm includes lookup table 364 containing the function Ĥ⁻¹, a first digital comparator 366, a first digital integrator 368, a second digital comparator 370 and a second digital integrator 372.

Vary−V_DAC−Compare−V_DAC(Algorithm 4)

An alternative force-current approach can be derived by interpreting the load line plot in FIG. 25 in another manner. Let V_DAC be the search variable and the desired V_DAC that yield I_out=I_force be V*_DAC. It can be seen from FIG. 25 that out of all the three V_DAC voltages, V_DAC2 will yield I_out=I_force, i.e., V*_DAC=V_DAC2. Also, at V_DAC2, the corresponding V_ADC voltage (V_ADC2) follows the following relationship:V_DAC2=H⁻¹(V_ADC2, I_force)|_Iout=Iforce  Equation 52Therefore, it follows that:V_DAC=H⁻¹(V_ADC, I_force)|_Iout=Iforce when V_DAC=V*_DAC  Equation 53Moreover, by observing the intersection points at V_DAC=V_DAC1 and V_DAC=V_DAC3, it can be shown that:V_DAC<H⁻¹(V_ADC, I_force)|_Iout=Iforce when V_DAC<V*_DAC  Equation 54V_DAC>H⁻¹(V_ADC, I_force)|_Iout=Iforce when V_DAC>V*_DAC  Equation 55

Using the definitions of k_DAC, k_ADC and Ĥ⁻¹ in equations 15, 31 and 37, the conditions in equations 54 and 55 can be summarized into the force current algorithm in Table 3. The algorithm described in Table 3 can be implemented by the circuit shown in FIG. 23(d). As shown, the search unit 208 implemented in accordance with this fourth algorithm includes lookup table 364 containing the function Ĥ⁻¹, a digital comparator 374 and a digital integrator 376. It can be observed that the circuit in FIG. 23(d) including 374, 376 and the feedback line from 376 to 374 is a unity gain buffer. As such, the circuit can be simplified to the implementation shown in FIG. 26.

In the circuit in FIG. 26 the search unit 208 is implemented by a memory module 364, such as a RAM. In a non-limiting implementation, the search unit 208 is implemented by a RAM without any addition digital circuitry for performing the searching functionality.

Calibration Techniques

In the first and second force-current algorithms described above, a lookup table 354 containing the function Ĥ is used in the circuit implementation as shown in FIGS. 23a) and 23b). Also in the third and fourth force-current algorithms, a lookup table 364 containing the function Ĥ⁻¹ is used in the circuit implementation as shown in FIGS. 23c) and 23d). The calibration procedures for these functions will be described herein below.

The function Ĥ defined in equation 34 can be found using the circuit shown in FIG. 27. In the circuit, I_out at output 204 is held constant at the desired forcing current I_force by current source 380. The function Ĥ is found by sweeping k_DAC over the full-scale range of the DAC 210, followed by recording k_ADC 85 at each step.

On the other hand, the function Ĥ⁻¹, defined in equation 37, can be found using the circuit shown in FIG. 28. In this circuit, I_out at output 204 is held constant at the desired forcing current I_force by current source 380. The function Ĥ⁻¹ is found by sweeping k_ADC over the full-scale range of the ADC 212, followed by recording k_DAC 87 at each step.

Note that, alternatively, the function Ĥ can be obtained by first finding the functional relationship. Ĥ⁻¹ followed by a switch of input and output variables. Similarly, the function Ĥ⁻¹ can be obtained by switching the input and output variable after finding Ĥ.

The lookup table Ĥ (or Ĥ⁻¹, whichever is employed in the force-current search algorithm) needs to be re-calibrated should a different forcing current I_force be needed.

Algorithm Modifications for an Inverting Load

The section below describes the circuit modifications when the internal load R₁ 216 used in the system 200 is an inverting load.

Force-Voltage-Measure-Current Algorithms

If the internal load R₁ 216 of the system 200 follows the inverting property defined in equation 3, the feedback loop in FIG. 8 needs to maintain negative feedback by switching the polarity of the comparator 82. The modified circuit is shown in FIG. 29 where the comparator 82 of FIG. 8 has been replaced by comparator 400 in FIG. 29. Note that the same change also applies to the measurement and calibration circuits in FIGS. 11, 12 and 14.

The current measurement and the first calibration method described above for a non-inverting internal load R₁ 216 do not need to be modified for an inverting R₁ because V*_DAC (as well as k*_DAC) is still a one-to-one corresponding function of I_out for any particular V_ADC=V_force. However, the second calibration method described needs to be modified for an inverting R₁. A modified calibration search algorithm is shown in Table 5. The modified calibration search circuit in FIG. 15 is shown in FIG. 30 where the comparators 82 and 108 of FIG. 15 have been replaced by comparators 400 and 402 respectively in FIG. 30.

TABLE 5
Modified Calibration Search Algorithm for an Inverting R1
Comparison		Adjustment of Itest in the next
Result	Implied Observation	iteration
k*DAC < k*DACx	Itest < Ix	increase Itest, otherwise decrease
		Itest

Force-Current-Measure-Voltage Algorithms

When the internal load R₁ 216 of the system 200 follows the inverting property as defined in equation 3, the feedback loop in the general force-V_ADC architecture in FIG. 18 needs to maintain negative feedback by switching the polarity of the comparator. The modified circuit is shown in FIG. 31 where the comparator 302 of FIG. 18 has been replaced by comparator 600 in FIG. 31. Similarly the V_ADC search logic module 304 in FIG. 18 is replaced by modified V_ADC search logic module 604 for certain algorithms (as in FIG. 32).

The modifications for each force-current algorithm is described herein below.

Vary−V_DAC−Compare−V_ADC (Algorithm 1)

When an inverting internal load R₁ 216 is used, this search algorithm is modified in order to maintain convergence of the target current value I_force. The modified search algorithm and its implementation are shown in Table 6 and FIG. 32. As shown, the search unit 208 implemented in accordance with this first algorithm when an inverting internal load R₁ 216 is used includes a lookup table 364 containing the function Ĥ⁻¹, a digital comparator 600 and a digital integrator 602.

TABLE 6
Force current algorithm for an inverting R1 (vary-VDAC-
compare-VADC)
	Implied	Adjustment of kDAC in the
kADC Comparison Result	Observation	next iteration
kADC > Ĥ(kDAC, Iforce)	kDAC > k*DAC	Decrease kDAC, otherwise
		increase kDAC

Vary−V_ADC−Compare−V_ADC (Algorithm 2)

There is no change in this search algorithm when an inverting internal load R₁ 216 is used. However, because this algorithm has a force-V_ADC circuit implementation, the modification in FIG. 31 is required. The modified circuit is shown in FIG. 33. As shown, the search unit 208 implemented in accordance with this second algorithm includes a lookup table 614 containing the function Ĥ, a first digital comparator 612, a first digital integrator 610, a second digital comparator 608 and a second digital integrator 606.

Vary−V_ADC−Compare−V_DAC (Algorithm 3)

When an inverting internal load R₁ 216 is used, this search algorithm has to change in order to maintain convergence of the target current value. The modified search algorithm is shown in Table 7. Also, because this algorithm has a force-V_ADC circuit implementation, the modification in FIG. 31 is required. The modified circuit is shown in FIG. 34. As shown, the search unit 208 implemented in accordance with this third algorithm includes a lookup table 616 containing the function Ĥ, a first digital comparator 612, a first digital integrator 610, a second digital comparator 608 and a second digital integrator 606.

TABLE 7
Force current algorithm for an inverting R1 (vary-VADC-
compare-VDAC)
			Adjustment
		Implied	of kDAC in the
	kDAC Comparison Result	Observation	next iteration
	kDAC < Ĥ−1 (kADC,	kADC > k*ADC	decrease kADC,
	Iout)|Iout = Iforce		otherwise
			increase kADC

Vary−V_DAC−Compare−V_DAC (Algorithm 4)

There is no change in this algorithm or its implementation if an inverting R₁ is used.

Special Case: Using an Internal Linear Resistive Load R₁ 216

This section deals with the special case where the internal load R₁ 216 used in the system 200 is a linear resistor. The general architectures of the system that forces V_DAC and V_ADC are shown in FIG. 35 and FIG. 36. In both figures, the internal load R₁ 216 is a linear resistor and is denoted by R_1L. In FIG. 36, the search unit 208 includes search logic 704, a digital comparator 702 and a digital integrator 701. Upon equilibrium, the current I_out at output 204 in both circuits is given by: $\begin{array}{cc}{I}_{\mathrm{out}}=\frac{k_{\mathrm{ADC}}\times V_{\mathrm{LSB}-\mathrm{ADC}}-k_{\mathrm{DAC}}\times V_{\mathrm{LSB}-\mathrm{DAC}}}{R_{1L}}+I_{\mathrm{offset}}& \mathrm{Equation}56\end{array}$where I_offset is a current offset term resulted from the offset voltages of the DAC 210 and the ADC 212. A calibration process is required to determine the values of the constants R_1L and I_offset. After these values are found, the lookup tables can be readily constructed for the force-voltage-measure-current or the force-current-measure-current algorithms.

