WO2001050119A1

WO2001050119A1 - Method and apparatus for measurement of electrochemical cell and battery impedances

Info

Publication number: WO2001050119A1
Application number: PCT/US2000/035044
Authority: WO
Inventors: James D. Kozlowski
Original assignee: The Penn State Research Foundation
Priority date: 2000-01-03
Filing date: 2000-12-22
Publication date: 2001-07-12
Also published as: AU4133601A

Abstract

A battery's impedance is measured by a technique that normally uses a current divider network which is connected to the battery. The circuit used according to this technique has a current generator (53) producing a regulated current signal (Ii) and has one or more sensing impedances (Zs) which are normally positioned electrically parallel, or in some alternate embodiments in series, with the battery. A DC-blocking capacitor prevents the battery voltage from draining into the one or more sensing impedances (Zs). A magnetic field sensor or comparable device measures the magnitude and/or phase of current passing through the sensing impedances. Substitution of a number of calibrated impedances into the circuit in place of the battery permits an initial mathematical computation of the battery's impedance utilizing this technique. Thereafter battery impedances can be computed with the current without using calibrated impedances.

Description

Method and Apparatus for Measurement of Electrochemical Cell and Battery Impedances

Background

This invention relates to techniques for measuring impedance in electrochemical cells. More particularly, the invention is directed to apparatuses and methods used for taking internal impedance measurements of electrochemical batteries and cells with improved sensitivity and noise/electromagnetic immunity as compared to currently existing methods.

Electrochemical batteries and cells have very low internal impedance. This is true in different types of cells, including those based on either lead acid or nickel cadmium chemistries

for which impedances can be on the order of milliohms (mΩ). For this reason, an effective

method for measuring impedance must be highly sensitive to small impedance values while being immune to noise and electromagnetic circuit interference. Prior methods of impedance measuring normally utilize one of five different types of electrical circuits: (1) bridge circuits; (2) voltage dividers; (3) 4-wire connections; (4) short circuits; and (5) time constant circuits. However, each of these methods is limited by the inherent characteristics of the particular circuit type used in performing the impedance measurement. Bridge circuits are commonly used to sense impedance changes in batteries. Such a bridge circuit 20 is depicted in FIG. 1, which shows the basic configuration of a circuit of this type which is powered by an AC voltage source Vj. These circuits generally include impedance elements 22 that are located along first and second current paths 24 and 26, the impedance elements 22 being located on either side of voltage divider points where the voltages V_A and VB can be measured. For battery impedance measurements, one of the impedance elements 22 in the bridge represents the battery being measured. The output of the bridge, V₀, is the potential difference between V_A and V_B. The voltages V_A and V_B are related to the input voltage, V; by the relation V_A = Vi Z3 and V_B = V Z4

Z1 + Z3 Z2 + Z4

under the condition that V₀ is equal to zero (i.e. VA - VB ), so that

(Z1) (Z4) = (Z2) (Z3).

For example, one way of using this circuit is to make one of the impedance elements 22 adjustable and adjust the value of the impedance until V₀ is equal to zero. The problem with this type of operation is that it requires continuous adjustment of the element for each frequency at which the measurement is made. This is because battery impedance is not constant over the frequency spectrum of interest.

An automated system for handling such a procedure is complex and difficult to implement. This circuit is typically used by picking nominal values of the three known impedance elements 22 to maximize the output voltage swing as the battery impedance changes through the sweep of frequencies and usable life. The sensitivity of the output is maximized when Z2 = Z3 and Zl = Z4. This implies that one of the impedance elements 22 must have a value that is the complex conjugate of the battery impedance.

Another limitation of bridge circuits relates to the fact that since internal impedance is very low for most cell types, voltage drops across the battery will also be very low. Fixing the values of all but one impedance element 22 and allowing only this battery impedance element to change implies that either V_A or V_B will remain constant. The bridge 20 reduces to a voltage divider for changes in the battery impedance. The output voltage is inversely proportional to changes in the battery impedance. Thus, as the impedance of the battery increases, output voltage becomes smaller. To get sufficiently large voltage drops at the output, a large amount of current is required. For example, if the magnitude of the battery impedance were 5mΩ, a 1A

current would be required to produce a 5mV drop at the output. Such a condition would place a high gain requirement on any sensing amplification equipment connected at the output of the bridge circuit. For example, the input impedance of such an amplifier would be the impedance of the bridge circuit 20 and would be very low due to the low battery impedance. Where such low input impedances are involved, such as those below 1Ω, amplifiers become highly susceptible to electrical field noise, whether self-generated or from other sources. This condition is compounded where the input signal is also very low. Adverse interference effects can be expected regardless of whether BJT or FET input stages are used. Although the addition of a transformer across V₀ is typically recommended in cases of low input impedance, the addition of such a device tends to contribute to circuit impedance, lowering the circuit's sensitivity. Alternatively, where a sufficiently high turns ratio is present, an added transformer can reduce the bandwidth of the output signal produced.

Since bridge circuits do not easily permit impedance sensing without adjustment of the known impedance elements 22, the circuit has no more sensitivity than the voltage divider circuit. Thus, such circuits are normally only usable in laboratory settings where the impedance elements can be adjusted.

