Patent Application
20230366942
This patent application pertains to testing and sorting batteries during and after the batteries are formed, and more particularly to improving the accuracy of measurements and reducing the time required to determine characteristics of the batteries.
Battery manufacturing is an important industry, particularly batteries based on lithium ion technology. Batteries are tested before use for quality control. Part of the battery production procedure includes final formation of and sorting of batteries. The present inventor's International Patent Application Pub. No. WO 2021/113161 A2 discloses a system for forming and testing batteries under mixed connection conditions in parallel and in series. The present inventor's U.S. Pat. No. 6,291,972 is titled “System for Battery Formation, Charging, Discharging and Equalization,” which discloses concepts for forming and testing batteries, is incorporated by reference in its entirety for all purposes. Tests on batteries after formation and for quality control and sorting include the measurement of an open-circuit voltage, internal resistance and a self-discharge rate for each battery.
Small electronic devices may use a single lithium ion battery. Electric vehicles and large-scale electric storage use many batteries connected together in a battery pack. A battery pack performs as well as the worst individual battery in the pack, so a significant effort is made to form a battery pack from batteries having consistent performance characteristics. A considerable amount of research has been conducted for how best to determine battery characteristics and then to sort batteries according to common characteristics. The battery manufacturing industry presently directly measures certain parameters and sorts batteries into groups having reasonably consistent values for the parameters. Other methods for grading and sorting batteries include modeling, algorithms and electrochemical analysis. A battery's parameters that can be measured directly include its voltage, which is typically an open-circuit voltage, its internal resistance and its self-discharge rate.
Batteries are typically sorted according to common values for open-circuit voltage, internal resistance and self-discharge rate. To determine the self-discharge rate, a battery's open-circuit voltage is measured, and then the battery is stored or aged for a period of time, which can be a period of days or weeks, after which the open-circuit voltage is measured again. The difference in voltage over the period of time is used to calculate the self-discharge rate. The difference in voltage is a small value, which is difficult to measure accurately. A longer period of time between the voltage measurements may provide a more accurate measurement of the self-discharge rate. However, there is a need to increase the production rate for making batteries and battery packs, so there is a need to improve the accuracy of the measurement of the self-discharge rate while at the same time reducing the aging period for the batteries between the measurements of open-circuit voltage before and after the aging period.
The present invention provides a difference voltmeter that comprises a reference voltage generator (RVG) configured to generate a reference voltage and output the reference voltage via its first lead; a first operational amplifier (OPA), the first OPA has a first input lead to be connected to a first terminal of an object to be measured, a second input lead connected to the first lead of the RVG, and the first OPA is configured to output via its output lead a difference voltage representing a difference between a voltage of the object and the reference voltage; and an analog-to-digital converter (ADC) configured to receive the difference voltage from the output lead of the first OPA and convert the difference voltage from analog signal to digital signal.
The difference voltmeter is used in a battery testing apparatus that comprises a charging and discharging module for providing a charging or discharging current to one or more batteries and a control unit configured to provide outputs indicating values of parameters for the one or more batteries. The control unit is preferably a microcontroller or a computing device or a computer or a personal computer. Typical parameters include an open-circuit voltage (OCV), a self-discharge (SD), a self-discharge rate (SDR) and/or an internal resistance (IR) of the batteries. The reference voltage generator is preferably configured to generate a reference voltage Vref, where the first OPA is configured to output a difference OCV of a battery at a time t1 (ΔOCV1) and at a later time t2 (ΔOCV2), where the difference OCV of a battery is the OCV of the battery minus Vref, and where the control unit is preferably configured to determine the SD for the battery as ΔOCV2−ΔOCV1 and to determine the SDR for the battery as (ΔOCV2−ΔOCV1)/(t2−t1).
The control unit is preferably configured to cause the charging and discharging module to deliver a current I1 passing through a battery and a different current I2 passing through the battery, where the control unit is configured to obtain a difference voltage ΔV1 while the current I1 is passing through the battery and a difference voltage ΔV2 while the current I2 is passing through the battery from the difference voltmeter, and where the control unit is preferably configured to determine the IR of each battery as equal to (ΔV2−ΔV1)/(I2−I1).