From equation 56, it can be seen that the quantities I_out, K_ADC and k_DAC are linearly related. As we will see in the following sections, this linearity property will dramatically deduce the amount calibration time required.

Current Measurement

Consider the force-voltage-measure-current circuit in FIG. 37. It is the same as the circuit in FIG. 11 except that internal load R₁ 216 is a linear resistor denoted by R_1L. For this circuit, we can write Ĝ defined in equation 18 as: $\begin{array}{cc}\begin{array}{c}{I}_{\mathrm{out}}=\hat{G}\left(k_{\mathrm{DAC}}^{*},V_{\mathrm{force}}\right)\\ =\frac{k_{\mathrm{ADC}}\times V_{\mathrm{LSB}-\mathrm{ADC}}-k_{\mathrm{DAC}}^{*}\times V_{\mathrm{LSB}-\mathrm{DAC}}}{R_{1L}}+I_{\mathrm{offset}}\end{array}& \mathrm{Equation}57\end{array}$where the values of R_1L and I_offset can be found using the calibration method described further on in the specification.

The relationship in equation 57 indicates that the lookup table Ĝ 706 in the current measurement system (shown in FIG. 37) can be generated by two calibration points. Moreover, from equation 57, it can also be seen that the lookup table Ĝ 706 need not be re-calibrated if a different forcing voltage is required. This is different from the lookup table Ĝ 90 shown in FIG. 11 for the generalized current measurement system, where the function Ĝ needs to be re-calibrated whenever a different forcing voltage (V_force) is needed.

Current Generation

With the relationship in equation 56, we can write Ĥ defined in equation 33 as follows: $\begin{array}{cc}\begin{array}{c}\hat{H}\left(k_{\mathrm{DAC}},I_{\mathrm{force}}\right)=k_{\mathrm{ADC}}=\frac{\left(I_{\mathrm{force}}-I_{\mathrm{offset}}\right)R_{1L}}{V_{\mathrm{LSB}-\mathrm{ADC}}}+\\ k_{\mathrm{DAC}}\times \frac{V_{\mathrm{LSB}-\mathrm{DAC}}}{V_{\mathrm{LSB}-\mathrm{ADC}}}\end{array}& \mathrm{Equation}58\end{array}$Similarly, we can write Ĥ⁻¹ defined in equation 37 as follows: $\begin{array}{cc}\begin{array}{c}{{\hat{H}}^{-1}\left(k_{\mathrm{ADC}},I_{\mathrm{force}}\right)}_{I_{\mathrm{out}}=I_{\mathrm{force}}}=k_{\mathrm{DAC}}\\ =-\frac{\left(I_{\mathrm{force}}-I_{\mathrm{offset}}\right)R_{1L}}{V_{\mathrm{LSB}-\mathrm{DAC}}}+\\ k_{\mathrm{ADC}}\times \frac{V_{\mathrm{LSB}-\mathrm{ADC}}}{V_{\mathrm{LSB}-\mathrm{DAC}}}\end{array}& \mathrm{Equation}59\end{array}$The relationships in equations 58 and 59 mean that the lookup table Ĥ and Ĥ⁻¹ in the force-current system described previously for the case of a general internal load R₁ 216 can be generated by two calibration points when internal load R₁ 216 is a linear resistor. Moreover, from equations 58 and 59, it can also be seen that the lookup tables need not be re-calibrated if a different forcing current is required. This is different from the lookup tables for the generalised current forcing system in Section 2.4, where the functions Ĥ and Ĥ⁻¹ need to be re-calibrated whenever a different forcing current (I_force) is needed.Calibration Circuits

The force-voltage circuit in FIG. 38 is analogous to the front end of the system in FIG. 35. In FIG. 38, a reference current source 708 with value I_ref is connected to the resistor 216 at output 204. The voltage V_DAC at output 204 is forced by the digital value k_DAC-cal 709. When the circuit in FIG. 38 is in equilibrium, the value of voltage V_ADC at output 204 is represented by k_ADC-cal 711. The reference current source I_ref 708 can be expressed as: $\begin{array}{cc}{I}_{\mathrm{ref}}=\frac{\begin{array}{c}{k}_{\mathrm{ADC}-\mathrm{cal}}\times V_{\mathrm{LSB}-\mathrm{ADC}}-\\ K_{\mathrm{DAC}-\mathrm{cal}}\times V_{\mathrm{LSB}-\mathrm{DAC}}\end{array}}{R_{1L}}+I_{\mathrm{offset}}& \mathrm{Equation}60\end{array}$

Another force-voltage circuit is shown in FIG. 39. This circuit is similar to the force-V_ADC architecture shown in FIG. 36. The voltage V_ADC at output 204 is set by the digital value k_ADC-cal 713. When the circuit is shown in FIG. 39 is in equilibrium, the value of voltage V_DAC at intermediate voltage point 214 is represented by k_DAC-cal 714. The expression in equation 60 can also be used to describe current I_ref 708.

A purpose of the calibration process is to determine the values of the constants R_1L and I_offset. To find these two constants, two calibration points are used. Note that either one of the circuits in FIG. 38 and FIG. 39 can be used as the calibration circuit as they both follow equation 60.

For the circuit in FIG. 38, a value of k_DAC-cal 709 and a reference current I_ref 708 is set at each calibration point. Then the resultant k_ADC value, k_ADC-cal 711, is recorded. For the circuit in FIG. 39, a value of k_ADC-cal 713 and a reference current I_ref 708 is set at each calibration point. Then the resultant k_DAC value, k_DAC-cal 714, is recorded.

The values of k_ADC-cal for the two calibration points will be designated as k_ADC-cal1 and k_ADC-cal2. Similarly, the calibration values for k_DAC-cal and I_ref can be written as k_DAC—cal1, k_DAC-cal2, I_ref1 and I_ref2. Using equation 60, two equations can be generated after two calibration points: $\begin{array}{cc}{I}_{\mathrm{ref1}}=\frac{k_{\mathrm{ADC}-\mathrm{cal}1}\times V_{\mathrm{LSB}-\mathrm{ADC}}-k_{\mathrm{DAC}-\mathrm{cal}1}\times V_{\mathrm{LSB}-\mathrm{DAC}}}{R_{1L}}+I_{\mathrm{offset}}& \mathrm{Equation}61\\ I_{\mathrm{ref2}}=\frac{k_{\mathrm{ADC}-\mathrm{cal}2}\times V_{\mathrm{LSB}-\mathrm{ADC}}-k_{\mathrm{DAC}-\mathrm{cal}2}\times V_{\mathrm{LSB}-\mathrm{DAC}}}{R_{1L}}+I_{\mathrm{offset}}& \mathrm{Equation}62\end{array}$Equations 61 and 62 can be solved to yield the values of R_1L and I_offset as follows: $\begin{array}{cc}{R}_{1L}=\frac{\begin{array}{c}\left(k_{\mathrm{ADC}-\mathrm{cal}2}-k_{\mathrm{ADC}-\mathrm{cal}1}\right)\times V_{\mathrm{LSB}-\mathrm{ADC}}+\\ \left(k_{\mathrm{DAC}-\mathrm{cal}1}-k_{\mathrm{DAC}-\mathrm{cal}2}\right)\times V_{\mathrm{LSB}-\mathrm{DAC}}\end{array}}{I_{\mathrm{ref2}}-I_{\mathrm{ref1}}}& \mathrm{Equation}63\\ I_{\mathrm{offset}}=\frac{\begin{array}{c}\left(I_{\mathrm{ref1}}\times k_{\mathrm{ADC}-\mathrm{cal}2}-I_{\mathrm{ref2}}\times k_{\mathrm{ADC}-\mathrm{cal}1}\right)\times \\ V_{\mathrm{LSB}-\mathrm{ADC}}+(I_{\mathrm{ref2}}\times k_{\mathrm{DAC}-\mathrm{cal}1}-I_{\mathrm{ref1}}\times \\ k_{\mathrm{DAC}-\mathrm{cal}2})\times V_{\mathrm{LSB}-\mathrm{DAC}}\end{array}}{\begin{array}{c}\left(k_{\mathrm{ADC}-\mathrm{cal}2}-k_{\mathrm{ADC}-\mathrm{cal}1}\right)\times V_{\mathrm{LSB}-\mathrm{ADC}}+\\ \left(k_{\mathrm{DAC}-\mathrm{cal}1}-k_{\mathrm{DAC}-\mathrm{cal}2}\right)\times V_{\mathrm{LSB}-\mathrm{DAC}}\end{array}}& \mathrm{Equation}64\end{array}$Special Case: Simplified Calibration and Measurement

Equations 57, 58 and 59 can be used to generate a lookup table with the generic calibration procedure described previously in the specification. The section below shows how to simplify equations 63 and 64 to reduce the computational complexity of generating a lookup table for a linear resistive load.