A second commonly used technique for impedance measuring uses a voltage divider circuit, which is typically preferred over bridge circuits when adjustment of impedance is not required. A voltage divider circuit 28 used for battery impedance measurements is shown in FIG. 2. The circuit 28, like most designs of this type, is driven by an AC current source 30 since voltage levels are typically in the range of millivolts and current in the range of amperes and thus amperage is easier to regulate than voltage. The circuit includes a sensing impedance Z_s and a battery impedance Zb in a series loop 29 with the AC current source. Each sensing and battery impedance has a respective sensor 31 that connects to the series loop 29 at the respective impedance's point of positive and negative potential. Each sensor 31 is separated from the series loop 29 by capacitors 32 used to block the battery's DC signal. This technique involves two measurements: (1) measurement of the voltage V_s across a sensing impedance Z_S; permitting measurement of the loop current given the known size of Z_s; and (2) measurement of the voltage Vb across the battery 34 being measured.

Voltage divider circuits used to measure battery impedance are limited by the same disadvantages as bridge circuits. Like bridge circuits, voltage measurements are taken in the millivolt signal level since batteries have very low impedance. Thus, voltage divider circuits, like bridge circuits, are susceptible to electrical field noise and have limited sensitivity.

A third technique utilizes a circuit known as a 4-wire or "Kelvin" connection. This is among the most frequently used techniques for measuring battery impedance and has been described in numerous patents and other references. The general configuration of a 4-wire connection 36 is shown in FIG. 3. In principle, this circuit is very similar to a voltage divider circuit, being driven by a current source 37. But the 4-wire connection 36 does not have a sensing impedance Z_s. A battery 38 is interrogated with a current signal, and the voltage drop Vb across the battery 38 is measured with a sensor 31 separated from the battery nodes by capacitors 32. As indicated above, for most lead acid and nickel cadmium cells, the internal impedance Zb is very low. This means that the battery 38 will be driven with amperes of current, and output signals will be on the order of millivolts of potential.

Most of the problems associated with bridge circuits and voltage dividers also apply to 4-wire connections. In fact, U.S. Patent No. 5,821,757 to Alvarez et al. specifically addresses the problem of reducing electromagnetic interference (EMI) that adversely affects the 4-wire connection described in U.S. Patent No. 5,281,920 to Warst with the addition of twisted, shielded-paired wires. Other attempts to reduce system noise have included the incorporation of ground isolation, the selection of driving frequencies away from known sources of electric field noise, and the combined techniques of windowing and averaging. The fact that there is a need for each of these attempted remedies demonstrates the inherent limitations of this type of circuit. In such a system, larger output signals require a larger input current signal. For example, output signals on the order of tens of millivolts require input signals on the order of tens of amperes. However, sensitivity tends to be inversely related to the impedance of a battery. Since larger cell sizes ultimately lead to progressively smaller internal impedances, then for progressively larger cells, output voltages produced using the 4- wire technique tend to be smaller for the same input current. It follows that this technique is generally inadequate for using in a broad range of cell sizes.

Another technique used to measure battery impedance is the short circuit configuration. This configuration is less common than others described above and has been used in applications where internal impedance magnitudes have been on the order of hundreds to thousands of ohms, such as in lithium iodine batteries used in pacemakers and related devices. A simplified illustration of a short circuit 40 is shown in FIG. 4. The circuit has a switch 41 connected to the positive node of a battery 42 having an impedance Zb. The battery's negative node is grounded, while the switch 41 connects the positive node to a current mirror 43 and to ground 45. This technique simply involves taking a voltage measurement on the unloaded battery 42 followed by a measurement of the short-circuited current to calculate a measure of the battery's internal impedance Zb. The short circuit is only applied long enough to get an accurate enough measurement of the discharge current. Although this technique is useful for calculating impedance in small, lithium iodine batteries, other larger battery types, including larger lithium iodine and most lead acid batteries, pose a serious explosion hazard when similarly short circuited. This technique is also limited in that it can only be used to get a bulk number to represent the battery's internal impedance, which eliminates all phase and frequency related information. One additional technique that is commonly used to measure battery impedance is the time-constant method. As demonstrated in the example circuit in FIG. 5, this method is based on the concept of an RC time response of a battery 44 where R is contributed from a battery 46 and a capacitor 45 is a selected known value C. The battery is connected between ground and a normally-open switch 47 which is connected through capacitor 45 to ground. The charge V_c across the capacitor 45 can be monitored through operational amplifier 49. In operation, switch 47 is closed, causing the battery voltage Vbto discharge through battery resistance Rb to charge capacitor 45. The time it takes to charge the capacitor 45 to the voltage Vb is used to determine battery resistance Rb since the capacitance C of capacitor 45 is known and the time τ = RbC.

This method has been incorporated into lithium iodine cells used in medical devices such as pacemakers. It includes switching a battery into a circuit with a parallel capacitor and then measuring the time response to determine the time constant, τ = RC.

As with other techniques, the battery's internal impedance is assumed to be a resistive element and the resulting measurement is reduced to a bulk number. No information about phase or frequency contributions is measured or determined. The technique is also limited in that there is a necessary tradeoff between capacitor size and processing speed of the detection circuit. A larger capacitor requires a larger amount of energy to be drawn from the battery, while the smaller the capacitors, the less time there is for the detection circuit to determine the time constant, affecting the sensitivity of the circuit. This relative dependence on the capacitor's size ultimately affects the circuit's sensitivity.