The present invention provides a battery testing machine in which the battery testing apparatus is used and is preferably configured to connect electrically to a battery cell; provide a reference voltage Vref; measure a ΔOCV1 at a first time t1, wherein ΔOCV1=OCV1−Vref; measure a ΔOCV2 at a second time t2, wherein ΔOCV2=OCV2−Vref; calculate the SD for the battery cell as the difference between ΔOCV1 and ΔOCV2; and calculate the SDR for the battery cell as SD divided by the difference between t2 and t1. The battery testing machine is preferably configured to: pass a first current I1 and a second current I2 through the battery cell, where I2 is not equal to I1; measure a first ΔV1=(V1−Vref) of the cell while passing the current I1 through the cell; measure a second ΔV2=(V2−Vref) of the cell while passing the current I2 through the cell; calculate a ΔV=(ΔV2−ΔV1); calculate a ΔI as I2−I1; and calculate the IR of the battery cell as equal to ΔV/ΔI.
In another embodiment, the present invention provides a method for determining the internal resistance (IR) of a battery using direct current (DCIR) that includes the steps of measuring a voltage V1 of a battery at a time t1 while passing a direct current I1 through the battery; changing the direct current passing through the battery to I2, wherein I2 is not equal to I1; measuring a voltage V2 of the battery at a time t2 while passing the direct current I2 through the battery; and using a difference voltmeter (DVM) to determine the values of V1 and V2. The values of V1 and V2 are determined using a first instrumentation amplifier (INA1); the DVM employs a reference voltage generator to generate a voltage Vref; the DVM employs a second instrumentation amplifier (INA2); V1 and Vref are input to the INA2, which provides an output of V1−Vref; V2 and Vref are input to the INA2, which provides an output of V2−Vref. Each of INA1 and INA2 have a full scale range (FSR), and the FSR of INA2 is less than the FSR of INA1, preferably less than 0.1* FSR of INA1. The IR of the battery is calculated as equal to ((V2−Vref)−(V1−Vref))/(I2−I1). The abs(Vref −V1) is preferably less than or equal to 0.1*abs(V1), and the abs(Vref−V2) is preferably less than or equal to 0.1*abs(V2).
A better understanding of the invention can be obtained when the detailed description of exemplary embodiments set forth below is considered in conjunction with the attached drawings in which:
Methods and equipment are disclosed herein that use a principle of “difference measurement and cancelation of base line” to provide a highly accurate measurement of DC voltage in comparison applications. Equipment is disclosed that provides a highly accurate measurement and comparison of OCV and DCIR during battery formation, testing and grading, especially for a battery manufacturing line. The methods and equipment are referred to as a Difference Voltage Measurement for Comparison (ΔVMFC) or ΔVMFC technology and equipment.
The self-discharge rate (SDR) and the internal resistance (IR) of a battery are important parameters for indicating the quality of a battery, especially for grading and sorting cells and for making a battery pack. Currently direct measurement of open-circuit voltage (OCV) is widely used for SDR measurement. The OCV of a battery cell is measured directly with a volt meter or a multimeter, which preferably measures with high accuracy.
The internal resistance (IR) of a battery is related to the charge/discharge current applicable to the battery. A battery with a high internal resistance will generate a higher voltage drop and will heat up more than a battery with a low IR. The ACR is used to evaluate the conductive resistance of a battery. A standard way for determining IR using a DCIR evaluation uses a step loading current test to calculate the DCIR value via the voltage difference caused by current change.
The self-discharge rate (SDR) of a battery is calculated as SDR=SD/Δt, where SD=ΔOCV(2−1)=OCV2−OCV1. The accuracy of the SDR measurement depends on the accuracy of the SD measurement and the Δt value. The accuracy of SDR can be expressed as (accuracy of SDR)=(daily accuracy of SDR)/Δt, which may be equivalently expressed as (daily accuracy of SDR)=(accuracy of SDR)*Δt=(accuracy of SD)/SD*Δt.