Equations 57, 58 and 59 can be simplified if constraints are imposed on the design of the system and on the calibration procedures. This section describes an example of such a set of constraints that enables effective implementation of the system with a linear element. It will be readily appreciated that other methods for simplifying the computations may also be used without detracting from the spirit of the invention.

To avoid unnecessary arithmetic due to LSB conversions, the system 200 can be designed such that:V_LSB-ADC=V_LSB-DAC  Equation 65In other words, the LSB voltage of ADC 212 and DAC 210 are the same. In the calibration procedure, the force-V_ADC circuit in FIG. 39 is used. The two calibration points are given in Table 8.

TABLE 8
Calibration Points
Calibration
Point	Setting of kADC-cal	Setting of Iref	Recorded kDAC-cal
1	kADC-cal1 = kforce-ref	Iref1 = Iforce-ref	kDAC-cal1
2	kADC-cal2 = kforce-ref	Iref2 = 0	kDAC-cal2

The zero reference current I_ref2 can be easily set by merely disconnecting the output 204. Hence, only one calibration point I_force-ref requires an external reference. The calibration process is summarised in Table 9 with reference to FIG. 39.

TABLE 9
Calibration of the System with Linear R1 216
Step Operation
1    Connect output 204 to the external current reference Iref 708
2    Set the external reference to make Iout = Iref1 = Iforce-ref
3    Force kADC = kforce-ref 713 and record kDAC-cal1 714
4    Disconnect output 204 from the external current reference Iref 708
     to make Iout = Iref2 = 0
5    Force kADC = kforce-ref 713 and record kDAC-cal2 714

With the system requirement in equation 65 and the calibration points in Table 8, equations 57, 58 and 59 can be simplified as follows: $\hspace{1em}\begin{array}{cc}{I}_{\mathrm{out}}=\hat{G}\left(k_{\mathrm{DAC}}^{*},V_{\mathrm{force}}\right)=\left[\frac{k_{\mathrm{DAC}}^{*}-k_{\mathrm{DAC}-\mathrm{cal}2}+k_{\mathrm{force}-\mathrm{ref}}-k_{\mathrm{force}}}{k_{\mathrm{DAC}-\mathrm{cal}1}-k_{\mathrm{DAC}-\mathrm{cal}2}}\right]I_{\mathrm{force}-\mathrm{ref}}& \mathrm{Equation}66\end{array}$ k_ADC=Ĥ(k_DAC, I_force)=k_DAC+L(I_force)  Equation 67k_DAC=Ĥ⁻¹(k_ADC, I_force)|_Iout=Iforce=k_ADC−L(I_force)  Equation 68where: $\begin{array}{cc}\begin{array}{c}L\left(I_{\mathrm{force}}\right)=\left(k_{\mathrm{DAC}-\mathrm{cal}2}-k_{\mathrm{DAC}-\mathrm{cal}1}\right)\frac{I_{\mathrm{force}}}{I_{\mathrm{force}-\mathrm{ref}}}+\\ (k_{\mathrm{force}-\mathrm{ref}}-k_{\mathrm{DAC}-\mathrm{cal}2)}\end{array}& \mathrm{Equation}69\end{array}$

These formulas can be reasonably easily implemented by a digital circuit. The force-voltage-measure-current process using the calibrated values is summarised in Table 10 with reference to FIG. 37.

TABLE 10
Force-Voltage Current Measurement Procedure
Step Operation
1    Connect output 204 to the external load R2 206
2    Set kforce at node 84 according to the required forcing voltage
3    Record k*DAC at node 87
4    Deduce the value of Iout 92 using equation 66

The calibrated system can also be used to force any arbitrary current I_out at output 204 to the external load R₂ 206 and measure the voltage V_ADC at output 204. Any force-current algorithm in FIGS. 23 or 26 can be used. This process is summarised in Table 11.

TABLE 11
Force-Current Voltage Measurement Procedure
Step Operation
1    Connect the output 204 to the external load R2 206
2    Run the force-current algorithm with Ĥ
     or Ĥ−1 given by equations 67 and 68.
     The value Iforce is set to the desired current value.
3    Record kADC

Circuit Implementations

The following part of this specification describes specific examples of implementations of the general system 200 shown in FIGS. 2a and 2b. Circuit implementations of the force-voltage-measure-current and the force-current-measure-voltage algorithms will be described.

General Architectures of the Force-Voltage/Force-Current Algorithms

In a non-limiting implementation, the system 200 shown in FIG. 2b may be implemented using digital logic components and a front-end circuit. The details of the front-end circuit are provided herein below.

Force-Voltage-Measure-Current Algorithm

The general structure of the system for implemented a force-voltage-measure-current circuit is shown in FIG. 40. As depicted, the system includes a lookup table 706 containing Ĝ and a front-end circuit, referred to as “V_ADC-Forcing circuit” 721 in FIG. 40. In this non-limiting implementation, lookup table 706 is implemented by a RAM. Lookup table 706 releases a current measurement at output 715. It will be appreciated that other suitable memory devices may be used without detracting from the spirit of the invention. The V_ADC-Forcing circuit 721 is shown in isolation in FIG. 41. Details of the V_ADC-Forcing circuit 721 will be described further on in the specification.

Force-Current-Measure-Voltage Algorithm

With reference to FIGS. 23a), b), c), d) and FIG. 26, two basic architectures have been described for the Force-Current-Measure-Voltage Circuit, depending on whether the search variable is the voltage V_ADC at output 204 or voltage V_DAC at the intermediate voltage point 214. The two basic architectures are shown in FIGS. 42 and 43. As shown, for both of these architectures, the search logic 720724 accesses a lookup table 722 where either function Ĥ or function Ĥ⁻¹ is stored. In this non-limiting implementation, lookup table 722 is implemented by a RAM device. It will be appreciated that other suitable memory devices may be used without detracting from the spirit of the invention.

For the force-current architecture that uses voltage V_ADC at output 204 as a search variable, shown in FIG. 42, the V_ADC-Forcing circuit 721 shown in FIG. 41 is used as a front-end circuit. For the force-current architecture that uses voltage V_DAC at the intermediate voltage point 214 as a search variable, shown in FIG. 43, the value voltage V_DAC at point 214 is reconstructed from value k_DAC at node 87 using DAC 210. The V_DAC-Forcing circuit 725 shown in FIG. 43 and in FIG. 44 includes DAC 210, internal load R₁ 216 and the ADC 212. Details of the V_DAC-Forcing circuit 725 will be described herein below.

V_DAC-Forcing Circuit 725

In a non-limiting implementation, the architecture of the V_DAC-Forcing Circuit 725 in FIG. 44 includes of a DAC 210 with essentially zero output impedance, hereinafter referred to as a “low-impedance DAC” 210, an internal load element R₁ 216 and an ADC 212 with infinite input impedance, hereinafter referred to as a “high-impedance ADC”. It is to be appreciated that the expression “infinite input impedance” is meant to designate a device, which in its effective operating range will draw a quantity of current that is considered to be essentially negligible. The components of the V_DAC-Forcing Circuit 725 are described in the following sections.

DAC 210 and the Internal Load Element R₁ 216

The partial front-end circuit with the low-impedance DAC 210 and internal load R₁ 216 is shown in FIG. 45. The combination of the DAC 210 and internal load R₁ 216 as separate components is one example of implementation of a circuit module having digital-to-analog conversion functionality and load functionality. The voltage V_ADC at output 204 is given by a function of voltage V_DAC at intermediate voltage point 214 and current I_out at output 204 as follows:V_ADC=H_R1(V_DAC, I_out)  Equation 70where the function H_R1 is dependent on the internal load R₁ 216. The voltage V_DAC at intermediate voltage point 214 is described by:V_DAC=k_DAC×V_LSB-DAC  Equation 71In this description, five alternative specific implementations of the circuit module having digital-to-analog conversion functionality and load functionality in FIG. 45 will be described. It will be readily appreciated that other configurations of the circuit in FIG. 45 may be used without detracting from the spirit of the invention.

Two of these implementations are shown in FIGS. 46a) and 46b). The implementations in FIGS. 46a) and 46b) are hereinafter referred to as Load Configurations A and B respectively. In both circuits, a PDM generator 801 generates digital pulses with a specific density of 1's (hereinafter referred to as “pulse density”) controlled by k_DAC. When the digital pulses pass through the low-pass filter (LPF) formed by R_Ax and C_Ax (where x=1 to m), the output voltage V_ADC at output 204 is settled at a specific DC value that is dependent on the pulse density. Also, it can be seen that for a particular pulse density, the DC value of voltage V_ADC at output 204 will change depending on current I_out at output 204. Therefore, the circuits in FIGS. 46a) and 46b) can be represented by a low-impedance DAC connected with a resistor. The parameters of the equivalent DAC 210 and the internal load R₁ 216 shown in FIG. 45 can be derived as follows. For the circuits in FIGS. 46a) and 46b), if the PDM generator 801 can generate 2ⁿ distinct pulse densities, V_LSB-DAC of the equivalent DAC 210 (in FIG. 45) is given by: $\begin{array}{cc}{V}_{\mathrm{LSB}-\mathrm{DAC}}=\frac{V_{\mathrm{DD}}}{2^n}& \mathrm{Equation}72\end{array}$where V_DD is the voltage representing the high value of the PDM generator 801. Also, the equivalent internal load R₁ 216 is linear and is given by the values in Table 12 below.