Most prior art methods of measuring internal impedance in batteries rely heavily on taking voltage measurements. Due to the very low impedance magnitudes involved, output signals are normally expected in the range of millivolts. This means that in order for most prior art methods to be operable, high gain amplifiers having a combination of low voltage signals and low input impedances to the amplifier must be used, implying a high level of susceptibility to noise and EMI. The related apparatus sensitivities of most prior art methods are also related to the impedance of the measured battery. As battery cells become progressively larger, internal impedance becomes smaller. Voltage measurements in turn become progressively smaller, thereby reducing the sensitivity of the measuring circuit. Although increasing input current can improve the output signal, such a step can be prohibitive since a magnification from amperes to tens of amperes may be required to achieve the desired effect.

Frequent measurements at high current levels not only impose a higher power requirement on the circuit, but also subject the battery to higher levels of energy. Such conditions can potentially contribute to heating and eventual disruptions in normal cell reactions. As confirmed by the number of past efforts to improve existing impedance measurement techniques, a new technique for measuring impedance is needed that is less sensitive to noise and EMI effects. Such a technique should also be less dependent on direct voltage measurements that are taken across the subject battery, while remaining usable for a variety of battery sizes and configurations.

Summary

In accordance with this invention, battery impedance Z_s of a battery is measured by a circuit, such as a current divider network which is connected to the battery. The circuit has a current generator producing a current signal Ii and has one or more sensing impedances Z_s which are normally positioned electrically parallel, or in some alternate embodiments in series, with the battery. A DC-blocking capacitor is positioned in series with the battery to prevent the battery voltage from draining into the one or more sensing impedances Z_s. A magnetic field sensor or comparable device for measuring the electromagnetic field produced by the current flowing through a wire, is then used to measure the current I_s passing through the sensing impedances Z_s. Either the magnitude or the phase, or both, of I_s can be measured to arrive at a usable value.

Once the value of I_s has been determined, the value of the impedance is mathematically determined. This can be done, for example, by substituting a number of calibrated impedances having values Z_oaιι through Z_cai_N, into the circuit in place of the battery and its impedance value Zb. In one illustrative method, for example, the mathematical determination includes accounting for the effects of the combined circuit and connector impedances Zl and Z2 which lead, respectively, into the parallel sensing and battery impedances Z_s and Zb. An additional accounting is made for the effect of the combined circuit and connector impedance Z3 leading out of the sensing impedance Z_s and the combined circuit, connector and blocking capacitor impedance Z4 leading out of the battery impedance Zb. This results in the relationship:

Z_b = I' (Zs + Zl + Z3) + (Z2 ₊ ZA) where r = is

Ii - Is, so that:

Z_b = I_s (Zs + Zl + Z3) + (Z2 ₊Z4)

where the values of Zl, Z2, Z3, and Z4 may be unknown. In this example of the technique, two calibrated impedances Z_caιι and Z_caι₂ are substituted into the circuit for the battery impedance Zb where:

Z_x ≡ Z3 + Zl + Z3 and Z_y ≡ Z2₊Z4 so that

Z_b = I' Z_x + Z_y

resulting in an I_s value of Ii when Z_cau is substituted for Z_b and resulting in an I_s value of I₂ when Z_caι is substituted for Z_b so that

■ χ = call " _cal Il - l2 , and so that

Zy = 1^~2 Zcall - IlZ_Cal2 Il - Ϊ2 with the final step in this example being the simple determination of the value of Z_b by substitution into the equation

Zb = r Z_X + z_y.

While the illustrative embodiment of this invention utilizes two calibrated impedances Z_CaiN, it will be appreciated that a larger number of calibrated values may be used to achieve the necessary measurement of Z_b as will be explained further in the following Detailed Description of the Preferred Embodiments. It will also be appreciated that, as reflected in the series embodiment of the incorporated circuit described below as an alternative embodiment of the invention, various circuit configurations are also possible in carrying out the disclosed impedance measuring technique and are fully contemplated as being within the scope of this invention. For example, the use of additional, parallel sensing impedances Z_s may be preferred and are contemplated.

Thus, the invention does not reside in any one of the features of the impedance measuring apparatus and method which is disclosed above and in the Detailed Description of the Preferred Embodiments and claimed below. Rather, this invention is distinguished from the prior art by its particular combination of features disclosed. Important features of this invention have been disclosed in the Detailed Description of the Preferred Embodiments of this invention which are shown and described below, to illustrate the best mode contemplated to date of carrying out this invention.

Those skilled in the art will realize that this invention is capable of embodiments which are different from those shown, and the details of the structure of the impedance measuring apparatuses and the details of the disclosed impedance measuring methods can be changed in various manners without departing from the scope of this invention. Accordingly, the drawings and description are to be regarded as illustrative in nature and are not to restrict the scope of this invention. Thus, the claims are to be regarded as including such equivalent apparatuses and methods as do not depart from the spirit and scope of this invention.

Brief Description of the Drawings

For a more complete understanding and appreciation of this invention and many of its advantages, reference will be made to the following, detailed description taken in conjunction with the accompanying drawings wherein:

FIG. 1 depicts a typical bridge circuit configuration of the type commonly used in one impedance measuring technique of the prior art;

FIG. 2 depicts a typical voltage divider configuration of the type commonly used in an additional impedance measuring technique of the prior art;

FIG. 3 represents a basic 4-wire configuration of the type used in the prior art;

FIG. 4 depicts a typical a short circuit configuration used in the prior art;

FIG. 5 is a prior art time constant configuration circuit;

FIG. 6 is a general form of a proposed circuit which is according to the principles of this invention;

FIG. 7 is a general form of a proposed circuit with connector, wire, and capacitor impedance values denoted before and after both the sensing impedance and the battery impedances;

FIG. 8 represents a magnetic interface for sensing current I_s according to the principles of this invention;

FIG. 9 is an alternative series embodiment of a proposed circuit according to the principles of this invention;

FIG. 10 is an electrical schematic of a prototype circuit implementing the invention concept; and FIG. 11 is a graphical representation of impedance of a nickel cadmium battery during discharge. Detailed Description of the Preferred Embodiments

Referring to the drawings, identical reference numbers and letters designate the same or corresponding parts throughout the several figures shown in the drawings.