The prior art for measurement of SD is by directly measuring OCV of a cell over a time period Δt (aging time) and calculating SD=ΔOCV(2−1)=OCV2−OCV1. The accuracy of SD depends on the accuracy of the OCV measurement. For example, the OCV for a lithium ion battery is about 4V, and the measurement range of the voltmeter is typically about ±5V. For a multimeter with an accuracy of ±0.01% of its FSR in a ±5V range, the accuracy of SD is ±0.01%*10V =±0.001V or ±1 mV. Assuming a typical SD is 10 mV over a 10-day period, then the SDR=10 mV/10 days=1 mV/day, and the (daily accuracy of SDR)=(accuracy of SD)/SD*Δt=±1 mV/10 mV*10 day=±100%*day. The ±100%*day of “daily accuracy of SDR” means a battery cell needs to age 10 days to get accuracy of SDR to ±10% (=±100%*day/10 days). The value for ΔOCV(2−1) is relatively small compared to the value measured for OCV, which makes it difficult to determine the SDR accurately without aging a battery cell over a relatively long period of time, such as about 10 days or more.
The present invention provides a reference voltage (Vref) and employs difference measurements of ΔOCV2=OCV2−Vref and ΔOCV1=OCV1−Vref at a much smaller voltage range. The self discharge of a battery cell is then calculated as SD=ΔOCV(2−1)=OCV2−OCV1=ΔOCV2−ΔOCV1 at much higher accuracy. For a voltage measurement with an accuracy of ±0.01% FSR in ±50 mV range, the accuracy of SD is ±0.01%*100 mV=±0.01 mV. Also assuming the typical SD is 10 mV over a 10-day period, the daily accuracy of SDR=(accuracy of SD)/SD*Δt=±0.01 mV/10 mV*10 days=±1%*day. The 1%*day of the (daily accuracy of SDR) means only 1 day of aging is needed to get an accuracy of SDR of ±1% (=±1%*day/1 day). The direct measurement of OCV in the paragraph above, without a reference voltage, required 100 days of aging for this same accuracy, while the difference measurement using a reference voltage required only 1 day of aging for this same accuracy of measurement. This shows that a difference measurement of ΔOCV2 and ΔOCV1 is much more accurate than the prior art direct measurement of OCV2 and OCV1. This improved accuracy can allow a significant reduction in aging time and a reduction in space required for aging batteries, while achieving an SDR determination that is as reliable as obtained using the prior art method. Alternatively, a much more accurate determination of SDR can be achieved using the prior art aging time, which allows batteries to be sorted into groups having a more consistent value for SDR.
It is appreciated that although the INAs 14a and 34 are employed to measure the voltage of the battery cell 12a and the difference voltage between the voltage of the battery cell 12a and the reference voltage generated by the generator 32 in the embodiment shown in
Assuming that t2−t1=200 hours, then the self discharge rate of the battery cell 12a can be calculated as SD/(t2−t1)=(OCV1−OCV2)/(t2−t1)=(3.812−3.791)/200=0.00011 V/hour, or as (OCV1Δ−OCV2 Δ)/(t2−t1)=((ΔOCV1+Vref)−(ΔOCV2+Vref))/(t2−t1)=(ΔOCV1−ΔOCV2)/(t2−t1)=(0.01225+0.00865)/200=0.0001045 V/hour. In order to get an accurate enough reading of SD and SDR, the voltage drop between t2 and t1 needs to be about 10 times the accuracy of the voltage reading. This means this voltage drop needs to be greater than 10 mV if we read the OCV from lead 20a because the accuracy reading on 20a is 1 mV. This means that t2−t1 needs to be at least 10 mV/0.00011 V/hour =91 hours. If one reads the OCV Δ by reading ΔOCV on lead 42 with accuracy of 0.01 mV, the voltage drop needs only to be greater than or equal to 0.1 mV, and the value of t2−t1 needs only to be 0.1 mV/0.0001045 V/hour=0.96 hr. That is about one hour versus 91 hours. This means one can reduce the self-discharge time by about 100 times to get the same result in evaluating the self-discharge rate. Or, one can get a much more precise measurement of the self-discharge rate of each battery for better grading and sorting in a relatively short period of time. The present invention can be used to change the grading and sorting procedure for batteries, which can save or reduce storage time and storage space, while at the same time improving the quality of batteries and battery packs and while improving the battery manufacturing process.