TABLE 12
Equivalent load resistances of Load Configurations A and B
Load
Configuration Value of Equivalent R1
A             Equation 73
              R                      1                                        =                                                                  ∑                                                  i                          =                          1                                                m                                            ⁢                                              R                        Ai
B             Equation 74
              R                      1                                        =                                                                  R                        B                                            +                                                                        ∑                                                      i                            =                            1                                                    m                                                ⁢                                                  R                          Ai

Alternatively, the circuit in FIG. 45 can be implemented using other configurations. Consider the three implementations shown in FIGS. 47(a) to (c) (hereinafter referred to as Load Configurations C to E). In these circuits, the PDM generator 801 and the low-pass filter (LPF) 805 corresponds to the low-impedance DAC 210 in FIG. 45. If the PDM generator 801 can generate 2ⁿ distinct pulse densities, V_LSB-DAC of the equivalent DAC 210 in FIG. 45 is given by equation 72.

Each of the MOS circuits 807 in FIGS. 47(a) to (c) correspond to the internal load R₁ 216 in FIG. 45. It can be seen that the equivalent internal load R₁ 216 for any one of the MOS circuits shown is inverting because the equivalent voltage V_DAC at intermediate voltage point 214 (not shown) increases while voltage V_ADC at output 204 decreases for a particular current I_out at output 204. Note that in practice, the current sources in FIGS. 47(b) and (c) of the MOS circuits 807 will generally have a positive differential output resistance. Hence, I_bias is dependent on voltage V_ADC at output 204. Moreover, load configurations D and E (shown in FIGS. 47(b) and (c)) can also be implemented with the bias current sources removed (i.e., I_bias=0). In this case, configuration D (in FIGS. 47(b)) can be used to measure/generate negative output currents (where I_out<0) only while configuration E (in FIGS. 47(c)) can be used to measure/generate positive output currents (where I_out>0) only.

The value of voltage V_ADC at output 204 for the five configurations described above can be easily deduced and is shown in Table 13. In Table 13, voltage V_DAC at intermediate node 214 is the equivalent DAC voltage given by equation 71. The functions H_NOT, H_PMOS and H_NMOS are DC transfer functions of the CMOS inverter, PMOS and NMOS circuits shown in FIGS. 48(a), (b) and (c), respectively, which can be written as:V_out1=H_NOT(V_in1, I_out1)  Equation 75V_out2=H_PMOS(V_in2, I_out2)  Equation 76V_out3=H_NMOS(V_in3, I_out3)  Equation 77

TABLE 13
Value of VADC for different load configurations
Load		Properties of
Configuration	Value of VADC	Equivalent R1
A	Equation 78	non-inverting,
		linear
	$V_{\mathrm{ADC}}=V_{\mathrm{DAC}}+\left[\sum _{i=1}^m{R}_{\mathrm{Ai}}\right]\times I_{\mathrm{out}}$
B	Equation 79
	$V_{\mathrm{ADC}}=V_{\mathrm{DAC}}+\left[R_B+\sum _{i=1}^m{R}_{\mathrm{Ai}}\right]\times I_{\mathrm{out}}$
C	Equation 80	inverting,
	VADC = HNOT(VDAC, Iout)	non-linear
D	Equation 81
	VADC = HPMOS(VDAC, Iout)
E	Equation 82
	VADC = HNMOS(VDAC, Iout)

It will be appreciated that although the examples shown in FIGS. 46(a), 46(b) and 47(a) to (c) include the PDM generator 801, the latter may be replaced by a suitable general-purpose pulse generator that provides a digital pulse.

ADC 212 (Digital Integration/Successive Approximation)

In a non-limiting implementation, the ADC 212 shown in FIG. 44 can be implemented by a delta-modulator of the type described in U.S. Pat. No. 2,605,361 issued to Cassius C. Cutler et al. in Jul. 29, 1952. The content of this document is hereby incorporated by reference. This implementation is shown in FIG. 49.

As shown, the analog-to-digital converter module 212 includes an analog comparator 504, a digital integrator 502 and a feedback circuit. The analog comparator 504 receives a signal indicative of the voltage at the output 204 and a tracking voltage V_track and generates a difference signal on the basis of the signals received. The digital integrator 502 receives the difference signal and generates successive digital approximations of the voltage signal at the output of the circuit device. The feedback circuit processes the successive digital approximations of the voltage signal to generate the tracking voltage V_track and provide the latter to the analog comparator 504. In a non-limiting implementation, the feedback circuit includes a digital-to-analog converter module 500.

In a non-limiting implementation, the DAC 500 in FIG. 49 can be implemented using the voltage reference described in M. M. Hafed, S. Laberge, G. W. Roberts, “A Robust Deep Submicron Programmable DC Voltage Generator”, Proc., IEEE International Symposium on Circuits and Systems, Vol. 4, pp. 5-8, May 2000 and depicted in FIG. 50. The content of this document is hereby incorporated by reference.

The output of the digital integrator 502, k_ADC, increases or decreases by a constant amount depending on the result from the comparator 504. Upon equilibrium, the tracking voltage V_track released by the DAC 500 will equal voltage V_ADC. The value k_ADC will become a digital representation of V_ADC.

In accordance with an alternative specific example of implementation, the ADC 212 shown in FIG. 44 can be implemented by successive-approximation circuit of the type described in D. A. Johns, K. Martin, Analog Integrated Circuit Design, John Wiley & Sons, Inc., pp. 492-493, 1997. The content of this document is hereby incorporated by reference. As shown in FIG. 51, the digital integrator 502 in the analog-to-digital converter module in FIG. 49 is replaced by a successive-approximation register (SAR) 506 module. In this implementation, the Successive-Approximation Register Logic (SAR Logic) 506 binary searches for the value of k_ADC that equalises V_ADC and V_track. The output value k_ADC will be a quantized representation of V_ADC given by: $\begin{array}{cc}{k}_{\mathrm{ADC}}=\frac{V_{\mathrm{ADC}}}{V_{\mathrm{LSB}-\mathrm{ADC}}}& \mathrm{Equation}83\end{array}$where V_LSB-ADC is the LSB voltage of the ADC circuit 212. If we denote the LSB voltage of the DAC 500 that is used in the ADC circuit in FIGS. 49 and 51 by V_LSB-DAC(ADC), the value of V_LSB-ADC will be given by:V_LSB-ADC=V_LSB-DAC(ADC)  Equation 84ADC 212 (Delta Modulator)

In yet another alternative implementation, the ADC 212 is implemented by a delta-modulator of the type described in D. J. G. Janssen, “Delta Modulation in DVM Design”, IEEE Journal of Solid-State Circuits, Vol. SC-7, pp. 503-507, Dec. 1972. The content of this document is hereby incorporated by reference. A non-limiting implementation of the circuit is shown in FIG. 52. As shown, the ADC 212 includes an analog comparator 504, a D-Flip-Flop (D-FF) 508, an RC filter 512 and a frame counter 510. It will be appreciated that although the example shown in FIG. 52 includes a the frame counter 510, the latter may be replaced by a general purpose digital filter implementing accumulation functionality such as a digital integrator for example.

At equilibrium, the tracking voltage V_track will equal voltage V_ADC at output 204. The DC value of V_track (and thus V_ADC) can be deduced by observing the density of 1's (hereinafter referred to as “pulse density”) from the D-FF 508. The frame counter 510 captures a frame of bits from the output of the D-FF 508 and counts the number of 1's in a frame. The resultant count k_ADC will become a digital representation of voltage V_ADC. For example, if the length of the frame captured is 2ⁿ bits, the output value k_ADC will be a quantized representation of V_ADC given by equation 83, where: $\begin{array}{cc}{V}_{\mathrm{LSB}-\mathrm{ADC}}=\frac{V_{\mathrm{DD}}}{2^n}& \mathrm{Equation}85\end{array}$Functional Relationship Between k_ADC, k_DAC and I_out

When the V_DAC-Forcing circuit 725 (FIG. 44) is used to implement the first and second current-searching algorithm, the corresponding function Ĥ defined in equation 33 is dependent on the internal load R₁. From equations 70, 71 and 83, Ĥ can be written as: $\begin{array}{cc}\hat{H}\left(k_{\mathrm{DAC}},I_{\mathrm{out}}\right)=k_{\mathrm{ADC}}=\frac{H_{\mathrm{R1}}\left(k_{\mathrm{DAC}}\times V_{\mathrm{LSB}-\mathrm{DAC}},I_{\mathrm{out}}\right)}{V_{\mathrm{LSB}-\mathrm{ADC}}}& \mathrm{Equation}86\end{array}$where V_LSB-ADC is the LSB voltage of the ADC 212 in the V_DAC-Forcing circuit 725. Using equation 86, the relationships in Table 13 can be rewritten as functions Ĥ listed in Table 14.