The proposed technique for making impedance measurements deals directly with the limitations of prior art measuring techniques. As noted in the Background above, most previous impedance measurement techniques rely on voltage measurements made across the battery and/or sensing impedance. These measurements are in the order of millivolts, driven at amps of current through milliohms of impedance. Thus, it is very difficult to measure such current levels by merely measuring the voltage across a sensing impedance Z_s. For example, as noted in the Background, a simple voltage measuring device is highly susceptible to the effects of noise, EMI, and the current drain of the battery itself. However, an alternate way of measuring a current I in given leg of an electrical circuit is to measure the magnetic field generated at a sensing branch of the circuit. Such a current I can be appropriately measured either by its magnitude or change in its phase angle or both.

FIG. 6 depicts a general form of a circuit which may be used to calculate a battery's impedance according to the proposed technique. The circuit has the general construction of a current divider network. C represents the capacitance of a DC-blocking capacitor 51 that is selected large enough for the AC range of interest. Selection of an appropriate size of capacitor

51 is dependent upon the lowest frequency level, or "half power frequency," which the capacitor allows to pass through it, determined by the formula

C = 1

2π Z_s 4 where C is the capacitor's capacitance value and f_h is the half power frequency, and Z_s is the magnitude of this impedance and does not include its phase angle. Without the DC-blocking capacitor 51, current from the battery 48 would drain into the sensing impedance Z_s. Ideally, Z_s is kept as close as possible to the anticipated impedance of the battery to be measured.

Zb represents the battery impedance that is being measured by the circuit, I; represents the input current signal and L. represents the sensing impedance current. Additionally, lb represents the AC current through the battery 48, though the actual current in the cell may contain a DC component if a load is connected to the battery 48. In this embodiment of the proposed circuit, current Ij travels from a current source 53 through a current signal path 50 to a current dividing connector 52 where it splits into portion I_s, flowing through sensing current path 54, and into portion lb, flowing through battery current path 56. The sensing current and battery current paths 54 and 56 eventually re-converge at current converging connector 58. If measurements are done online, the AC current signal can be choked with an appropriate AC choke 60, if necessary, so the load impedance, Z is much larger than Z_s and Zb.

To determine the battery impedance, Zb, requires recognition of the basic relations

I_; = I_s +I_b and

Neither the battery voltage Vb nor the sensing impedance voltage V_s are measured directly. However, since the I_s and lb branches are parallel, v_b =v_s =ι_sz_s then

When battery sizes become very large, cell internal impedance becomes very small, so that impedance contributions from connectors and wires cannot be ignored. Thus, the circuit must be remodeled to include these additional impedance elements. Each of these particular elements is included in FIG. 7, with the load being omitted. In this particular embodiment of the circuit, the impedance resulting from capacitance C of DC-blocking capacitor 51 is lumped with the impedance of the connector and local segment of the battery current path 56 as Z4. Zl, Z2, and Z3 represent the connector and wire impedances of sensing element 62 and battery 48. Z_s is a known value, and as noted above, is ideally kept as close as possible to the impedance of the battery to be measured. However, unlike Z_s, the values of Zl, Z2, Z3, and Z4 are not known explicitly. For this circuit, it is known that

v_s =v_b which can be written as

I_s(Z₁ +Z_s +Z₃)= I_b(Z₂ +Z_b +Z₄) .

The definitions

Z_x — Z_s +Z_j +Z₃ and Z_γ ≡ Z₂ +Z₄

allow for the relation

Further, the definition ϊ' ≡ -

I. - I. allows for the relation

Z_b - I z_x +Z_Y .

For this equation, I' is known and measured. However, Zx and Zy are not precisely known, and may not be known at all. Z_b is the value that must be determined.

As an example of this technique, the values of Z and Zy can be determined experimentally using two different calibrated impedances, Z_caπ and Z_caι₂. The example procedure involves removing the battery 48 from the circuit and replacing it with Z_calι first and then Z_cai2. It will be appreciated, however, that removal of the battery 48 may not always be necessary in order to complete the required measurements according to the invention, for example if the values of Zl, Z2, Z3 and Z4 are small in relation to Zb and Z_s, or if the values of Zl, Z2, Z3 and Z4 are already known. In this example, a measurement of I_s is made for each of the two calibrated impedances Z_caιι and Z_caι . This results in two equations derived from the previous equation for Zb,

Z_Call ⁼ I_lZχ +Z_γ

Solving for Zx and Zy leaves the relations,

Once Zx and Zy are determined, the battery 48 can be placed back into the circuit and the previous equation

Z_b = 1 Z_x + Z_Y

can be used to later determine Zb after a value for I_s is determined.

It will be appreciated that, while the impedance determination of this example is made using two impedance values, it is also possible to use more than two calibration values, and this possibility is fully contemplated to be within the scope of the invention. For example, given the relation

_b ⁼ ι z_x + Z_Y,

in which Z_x and Z_y are not precisely known, a number N of known calibration impedances Z_oaιι through ZoaiN can be substituted for Z_b in the circuit, allowing for a measurement of I_s to be made for each of the calibration impedances leaving N separate relations

Zcall ⁼ IlZχ + Zy

ZoaiN — INZ_X + Z_y, which, depending on the range of the calibration impedances, can be solved linearly, piecewise- linearly, or nonlinearly for Z_x and Z_y. It should be noted that, unlike the prior art techniques discussed above, the voltage drops across Z_s and Zb are never measured directly. Once the circuit is calibrated for Z_x and Z_y, the battery 48 can be placed back into the circuit for the determination of Z_b from the measurement of I_s.

Detection of I_s is accomplished by magnetically coupling I_s with a magnetic field sensor 64 such as a Hall effect or a magnetoresistive sensor or any other device which can determine the magnitude and phase of a magnetic field. This is possible if Z_s is an electrical conductor such as copper. An appropriate magnetic interface 66 is depicted in FIG. 8. The figure shows how the magnetic interface links the sensing current, I_s to the magnetic sensor 64. The interface includes a ferromagnetic core 68 which is coupled to the magnetic field sensor 64 and which need not be wound. The sensing current I_s travels along the sensing impedance Z_s throughout the ferromagnetic core 68 resulting in a magnetic flux 70. The size and shape of the conductor Z_s should be selected to maximize sensitivity of the current change when the battery impedance Zb changes and maximizes the flux linkage to the magnetic circuit.

Zs represents a single turn winding on the magnetic core 68 and the magnetic sensor 64 sits in the air gap that dissects the core path. The magnetic flux density that the magnetic field sensor 64 is exposed to is given by the equation

This equation assumes the cross sectional area of the air gap is the same as the core. B_gap is the

magnetic flux density of the air gap and £_gΑp is the effective gap length. Also, ^₀or_e is the

effective core flux path length, μ_core is the core permeability factor and μ₀ is the permeability of

free space. It should be noted that this equation was derived under static assumptions and nonlinear and dynamic properties such as hysteresis and core saturation are considered negligible factors for purposes of this invention description. The equation shows that the air gap length is the dominant factor for the sensitivity of the flux density to the sensing current I_s. If a large enough current is used and if the magnetic field sensor 64 is sufficiently sensitive, a core 68 may not be necessary.

The choice of magnetic field sensor 64 must include considerations such as the sensor's ability for mounting in the core path and the ability to provide sufficient sensitivity to detect the sensing current I_s. Some suitable types that have been successfully implemented include Hall effect and anisotopic magnetoresistive (AMR) sensors which are readily available. AMR sensors, such as the Honeywell HMCIOOI, have demonstrated greater levels of sensitivity than Hall effect sensors, such as the Optek OHN-xx, for operation in small magnetic fields. AMR sensors also offer a much wider bandwidth at approximately the same cost as Hall effect sensors. Other alternative sensor types may present problems due to cost, bandwidth, and size. Some alternatives, such as the use of secondary windings as the magnetic sensor 64, present an additional limitation in that they tend to introduce impedance into the circuit and constrain the bandwidth. However, it is contemplated that additional, satisfactory magnetic field sensors will be developed over the years for use according to this invention.

Advantages of the invention over previous impedance measuring techniques include greater sensitivity and greater immunity to noise and EMI. In the proposed circuit, sensitivity is controlled mainly by the selection of the sensing impedance and the gap size of the core 68. The fact that no potential measurements are taken across low voltage and impedance sources results in greater immunity to noise and EMI. Most previous methods require that measurements of voltage drops be made across the battery 48 and/or sensing impedance. In comparison, the proposed technique requires only that a current measurement be made.

For prior art methods of measuring battery impedance, such as the 4-wire circuit, the voltage drop across the battery will become progressively smaller as cell sizes increase. This results in the circuit becoming increasingly less sensitive as the magnitude of each measurement falls. It follows that the change in sensitivity is dependent on the internal impedance of the battery being monitored. Since the technique proposed by this invention measures only current with proper selection of a sensing impedance, sensitivity becomes independent of the measured internal impedance. Thus, given the proper selection of the sensing impedance, circuit sensitivity using the technique of this invention will be approximately the same in both large and small capacity cells. In addition, the overall measurement sensitivity of the disclosed technique is superior to that of previous measurement techniques. When, by way of example, an AMR sensor is used in conjunction with the technique of this invention, sensitivity has been shown experimentally to improve 26 times over the level of a prior art 4-wire circuit. Reducing the core air gap size may further increase sensitivity of this technique.

A further advantage of the invention relates to inherent noise and EMI limitations of previous techniques, such as the 4-wire circuit. In such previous techniques, sensing amplifiers are required to amplify signals in the range of millivolts from a low impedance source, the battery, or sensing impedance. Such amplification typically requires the use of transistor amplifiers, such devices being highly susceptible to electric field noise sources when the input source impedance is low.

In the proposed circuit configuration, sufficiently high current, typically in the range of milliamps or greater, is sensed by the magnetic interface 66 and is more immune to noise and EMI than voltage gain amplifiers would be. The sensing amplifier for this circuit is connected to the magnetic field sensor 64, which, as noted above, has much higher impedance and voltage levels due to the improved sensitivity. For example, a Honeywell HMCIOOI has a source impedance of 850Ω while, with other AMR sensors typically having source impedances in the range of hundreds to thousands of ohms. It will be further appreciated that alternate forms of the disclosed circuit may be implemented with the proposed technique and are contemplated to be within the scope of this invention. One such alternate form of the proposed circuit is depicted in FIG. 9 and can be a voltage driven circuit having a voltage source 59 and also having a magnetic coupling in series with the battery 48 as shown in the figure. In this embodiment, the DC-blocking capacitor 61 is positioned in series between the battery impedance Zb and sensing impedance Z_s. According to the proposed technique, an impedance measurement of the battery 48 would be accomplished first by noting that

V = v_b + v_s and by noting that the sensing current I_s, which in this embodiment is also the battery current, can be expressed by the relation

Zb + Z_s .