TABLE 14
Value of kADC for different load configurations
Load
Configuration Definition of Ĥ
A             Equation 87
              H                        ^                                            ⁡                                              (                                                                              k                            DAC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                                                        k                            DAC                                                    ×                                                      V                                                          LSB                              -                              DAC                                                                                                      +                                                                              [                                                                                          ∑                                                                  i                                  =                                  1                                                                m                                                            ⁢                                                              R                                Ai                                                                                      ]                                                    ×                                                      I                            out                                                                                                                      V                                                  LSB                          -                          ADC
B             Equation 88
              H                        ^                                            ⁡                                              (                                                                              k                            DAC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                                                        k                            DAC                                                    ×                                                      V                                                          LSB                              -                              DAC                                                                                                      +                                                                              [                                                                                          R                                B                                                            +                                                                                                ∑                                                                      i                                    =                                    1                                                                    m                                                                ⁢                                                                  R                                  Ai                                                                                                                      ]                                                    ×                                                      I                            out                                                                                                                      V                                                  LSB                          -                          ADC
C             Equation 89
              H                        ^                                            ⁡                                              (                                                                              k                            DAC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                            H                          NOT                                                ⁡                                                  (                                                                                                                    k                                DAC                                                            ×                                                              V                                                                  LSB                                  -                                  DAC                                                                                                                      ,                                                          I                              out                                                                                )                                                                                            V                                                  LSB                          -                          ADC
D             Equation 90
              H                        ^                                            ⁡                                              (                                                                              k                            DAC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                            H                          PMOS                                                ⁡                                                  (                                                                                                                    k                                DAC                                                            ×                                                              V                                                                  LSB                                  -                                  DAC                                                                                                                      ,                                                          I                              out                                                                                )                                                                                            V                                                  LSB                          -                          ADC
E             Equation 91
              H                        ^                                            ⁡                                              (                                                                              k                            DAC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                            H                          NMOS                                                ⁡                                                  (                                                                                                                    k                                DAC                                                            ×                                                              V                                                                  LSB                                  -                                  DAC                                                                                                                      ,                                                          I                              out                                                                                )                                                                                            V                                                  LSB                          -                          ADC

V_ADC-Forcing Circuits 721

V_ADC-Forcing Circuits 721, of the type shown in FIG. 42, are used in the force-voltage-measure-current algorithm as well as the third and fourth force-current-measure-voltage algorithms. In this section, two specific non-limiting implementations of the V_ADC-Forcing circuits 721 will be described. It will be readily apparent that other suitable implementations are possible without detracting from the spirit of the invention. Before the description of the two implementations, the definition of a function that relates the key variables of the V_ADC-Forcing Circuit 721 is provided in the following.

When a V_ADC-Forcing Circuit 721 is in equilibrium, the quantities k_DAC, k_ADC and I_out will be related by a function which is dependent on the internal load R₁ 216. That function will be described in the following. Let us define a DC transfer characteristic W_R1 for internal load R₁ 216 such that voltage V_DAC at intermediate voltage point 214 is a function of the voltage V_ADC and the current I_out at the output 204. Mathematically, this can be expressed as follows:V_DAC=W_R1(V_ADC, I_out)  Equation 92

When the circuit in FIG. 44 is in equilibrium, k_DAC can be expressed as: $\begin{array}{cc}{k}_{\mathrm{DAC}}=\frac{W_{\mathrm{R1}}\left(k_{\mathrm{ADC}}\times V_{\mathrm{LSB}-\mathrm{ADC}},I_{\mathrm{out}}\right)}{V_{\mathrm{LSB}-\mathrm{DAC}}}& \mathrm{Equation}93\end{array}$where V_LSB-ADC and V_LSB-DAC are the LSB voltages of the equivalent ADC 212 and the DAC 210 employed in the V_ADC-forcing circuit 721, respectively. Function Ŵ_R1 can be defined as: $\begin{array}{cc}{\hat{W}}_{\mathrm{R1}}\left(k_{\mathrm{ADC}},I_{\mathrm{out}}\right)=k_{\mathrm{DAC}}=\frac{W_{\mathrm{R1}}\left(k_{\mathrm{ADC}}\times V_{\mathrm{LSB}-\mathrm{ADC}},I_{\mathrm{out}}\right)}{V_{\mathrm{LSB}-\mathrm{DAC}}}& \mathrm{Equation}94\end{array}$

For each variation of the V_ADC-Forcing Circuit 721 described in the following sub-sections, the function Ŵ_R1 will be derived and listed. At the end of the section, we will show that the function Ŵ_R1 for the V_ADC-Forcing circuits described is related to the functions Ĝ and Ĥ used the force-voltage/force-current algorithms.

V_ADC-Forcing Circuit 721 (Digital Integration/Successive Approximation)

The V_ADC-Forcing Circuit 721 in FIG. 41 is redrawn in FIG. 53. It can be seen that the V_ADC-Forcing circuit 721 includes a V_DAC-Forcing circuit 725 with a digital comparator 702 and a digital integrator 701. Therefore, a V_ADC-Forcing Circuit 721 can be implemented using a V_DAC-Forcing circuit 725 described above in addition to the digital logic shown in FIG. 53.

In a non-limiting implementation where the combination of DAC 210 and internal load R₁ 216 is of the type shown in FIGS. 46a) or b) (type A or B), the equivalent internal load R₁ 216 is linear and the corresponding values are shown in Table 12 above.

If an alternative non-limiting implementation where the combination of DAC 210 and the internal load R₁ 216 is of the type shown in FIGS. 47a), b) or c) (type C or D or E), the equivalent internal load R₁ 216 will be an inverting load. For these implementations, the polarity of the digital comparator 702 in FIG. 53 should be inverted to maintain negative feedback. This modification is shown in FIG. 54 where digital comparator 702 was replaced by digital comparator 730.

Similarly to the ADC circuit 212 described above with reference to FIGS. 49 and 51, the digital integrator 700 in FIGS. 53 and 54 can be replaced by an SAR Logic unit such that the value of k_DAC can be obtained by a binary search.

Upon equilibrium, the function Ŵ_R1 for each load configuration is given by Table 15. In Table 15, V_LSB-ADC and V_LSB-DAC(PDM) are the LSB voltages of the equivalent ADC 212 and the DAC 210. The functions W_NOT, W_PMOS and W_NMOS represent the DC transfer characteristics of the CMOS inverter, PMOS and NMOS circuits shown in FIGS. 48(a), (b) and (c), which can be written as:V_in1=W_NOT(V_out1, I_out1)  Equation 95V_in2=W_PMOS(V_out2, I_out2)  Equation 96V_in3=W_NMOS(V_out3, I_out3)  Equation 97

TABLE 15
The Value of kDAC for different load configurations
Load
Configura-
tion Definition of the function ŴR1
A    Equation 98
     W                          ^                                                R1                                            ⁡                                              (                                                                              k                            ADC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                                                        k                            ADC                                                    ×                                                      V                                                          LSB                              -                              ADC                                                                                                      -                                                                              [                                                                                          ∑                                                                  i                                  =                                  1                                                                m                                                            ⁢                                                              R                                Ai                                                                                      ]                                                    ×                                                      I                            out                                                                                                                      V                                                  LSB                          -                                                      DAC                            ⁡                                                          (                              PDM                              )
B    Equation 99
     W                          ^                                                R1                                            ⁡                                              (                                                                              k                            ADC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                                                        k                            ADC                                                    ×                                                      V                                                          LSB                              -                              ADC                                                                                                      -                                                                              [                                                                                          R                                B                                                            +                                                                                                ∑                                                                      i                                    =                                    1                                                                    m                                                                ⁢                                                                  R                                  Ai                                                                                                                      ]                                                    ×                                                      I                            out                                                                                                                      V                                                  LSB                          -                                                      DAC                            ⁡                                                          (                              PDM                              )
C    Equation 100
     W                          ^                                                R1                                            ⁡                                              (                                                                              k                            ADC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                            W                          NOT                                                ⁡                                                  (                                                                                                                    k                                ADC                                                            ×                                                              V                                                                  LSB                                  -                                  ADC                                                                                                                      ,                                                          I                              out                                                                                )                                                                                            V                                                  LSB                          -                                                      DAC                            ⁡                                                          (                              PDM                              )
D    Equation 101
     W                          ^                                                R1                                            ⁡                                              (                                                                              k                            ADC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                            W                          PMOS                                                ⁡                                                  (                                                                                                                    k                                ADC                                                            ×                                                              V                                                                  LSB                                  -                                  ADC                                                                                                                      ,                                                          I                              out                                                                                )                                                                                            V                                                  LSB                          -                                                      DAC                            ⁡                                                          (                              PDM                              )
E    Equation 102
     W                          ^                                                R1                                            ⁡                                              (                                                                              k                            ADC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                            W                          NMOS                                                ⁡                                                  (                                                                                                                    k                                ADC                                                            ×                                                              V                                                                  LSB                                  -                                  ADC                                                                                                                      ,                                                          I                              out                                                                                )                                                                                            V                                                  LSB                          -                                                      DAC                            ⁡                                                          (                              PDM                              )

V_ADC-Forcing Circuit 721 (Delta Modulation)

The V_ADC-Forcing Circuits 721 in FIGS. 53 and 54 can be simplified into a group of circuits derived from a Delta-Modulator structure. Specific examples of implementation of such circuits are shown in FIGS. 55(a) to (e). Each circuit shown in FIG. 55 includes a DAC 904, an analog comparator 902 (916 in FIGS. 55(c) to (e)), a D-FF 906 and a digital filter in the form of a frame counter 900. It will be appreciated that although the example shown in FIGS. 55(a) to (e) includes frame counter 900, the latter may be replaced by a general purpose digital filter implementing accumulation functionality such as a digital integrator for example. The corresponding set-ups for load configurations shown in FIGS. 46(a) and (b) and in FIGS. 47(a) to (c) (i.e. Load Configurations A to E) are shown in FIG. 55(a) to (e) respectively.

Upon equilibrium, the voltage V_ADC at output 204 will be equal to the DC value set by k_ADC and the DAC 904. As previously described, the value of k_DAC can be deduced by observing the pulse density from the D-Flip Flop 906 using a frame counter 900.

The function Ŵ_R1 for the different load configurations are given by equations 105 to 109 shown in Table 16. In the table, V_LSB-ADC is given by:

 V_LSB-ADC=V_{LSB-DAC(Force-VADC)}  Equation 103

where V_{LSB-DAC(Force-VADC)} is the LSB voltage of the DAC used in the Force-V_ADC circuits in FIGS. 55(a) to (d).

The value of V_LSB-DAC in Table 16 depends on the length of a captured frame in the frame counter 900. For a captured frame with 2ⁿ bits in length, V_LSB-DAC is given by: $\begin{array}{cc}{V}_{\mathrm{LSB}-\mathrm{DAC}}=\frac{V_{\mathrm{DD}}}{2^n}& \mathrm{Equation}104\end{array}$

TABLE 16
Value of kDAC for different load configurations
Load
Configuration Definition of the function ŴR1
A             Equation 105
              W                          ^                                                R1                                            ⁡                                              (                                                                              k                            ADC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                                                        k                            ADC                                                    ×                                                      V                                                          LSB                              -                              DAC                                                                                                      -                                                                              R                            A1                                                    ×                                                      I                            out                                                                                                                      V                                                  LSB                          -                          ADC
B             Equation 106
              W                          ^                                                R1                                            ⁡                                              (                                                                              k                            ADC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                                                        k                            ADC                                                    ×                                                      V                                                          LSB                              -                              DAC                                                                                                      -                                                                              (                                                                                          R                                A1                                                            +                                                              R                                B                                                                                      )                                                    ×                                                      I                            out                                                                                                                      V                                                  LSB                          -                          ADC
C             Equation 107
              W                          ^                                                R1                                            ⁡                                              (                                                                              k                            ADC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                            W                          NOT                                                ⁡                                                  (                                                                                                                    k                                ADC                                                            ×                                                              V                                                                  LSB                                  -                                  DAC                                                                                                                      ,                                                          I                              out                                                                                )                                                                                            V                                                  LSB                          -                          ADC
D             Equation 108
              W                          ^                                                R1                                            ⁡                                              (                                                                              k                            ADC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                            W                          PMOS                                                ⁡                                                  (                                                                                                                    k                                ADC                                                            ×                                                              V                                                                  LSB                                  -                                  DAC                                                                                                                      ,                                                          I                              out                                                                                )                                                                                            V                                                  LSB                          -                          ADC
E             Equation 109
              W                          ^                                                R1                                            ⁡                                              (                                                                              k                            ADC                                                    ,                                                      I                            out                                                                          )                                                              =                                                                                            W                          NMOS                                                ⁡                                                  (                                                                                                                    k                                ADC                                                            ×                                                              V                                                                  LSB                                  -                                  DAC                                                                                                                      ,                                                          I                              out                                                                                )                                                                                            V                                                  LSB                          -                          ADC

Comparing the equations in Table 16 to that in Table 15, it can be seen that the structures in FIG. 55(a) to (d) can also be mapped into the generalised architecture of the V_ADC-forcing circuit in FIG. 41. For the load configuration of type shown in FIGS. 46(a) and (b) (type A or B), the equivalent internal load R₁ 216 is linear and the corresponding values are shown in Table 17. For load configurations shown in FIGS. 47(a) to (c) (type C, D or E) the equivalent internal load R₁ 216 will be an inverting load.

TABLE 17
Equivalent load resistances
Load Configuration Value of Equivalent R1
A                  Equation 110
                   R1 = RA1
B                  Equation 111
                   R1 = RA1 + RB

Functional Relationship Between k_ADC, K_DAC and I_out

When the V_ADC-Forcing circuit 721 (FIG. 41) is used to implement the force-voltage-measure-current or the third or fourth current-searching algorithm, the corresponding functions Ĝ and Ĥ, defined in equations 18 and 33, will be dependent on the internal load R₁.

Comparing equation 94 to equations 20 and 37, it can be seen that Ŵ_R1 represents Ĝ⁻¹ in the force-voltage-measure-current algorithm and Ĥ⁻¹ in the force-current measurement voltage algorithm, i.e.:Ĝ⁻¹(k_ADC, I_out)|_kADC=kforce=k*_DAC=Ŵ_R1(k_ADC, I_out)|_kADC=kforce  Equation 112 Ĥ⁻¹(k_ADC, I_out)|_Iout=Iforce=k_DAC=Ŵ_R1(k_ADC, I_out)|_Iout=Iforce  Equation 113Therefore, for all the V_ADC-Forcing circuit 721 with Ŵ_R1 defined in equations 105 to 109, the corresponding Ĝ and Ĥ for the force-voltage-measure-current algorithm and the third and fourth force-current-measure-voltage algorithm are given by equations 112 and 113.

SPECIFIC EXAMPLES OF IMPLEMENTATION

Example 1

System with Non-linear Inverting Load Element R₁ 216

In a first alternative specific example, depicted in FIG. 56, the system includes a V_ADC-Forcing circuit described in FIG. 55(c) with the DAC 904 implemented by the circuit shown FIG. 50. A first-order RC filter is used for the DAC 904. In the figure, both the PDM generator and the frame counter 900 have 8-bit (256 distinct levels) resolution. The analog full-scale voltage range is from 0 V to 5 V.

Current Measurement

Using the V_ADC-Forcing circuit in FIG. 56, a force-voltage-measure-current system can be constructed as described previously in the specification. The resultant system is shown in FIG. 57. In this example, let us assume that we need to force 3.75 V at the output 204 (V_ADC). According to equations 72 and 103, the value of k_ADC should be set as follows: $\begin{array}{cc}{k}_{\mathrm{ADC}}=k_{\mathrm{force}}=\frac{V_{\mathrm{ADC}}}{V_{\mathrm{LSB}-\mathrm{DAC}\left(\mathrm{Force}-\mathrm{VADC}\right)}}=\frac{3.75V}{\left(\frac{5V}{256}\right)}& \mathrm{Equation}114\end{array}$

To calibrate this ammeter for a 5-bit current resolution, 32 currents, each with an increment of {fraction (1/32)} of the full-scale current, is applied to the output 204 in the manner shown in FIG. 58. For each current increment, the corresponding k*_DAC from the frame counter is recorded. The result is tabulated as that listed in Table 18. In this example, the full-scale current range is from −1 mA to 0.9375 mA. Note that Table 18 is a mapping of the function Ĝ⁻¹ defined in equation 18. To find Ĝ, we simply invert the second and third column of the table.

TABLE 18
A 5-bit look-up table for the ammeter
Entry #	Equivalent Isweep 924	Measured k*DAC
0	−1.0000 mA	1
1	−0.9375 mA	5
.	.	.
.	.	.
.	.	.
16	0.0000 mA	105
17	0.0625 mA	108
.	.	.
.	.	.
.	.	.
30	0.8750 mA	249
31	0.9375 mA	254

After the calibration, the system in FIG. 57 can then be used to set 3.75 V at an unknown external load 206, from which the appropriate load current I_out flows. The resultant bit code k*_DAC is then compared to the counts found in Table 18. For example, if an unknown current is applied and a k*_DAC of 108 is produced, the unknown current will be found to be 0.0625 mA.