It would then follow that

V = Is (Z_b + Zs) and Zb - Vi - LZ_S

Is •

Thus, implementation of this circuit embodiment requires knowledge of the source voltage V, sensing impedance Z_s, as well as measurement of the sensing current I_s. In practice, this configuration is capable of making impedance measurements that are similar to other embodiments of the invention. However, unlike other embodiments, this specific configuration of the proposed circuit requires the inclusion of a voltage source V; which must maintain potential magnitudes on the order of millivolts. Unlike other circuit configurations of this invention, this embodiment requires that for progressively larger battery cells, the voltage source Vi must become progressively smaller in order to maintain operability. Regulation of the voltage source Vi may become increasingly difficult for smaller battery impedances without an increase in the level of current delivered. FIG. 10 is a schematic illustration of an example prototype circuit 72 incorporating the invention that can be used to collect battery impedance data from lead acid, nickel cadmium and lithium battery cells. The circuit 72 is designed to be used with a PC based data acquisition (DAQ) board. The DAQ board is used to inject a waveform (V_DAQ) into the circuit at input 71 and toggle the set/reset circuit 82 of the AMR sensor with a signal (V_S/ ) at input 73. To sample the injected current waveform and sensed current waveform, the DAQ board has access to the feed current (V_INJ) and sensor response signal (V_SENS). The circuit section 74 in the upper left dashed box 75 represents the current source 76 and current divider circuit 78. This section drives the current into the battery 79 and sensing element 81. The injected waveform V_DAQ is passed to the signal amplifier Ul which in turn generates the current I;. The differential amplifier U2 is used to detect and to determine the value of Ii as it exists at the negative node of resistor R8 and enters the current signal path 50. A ferromagnetic core XI of the magnetic sensor detects the value of I_s at the sensing impedance 81.

The circuit section 80 in the upper right dashed box 83 represents the current sensing circuit that is linked to the sensing element via the ferromagnetic core XI and the magnetic sensor SI. Within the magnetic sensor SI, the combined resistances of component resistors 85 change in proportion to the magnetic field they encounter at the sensing impedance 81, with SI essentially comprising the ferromagnetic core XI. Within SI, a degaussing resistor 86 comprises a coil used for demagnetizing the component resistors 85 of the ferromagnetic core XI. In the event that an external magnetic signal interferes with the core's operation, the interference (magnetic offset) can be minimized by passing a positive voltage through the degaussing resistor 86 proximate to the component resistors 85.

The circuit section of the bottom dashed box 82 of FIG. 10 shows the set/reset circuit used for minimizing the magnetic offset in the magnetic sensor. In the circuit, a signal V_S/R at input 73 is electrically isolated from the rest of the circuit by an optical coupling U5. To minimize a magnetic offset in the sensor SI, the signal V_S/R, which is normally on the order of +5V, is manually dropped to 0V for a duration of, for example, one second. While the optical coupling U5 electrically isolates the signal V_S/R from the rest of the circuit, U5 still permits a mimicking +5V signal to pass from mimicking potential 88 through resistor R9 in response to each positive (+5V) condition for V_S/R. Mimicked signals are fed through four digital inverters U3, all of which may be contained on a single electronic chip allowing the signal to pass to a complementary MOSFET pair Q2. The MOSFET pair Q2 includes an N-channel MOSFET 90 and an E-channel MOSFET 92, the pair functioning together as a combination toggle switch for effecting positive voltage through resistor 86. All four digital inverters act in concert to alternate the positive conditions of the N- and E-channel MOSFETS in order to toggle positive voltage through the degaussing resistor 86 depending on whether V_S/R is currently in its positive +5V condition. Minimizing the magnetic offset in this way insures that the magnetic sensor does not saturate or drift from the zero field point, helping to maintain high sensor resolution. This further enables the circuit to be used to collect impedance data on batteries while charging and discharging.

FIG. 11 is a sample plot of the impedance measured for a D-size, 4.3-Ahr nickel cadmium cell during discharge. Each point in the plot represents the cell impedance Zb at different frequencies ranging from 1 Hz at the upper right end of the curve to 17.7kHz at the lower left end of the curve. The data for this plot was obtained using the prototype circuit of FIG. 10 on a nickel cadmium cell during a discharge cycle. It will be appreciated that those skilled in the art will normally test a particular battery at one or more frequencies to determine the battery's impedance Z_b at each frequency. It will be further appreciated that a subject battery may be tested at many selected frequencies to determine the condition of the battery in question, and that the invention permits those skilled in the art to perform testing across such frequency ranges as appropriate. Those skilled in the art will recognize that the various features of this invention described above can be used in various combinations with other elements without departing from the scope of the invention. Thus, the appended claims are intended to be interpreted to cover such equivalent impedance measuring techniques that do not depart from the spirit and scope of the invention.

Claims

1. An apparatus for measuring impedance in an electrochemical cell comprising: a current generator for producing an input current signal Ii along a current signal path in a current divider network; said current divider network having at least a sensing current path and a battery current path, said sensing and battery current paths being in parallel to each other, each of said sensing and battery current paths also being in series with said current signal path, said current signal, sensing, and battery paths interconnecting with each other at a current dividing connector and at a current converging connector; said sensing current path having a measurable sensing impedance Z_s located along said sensing current path, a connector-wire impedance Zi between said current dividing connector and said sensing impedance Z_s, and a connector-wire impedance Z₃ between said sensing impedance Z_s and said current converging connector, wherein a portion I_s of current signal Ii flows through said sensing current path; said battery current path having a battery impedance Zb located along said battery current path, a connector-wire impedance Z₂ between said current dividing connector and said battery impedance Zb, and a connector-wire-capacitor impedance Z₄ between said battery impedance Zb and said current converging connector, wherein a portion lb of current signal I; flows through said battery path; a DC-blocking capacitor in series with said battery impedance Zb and located along said battery current path to prevent said battery from discharging current through said sensing impedance Z_s; said sensing current path also having a magnetic field sensor with a coupling attached thereto for measuring the magnitude and phase of I_s, thereby permitting mathematical determination of the value of Zb.

2. The apparatus for measuring impedance in an electrochemical cell of Claim 1 in which the magnetic field sensor used for measuring the magnitude and phase of I_s is a magnetoresistive sensor magnetically coupled to the sensing current path.

3. The apparatus for measuring impedance in an electrochemical cell of Claim 1 in which the magnetic field sensor used for measuring the magnitude and phase of I_sis a Hall effect sensor magnetically coupled to the sensing current path.

4. The apparatus for measuring impedance in an electrochemical cell of Claim 1 in which the magnetic field sensor used for measuring the magnitude and phase of I_s is a magnetic resistive sensor magnetically coupled to the sensing current path; said apparatus being further configured to experimentally determine Zb by measuring the variable I_s when at least two different calibrated impedances Z_oaιι through Z_caiN are substituted for Zb into the circuit of said apparatus.

5. The apparatus for measuring impedance in an electrochemical cell of Claim 1, said apparatus being further configured to experimentally determine Zb from the measurement of the variable I_s when at least two different calibrated impedances Z_o ιι through Z_caiN are substituted into said current divider network of said apparatus.

6. An apparatus for measuring impedance in an electrochemical cell comprising: a current generator for producing an input current signal Ii along a current signal path in a current divider network; said current divider network having a sensing current path and a battery current path, said sensing and battery current paths being in parallel to each other, each of said sensing and battery current paths also being in series with said current signal path, said current signal, sensing, and battery paths interconnecting with each other at a current dividing connector and at a current converging connector; said sensing current path having a measurable sensing impedance Z_s located along said sensing current path, a connector-wire impedance Zi between said current dividing connector and said sensing impedance Z_s, and a connector-wire impedance Z₃ between said sensing impedance Z_s and said current converging connector, wherein a portion I_s of current signal I; flows through said sensing current path; said battery current path having a battery impedance Zb located along said battery current path, a connector-wire impedance Z₂ between said current dividing connector and said battery impedance Zb, and a connector-wire impedance Z₄ between said battery impedance Zb and said current converging connector, wherein a portion lb of current signal I; flows through said battery path; a DC-blocking capacitor in series with said battery impedance Zb and located along said battery current path to prevent said battery from discharging current through said sensing impedance Z_s; said sensing current path also having a magnetic field sensor attached with a coupling thereto for measuring the magnitude and phase of I_s and for determining the value of Zb using the equation z_b = r (Zs + Z_! + z₃) + (z_{2 +}z₄) where r = _i

Ii - ι_s, so that: Z_b = Is (Zs + Zi + Z₃) + (Z_{2 +} Z₄). Ii - Is

7. The apparatus for measuring impedance in an electrochemical cell of Claim 6 further comprising an AC-choked load impedance Z located along said battery current path in electrical parallel to said battery impedance Z , said load impedance Z also being located between said connector-wire impedance Z₂ and said connector-wire impedance Z₄ for permitting online measurements of Zb, Z being much larger than Z_s and Zb.

8. The apparatus for measuring impedance in an electrochemical cell of Claim 6 in which said measuring instrument for measuring the magnitude of I_s includes a magnetic core having a magnetic core path and an air gap that dissects said core path, said sensing impedance Z_s thereby representing a single turn winding on said magnetic core, and a magnetic field sensor being positioned in said air gap, the magnetic flux density to which said magnetic field sensor is exposed is given by the equation

where B_gap is the magnetic flux density of the air gap, ^_gap is the effective gap length, £_core is the

effective core flux path length, μ_core is the core permeability factor, and μ₀ is the permeability

of free space.

9. The apparatus for measuring impedance in an electrochemical cell of Claim 6, said apparatus being configured to experimentally determine Z by measuring the variable I_s when at least two different calibrated impedances Z_caιι and Z_caι₂ are substituted for Zb using the equation

Z_b = I' (Z_s + Z₁ + Z₃) + (Z_{2 +}Z₄) wherein Z_x ≡ I' Z₃ + Z_\ + Z₃ and Z_y ≡ Z₂₊Z

so that

resulting in an I_s value of Ii when Z_caπ is substituted for Z_b and resulting in an I_s value of I₂ when Z_cal2 is substituted for Z_b so that

and so that

Zy = I2 Zcall - I_Zcal2 Il - l2

making the value of Zb solvable by substitution into the equation

Z_b = r z_x + z_y.