Current Generation

Using the V_ADC-Forcing circuit in FIG. 56, the force-current-measure-voltage system shown in FIG. 33 can be constructed. The resultant system is shown in FIG. 59. This system employs the third current-forcing algorithm described previously, where k_ADC is controlled by a search unit 928 that observes the value of k_DAC. As shown in FIG. 59, the digital logic 928 includes a look-up table, a digital comparator 930 and an integrator 932.

The lookup table 933 containing Ĥ⁻¹ needs to be calibrated whenever a new value of I_force applied at input 202 is required. That can be done using the set-up shown in FIG. 60. During calibration, the I_out is set to I_force by an external current reference 926 while k_ADC is swept. For each K_ADC increment, the corresponding value k_DAC from the frame counter 900 is recorded. The resultant lookup table lists the value of k_DAC as a function of k_ADC, i.e., the function Ĥ⁻¹ defined in equation 37. An example lookup-table is shown in Table 19.

TABLE 19
An 8-bit look-up table for the current source
kADC	Actual VADC	Measured kDAC
0	0.00 V	254
1	0.02 V	253
.	.	.
.	.	.
.	.	.
254	4.96 V	1
255	4.98 V	1

After calibration, an external load 206 can be applied to the output 204 of the system as shown in FIG. 59. A current of value I_force will be forced into the external load R₂ 206 when the system is in equilibrium.

Example 2

System with Linear Load Element R₁

In a second alternative specific example, depicted in FIG. 61, the system is constructed using the V_ADC-Forcing circuit described in FIG. 55(a) with the DAC 904 implemented by the circuit in FIG. 50. Again, a first-order RC filter is used for the DAC 904. In the FIG., both the PDM generator and the frame counter 900 have 10-bit (1024 distinct levels) resolution. The analog full-scale voltage range is from 0 V to 2.56 V.

The architectures/operations of the Force-voltage and Force-current systems derived from the V_ADC-Forcing circuit in FIG. 61 is similar to those described with regards to the first alternative specific example described above. However, because the internal load element R₁ is linear, calibration/measurement method described in connection with a linear resistive internal load R₁ 216 can be used.

As both the equivalent ADC and DAC 904 in the V_ADC-Forcing circuit shown in FIG. 61 have 10-bit resolution over a full scale of 0 V to 2.56 V, the LSB voltage (V_LSB) can be defined as follows: $\begin{array}{cc}{V}_{\mathrm{LSB}}=V_{\mathrm{LSB}-\mathrm{DAC}}=V_{\mathrm{LSB}-\mathrm{ADC}}=\frac{2.56V}{1024}=2.5\mathrm{mV}& \mathrm{Equation}115\end{array}$

In a non-limiting implementation, the resistance R₂ in the RC circuit 932 in FIG. 61 is. 2.5 kohms, then from equation 110, the equivalent internal load R₁ 216 in the system will also equal to 2.5 kohms. This resistance value described in equation 110 into equation 56 to get: $\begin{array}{cc}{I}_{\mathrm{out}}=\frac{k_{\mathrm{ADC}}-k_{\mathrm{DAC}}}{R_{1L}}\times V_{\mathrm{LSB}}+I_{\mathrm{offset}}=\left[k_{\mathrm{ADC}}-k_{\mathrm{DAC}}\right]\times 1\mathrm{\mu A}+I_{\mathrm{offset}}& \mathrm{Equation}116\end{array}$Calibration

The method summarised in Table 9 can be used to calibrate the system depicted in FIG. 61. In this example, we first set k_ADC to an arbitrary reference level (k_force-ref) at 612. For the calibration process, in a first step a current reference (I_force-ref) 936 of +200 uA is applied to the system as shown in FIG. 62 and the value of k_DAC as k_DAC-cal1 is recorded. In a second step, the output 204 is disconnect from any external sources/loads to make current I_out zero and the corresponding k_DAC as k_DAC-cal2 is recorded. The resultant calibration points are summarised in Table 20. These calibration values will be used in generating the lookup tables required in the force-voltage/force-current algorithms described herein below.

TABLE 20
Calibration Points in the Example
Calibration Point	Setting of kADC-cal	Setting of Iref	Recorded kDAC-cal
1	kADC-cal1 =	Iref1 =	kDAC-cal1 = 405
	kforce-ref = 612	Iforce-ref = 200
2	kADC-cal2 =	Iref2 = 0	kDAC-cal2 = 605
	kforce-ref = 612

Current Measurement

In a non-limiting implementation, the force-voltage-measure-current system described in connection with FIG. 37 and with reference to equation 57 can be constructed using the V_ADC-Forcing circuit shown in FIG. 61. The resultant system is shown in FIG. 63. For this current-measurement system, the lookup table Ĝ 936 is a simple expression generated by calibration data. If the calibration points in Table 20 are substituted into equation 66 and the equation is simplified, the following can be obtained:I_out=Ĝ(k*_DAC,k_force)=[k_force−k*_DAC−7]×1 μA  Equation 117Knowing equation 117, the calibrated system in FIG. 63 can then be used to set an arbitrary voltage at output 204, from which a current I_out flows. In this example, let us assume that we need to force 1.28 V at the output 204. (V_ADC). That means the corresponding value of k_ADC will be given by equations 72 and 103 as: $\hspace{1em}\begin{array}{cc}{k}_{\mathrm{ADC}}=k_{\mathrm{force}}=\frac{V_{\mathrm{ADC}}}{V_{\mathrm{LSB}-\mathrm{DAC}\left(\mathrm{Force}-\mathrm{VADC}\right)}}=\frac{1.28V}{\left(\frac{2.56V}{1024}\right)}=512& \mathrm{Equation}118\end{array}$

The resultant bit code k*_DAC can then be substituted in equation 117 to find the current value for I_out at output 204. For example, if an unknown current is applied and a k*_DAC of 866 is produced, the value of the unknown current I_out will be given by:I_out=Ĝ(866,512)=[512−866−7]×1 μA=−361 μAI_out={acute over (G)}(866,512)=[512−866−71]×1 μA=−361 μA  Equation 119Current Generation

In accordance with a specific non-limiting implementation, an embodiment of the force-current-measure-voltage system shown in FIG. 23(c) can be constructed using the V_ADC-Forcing circuit in FIG. 61. The resultant system is shown in FIG. 64. For this force-current system, the lookup table {acute over (H)}⁻¹ is a simple expression generated by calibration data. If we substitute the calibration points in Table 20 into equation 69 and 68, the following is obtained: $\begin{array}{cc}L\left(I_{\mathrm{force}}\right)=\frac{I_{\mathrm{force}}}{1\mu A}+7& \mathrm{Equation}120\\ k_{\mathrm{DAC}}={\hat{H}}^{-1}\left(k_{\mathrm{ADC}},I_{\mathrm{out}}\right){|}_{I_{\mathrm{out}}=I_{\mathrm{force}}}=k_{\mathrm{ADC}}-\frac{I_{\mathrm{force}}}{1\mu A}-7& \mathrm{Equation}121\end{array}$

Knowing the result of equation 121, the system can then be used to set an arbitrary current at output 204 node. In this example, let us assume that a current I_force of 50 uA must be forced at the output 204. The corresponding relationship in equation 121 can be simplified as:

 k_DAC=Ĥ⁻¹(k_ADC, I_out)|_{Iout=50 μA}=k_ADC−57  Equation 122

After defining the lookup table Ĥ⁻¹ from the calibration data, an external load R₂ 206 can be applied to the output 204 as shown in FIG. 64. With Ĥ⁻¹ defined in equation 122, a current of value I_force=50 uA will be forced into the load R₂ 206 when the system is in equilibrium.

The general structure of the circuits shown in FIGS. 63 and 64 is shown in FIGS. 65 and 66 respectively. The set of circuit components labelled 1050 in FIGS. 63 to 66 implement an equivalent functionality.

As shown in FIG. 65, the system includes an input 1012 for receiving a digital forced voltage signal, an output 204 suitable for releasing to an external load a signal approximating the digital forced voltage signal, a first circuit segment and a second circuit segment. The first circuit segment is between the input 1012 and the output 204. The second circuit segment is connected in a feedback arrangement between the output 204 and the first circuit segment and provides the first circuit segment with the voltage associated with the output 204. The first circuit segment includes a first digital-to-analog converter module 904, a difference module 934, an analog-to-digital converter module 906, a second digital-to-analog converter module 1002, an analog accumulation device 1000, an intermediate voltage point 214 and in internal load 216 between the intermediate voltage point 214 and the output 204. The first digital-to-analog converter module 904 processes the digital forcing parameter signal and generates an analog value of the digital forcing parameter signal. The difference module 934 processes the analog value of the digital forcing parameter signal as well as the voltage associated with the output of the circuit device and generates a difference signal. The polarity of the difference module 934 is reversed if internal load 216 is an inverting load. The analog-to-digital converter module 906 processes the difference signal and generates a digital representation of the difference signal. The second digital-to-analog converter module 1002 processes the digital representation of the difference signal and generates an analog representation of the difference signal. The analog accumulation device 1000 generates a second voltage signal on the basis of the analog representation of the difference signal and applies the second voltage signal to the intermediate voltage point 214. The application of the second voltage signal to the intermediate voltage point 214 causes a change in either one of the voltage signal or the current signal at the output such that a voltage approximating the forced voltage signal is caused at the output 204.