10. The apparatus for measuring impedance in an electrochemical cell of Claim 6, said apparatus being configured to experimentally determine Zb by measuring the variable I_s when multiple calibrated impedances Z_cau through Z_calN are substituted for Zb using the equation

Zb = I' (Z_s + Z₁ + Z₃) + (Z_{2 +}Z₄) wherein Z_x ≡ Z_s + Zi + Z₃ and Z_y ≡ Z₂₊Z so that

Zb = I' Z_x + Z_y resulting in an I_s value of Ii when Z_oaιι is substituted for Zb and resulting in an I_s value of I_N when Z_calN is substituted for Zb so that

Zcall = IlZ_x+ Zy, and so that

ZcalN = lNZ_x+ Zy,

making the value of Z_x through Z_N solvable and thereby making the value of Zb solvable by substitution into the equation Zb = I Z_x + Zy.

11. A method for measuring impedance in an electrochemical cell comprising: producing an input current signal I; with a current generator along a current signal path in a current divider network; establishing a sensing current path and a battery current path along said current divider network so that said sensing current and battery current paths are electrically parallel to each other and so that said parallel sensing current and battery current paths are in series with said current generator; locating a measurable sensing impedance Z_s along said sensing current path, locating a connector-wire impedance Zi between said current signal path and said sensing impedance Z_s so that a portion of current signal I; flows from said current signal path through said connector- wire impedance Zi to said sensing impedance Z_s, and locating a connector-wire impedance Z₃ between said sensing impedance Z_s and said current signal path so that current flows from said sensing impedance Z_s through said connector-wire impedance Z₃ to said current signal path; locating a battery impedance Z_b along said battery current path, locating a connector- wire impedance Z₂ between said current signal path and said battery impedance Zb so that a portion of current signal I; flows from said current signal path through said connector-wire impedance Z₂ to said battery impedance Zb, and locating a connector-wire impedance Z between said battery impedance Z_b and said current signal path so that current flows from said battery impedance Z through said connector-wire impedance Z to said current signal path; and locating a DC-blocking capacitor in series with said battery impedance Zb and located along said battery current path to prevent said battery from discharging current through said sensing impedance Z_s; instrumentally measuring the magnitude and phase of I_s with a magnetic field sensor and determining the value of Zb using the equation Z_b = I' (Z_s + Z₁ + Z₃) + (Z₂4-Z₄) where r = is Ii - I_s , so that

Z_b = s (Z_s + Zι + Zi) + (Z₂4- Zi).

Ii - Is

12. The method for measuring impedance in an electrochemical cell of Claim 11 in which the magnetic field sensor used for measuring the magnitude and phase of I_s is a Hall effect sensor magnetically coupled to the sensing current path.

13. The apparatus for measuring impedance in an electrochemical cell of Claim 11 in which the magnetic field sensor used for measuring the magnitude and phase of I_s is a magnetoresistive sensor magnetically coupled to the sensing current path.

14. The method for measuring impedance in an electrochemical cell of Claim 11 in which the value of battery impedance Z_b is experimentally determined by measuring the variable I_s when two different calibrated impedances Z_cau and Z_cai2 are substituted for Z_b using the equation

Z_b = I' (Zs + Z₁ + Z₃) + (Z_{2 +}Z₄) wherein

Z_x ≡ I' Z₃ + Zi + Z₃ and Z_y ≡ Z₂₊Z₄ so that

Z - V Z_κ + Zy, resulting in an I_s value of Ii when Z_calι is substituted for Zb and resulting in an I_s value of I₂ when Z_caι₂ is substituted for Zb so that Z_x = Z_caH - Zcal2

I1 - I2 and so that

Zy = I₂ Zcall ' IlZ_Cal2

and then the value of Zb is solved for by substituting Z_x and Z_y into the equation z = r Z_x + Zy.

15. The method of measuring impedance in an electrochemical cell of claim 11 in which the instrumental measurement of the magnitude of I_s is accomplished using a magnetic field sensor having a magnetic core that has a magnetic core path and an air gap that dissects said core path, said sensing impedance Z_s thereby representing a single turn winding on said magnetic core, a magnetic field sensor being positioned in said air gap, the magnetic flux density to which said magnetic field sensor is exposed is given by the equation

B„ m Oore T

^ coorree + μ ",core

where B_gap is the magnetic flux density of the air gap, £_gΑp is the effective gap length, £_coce is the

effective core flux path length, Uc_0re is the core permeability factor, and μ₀ is the permeability

of free space.

16. The method for measuring impedance in an electrochemical cell of Claim 11 further comprising: establishing a load impedance Z_L located along said battery current path in electrical parallel to said battery impedance Z_b, said load impedance Z also being located between said connector-wire impedance Z₂ and said connector-wire impedance Z ; and

AC-choking said load impedance ZL for permitting online measurements of Zb, Z_L being much larger than Z_s and Zb.

17. An apparatus for measuring impedance in an electrochemical cell comprising: a voltage generator for producing an input current signal l_b that flows along a current signal path through a battery having a battery impedance Z_b, said battery being located along said current signal path; said current signal path having a measurable sensing impedance Z_s located along said current signal path in series with said battery impedance Z_b; a DC-blocking capacitor in series with said battery impedance Zb located along said current signal path in series with said battery impedance Z_b and interpositioned along said path between said battery impedance Z_b and said sensing impedance Z_s to prevent said battery from discharging through said sensing impedance Z_s; said sensing impedance Z_s having a sensing current I_S; a known source voltage Vi, and a magnetic field sensor with a coupling attached thereto for measuring the magnitude and phase of the sensing current I_s using the relation

Zb + Zs

making said battery impedance Zb determinable by the relation