In a first non-limiting implementation, the analog accumulation device 1000 includes an analog integrator module. In a second non-limiting implementation, the analog accumulation device 1000 includes a low-pass filter unit. It will be appreciated that other suitable equivalent devices instead of a low-pass filter unit may be used without detracting from the spirit of the invention.

The system shown in FIG. 65 also includes a current measurement circuit having a first input for receiving the digital representation of the difference signal generated by the analog-to-digital converter module 906, a second input for receiving a digital forced voltage signal from input 1012, a digital filter implementing accumulation functionality 900, search logic 936 and an output 938. The digital filter 900 is adapted for processing the difference signal received at the first input for generating an average voltage value. The search logic 936 receives the digital forced voltage signal and the average voltage value from the second input and the digital filter 900. The search logic 936 derives a certain current measurement on the basis of the digital forced voltage signal and the average voltage value such that, when the system is in equilibrium, the certain current measurement is indicative of an approximation of measurement of the current flowing at the output 204 of the voltage generating circuit. The certain current measurement is then released at the output 938.

FIG. 67 shows an alternative general structure of the circuit shown in FIG. 66. In this alternate configuration, the digital filter 900 of FIG. 66 present in the feedback path has been replaced by digital filter 1090 located between ADC 906 and DAC 1002. Digital filter 1090 may implement any suitable accumulation functional element such as for example a digital integrator. The output of digital filer 1090 is provided to the search control logic 940. It will be appreciated that binary searches can also be performed by replacing the digital filter 1090 with a successive-approximation circuit (SAR) module without detracting from the spirit of the invention. The corresponding alternate structures associated with the circuit shown in FIG. 65 will be readily apparent to the person skilled in art upon a reading of this specification and as such will not be described further here.

Other specific examples of implementation of this invention are presented in C. K. L. Tam, G. W. Roberts, “A Robust DC Current Generation and Measurement Technique for Deep Submicron Circuits”, Proc. IEEE International Symposium on Circuits and Systems, Vol. 1, pp. 719-722, May 6, 2001. The contents of this document are hereby incorporated by reference.

Specific Physical Implementation

Those skilled in the art should appreciate that in some embodiments of the invention, all or part of the functionality previously described herein with respect to the circuit device and system may be implemented as pre-programmed hardware or firmware elements (e.g., application specific integrated circuits (ASICs), FPGA chips, ROM, PROM, EPROM, etc.), or other related components.

For example, the above described circuits may be incorporated in IC generally, diagnostic tools, IC testing equipment, on-chip testing and IC including on-chip testing functionality amongst others.

Specific non-limiting examples of use of the above-described system include:

• • the measurement of leakage currents in bonding pads of ICs;
  • continuity tests for bonding wires;
  • parametric testing of logic output pins;
  • parametric testing of output drivers (analog/digital);
  • IDDQ testing of an IC or modules on an IC;
  • parametric testing of resistors/diodes/transistors;
  • parametric testing of voltage/current sources, bias networks or other analog components.

Although the present invention has been described in considerable detail with reference to certain preferred embodiments thereof, variations and refinements are possible without departing from the spirit of the invention. Therefore, the scope of the invention should be limited only by the appended claims and their equivalents.

Claims

1. A method of performing a DC parametric test on a first load, comprising the steps of: a) electrically coupling an output of a second load to the first load; b) applying a first electrical signal to said second load so as to cause a second electrical signal at said output; c) feeding back said second electrical signal to a search entity; d) inputting a forcing parameter signal to said search entity; and e) determining said first electrical signal as a function of said second electrical signal and said forcing parameter signal.

2. A method according to claim 1, further comprising the step of measuring at least one DC parametric signal corresponding to said first load.

3. A method according to claim 2, wherein step a) includes forming an electrically conductive pathway between said second load and said first load and the step of measuring said at least one DC parametric signal includes measuring said at least one DC parametric signal along said electrically conductive pathway.

4. A method according to claim 1, wherein said search entity comprises digital search control logic and step e) comprises converting said second electrical signal from an analog signal to a digital signal and converting said first electrical signal from a digital signal to an analog signal, each prior to applying said first electrical signal to said second load.

5. A method according to claim 4, wherein step e) further comprises converting said forcing parameter signal from an analog signal to a digital signal.

6. A method according to claim 1, wherein step e) comprises generating a digital difference voltage signal dependent on a difference between said forcing parameter signal and a digital approximation of said second electrical signal.

7. A method according to claim 6, wherein step e) further comprises converting said forcing parameter signal from an analog signal to a digital signal.

8. A method according to claim 6, wherein step e) further comprises processing said digital difference signal to derive a digital voltage signal.

9. A method according to claim 8, wherein said first electrical signal is an analog signal and step e) further comprises converting said digital voltage signal to said first electrical signal.

10. A method according to claim 8, further comprising the step of allowing said digital voltage signal to settle to a digital target voltage over a period of time.

11. A method according to claim 10, further comprising, after the step of allowing said digital voltage signal to settle, the step of measuring at least one DC parametric signal corresponding to said first load.

12. A method according to claim 1, wherein step d) comprises inputting a forcing voltage signal.

13. A method according to claim 1, wherein step d) comprising inputting a forcing current signal.

14. A method according to claim 1, wherein the first load responds to said second electrical signal with an output current flowing between the first load and said second load and the method further comprises the step of measuring said output current.

15. A method according to claim 14, wherein the step of measuring said output current includes determining said output current as a function of said forcing parameter signal and said first electrical signal.

16. A method according to claim 15, wherein the step of measuring said output current includes the step of determining said output current as a function of a digital approximation of said forcing parameter signal and a digital approximation of said first electrical signal.

17. A method according to claim 16, wherein the step of determining said output current comprises performing a look-up using said digital approximation of said forcing parameter signal and said digital approximation of said first electrical signal.

18. A method of performing a DC parametric test on a first load, comprising the steps of: a) applying a first electrical signal to the first load; b) generating a second electrical signal as a function of said first electrical signal and a forcing parameter signal; and c) passing said second electrical signal through a second load so that said second load outputs said first electrical signal.

19. A method according to claim 18, wherein step b) comprises converting said first electrical signal into a digital signal.

20. A method according to claim 19, wherein step b) further comprises performing a look-up to determine said second electrical signal.

21. A method according to claim 18, wherein said first electrical signal comprises an output current and the method further comprises the step of measuring said output current.

22. A method according to claim 21, wherein the step of measuring said output current includes determining said output current as a function of said forcing parameter signal and said second electrical signal.

23. A method according to claim 22, wherein the step of measuring said output current includes the step of determining said output current as a function of a digital approximation of said forcing parameter signal and a digital approximation of said second electrical signal.

24. A method according to claim 23, wherein the step of determining said output current comprises performing a look-up using said digital approximation of said forcing parameter signal and said digital approximation of said second electrical signal.

25. A method of performing a DC parametric test on a first load, comprising the steps of: a) applying a first electrical signal to the first load; b) receiving a forcing parameter signal; c) causing said first electrical signal to approximate said forcing parameter signal as a function of a feedback signal of said first electrical signal and said forcing parameter signal; and d) measuring a DC parametric signal of the first load, said DC parametric signal being a response of the first load to said first electrical signal.

26. A method according to claim 25, wherein step a) includes applying said first electrical signal to the first load along an electrically conductive path and said DC parametric signal occurs along said electrically conductive path.

27. A method according to claim 26, wherein said DC parametric signal is a current signal and step d) comprises determining said current signal as a function of said forcing parameter signal and a second electrical signal applied to a second load that generates said first electrical signal as a function of said second electrical signal.

28. A method according to claim 27, wherein step d) includes the step of determining said current signal as a function of a digital approximation of said forcing parameter signal and a digital approximation of said second electrical signal.

29. A method according to claim 28, wherein the step of determining said current signal comprises performing a look-up using said digital approximation of said forcing parameter signal and said digital approximation of said second electrical signal.

30. A method according to claim 25, wherein step c) comprises generating a second electrical signal and passing said second electrical signal through a second load so as to cause said second load to output said first signal.

31. A method according to claim 30, wherein the step of generating said second electrical signal includes generating said second electrical signal as a function of said first electrical signal and said forcing parameter signal.

32. A method according to claim 31, wherein the step of generating said second electrical signal includes generating a digital difference voltage signal dependent on a difference between said forcing parameter signal and a digital approximation of said second electrical signal.

33. A method according to claim 32, further comprising processing said digital difference signal to derive a digital voltage signal.

34. A method according to claim 33, wherein said second electrical signal is an analog signal and step c) further comprises converting said digital voltage signal to said second electrical signal.

35. A method according to claim 25, further comprising, prior to step d), the step of allowing said first electrical signal to settle over a period of time.