The value for Vref for ΔOCV2 and ΔOCV1 should be the same value so that the value for Vref cancels out properly in the calculation for SD=ΔOCV(1−2)=OCV1−OCV2=ΔOCV1−ΔOCV2. For the same reason, the fluctuation/noise of Vref during measurement should be smaller than the error level of ΔOCV measurement (+0.01 mV).
In one embodiment, a rough range of the OCV of the battery cell 12a is known in a specific implementation, therefore the reference voltage Vref to be generated by the reference voltage generator 32 may be predefined, for example, the reference voltage Vref may be preconfigured as 3.80000 volt in the above example. In another embodiment, the reference voltage Vref is set based on a measured value of the voltage of the battery cell 12a. In the voltmeter 30 of
It is appreciated that the function of reference voltage generator 32 and INA 34 in
Ro=abs((Vro−V1)/(I2−I1))=abs((ΔVro−ΔV1)/(I2−I1));
Rp=abs((V2−Vro)/(I2−I1))=abs((ΔV2−ΔVro)/(I2−I1));
IR=Ro+Rp=abs((V2−V1)/(I2−I1))=abs((ΔV2−ΔV1)/(I2−I1)); and
With reference to
An IR value can be measured by a DCIR method, where IR=ΔV/ΔI=(V2−V1)/(I2−I1). The accuracy depends on the accuracy of ΔV and the value of ΔI. The prior art DCIR method is by directly measuring Vcell of the cell over the current change as ΔV/ΔI=(V2−V1)/(I2−I1).
The principle of a ΔDCIR method according to the present invention is to employ a reference voltage Vref and a difference measurement, which measures the difference of (Vcell−Vref) instead of Vcell so that IR=ΔV/ΔI=((V2−Vref)−(V1−Vref))/(I2−I1)=(ΔV2−ΔV1)/(I2−I1). The accuracy improvement is proportional to the ratio of the voltage measurement range of full Vcell over the voltage measurement range for Vcell−Vref. Vcell−Vref can be much smaller than Vcell, and the voltage range to measure ΔVcell=Vcell−Vref can be much smaller than the voltage range of Vcell. Consequently, one can determine the value for ΔDCIR much more accurately (up to 100 times higher when the ratio of the voltage measurement range of full Vcell over the voltage measurement range for Vcell−Vref is 100) with the option of using a much smaller current step (I2−I1) than used in the DCIR method, which can be as low as 1% to get the same accuracy.
The value for Vref for ΔV2 and ΔV1 should be the same so that the value cancels out properly. and IR=ΔV/ΔI=((V2−Vref)−(V1−Vref))/(I2−I1)=(ΔV2−ΔV1)/(I2−I1) is determined properly. The fluctuation/noise level of Vref should be smaller than the error level of ΔDCIR measurement. As mentioned above, the reference voltage Vref to be generated by the reference voltage generator 32 may be predefined or may be set based on a measured value of the voltage of the battery cell 12a or 12b.
The principles of the ΔVM and ΔDCIR methods include the following:
Abs(V−Vref)<<V
ACR is generally preferred for determining Ro than traditional DCIR, where the traditional DCIR directly measures V1 and Vro and determines Ro as abs((V1−Vro)/(I2−I1)). The accuracy of the value of ΔV=V1−Vro in directly measuring V1 and Vro is same as accuracy of V1 and Vro. Because the value of ΔV=V1−Vro is usually very small in comparison to the value of V1, the accuracy for determining V1−Vro is fairly poor. For example, if the accuracy for determining a value for V1 is ±0.02% FSR or ±2 mV when FSR=10V and V1−mVro is ˜10 mV, the accuracy for determining a value for V1−Vro is about ±2 mV/10 mV=±20%.
It is preferred to use a current as small as possible in IR measurement to avoid disrupting the electrochemical system of a battery. A small change in current flow results in a small change in voltage. For a small current change of I2−I1, the voltage change of V1−Vro will also be small, and the accuracy of the measurements will be poor.
ACR can usually get to ±0.5˜1% or better with relatively small current because it measures directly the voltage fluctuation of V1 instead of V1 itself ACR uses a difference mechanism for voltage measurement through the isolation of a very small AC signal from a large DC voltage baseline, which is approximately the cell voltage without an AC current passing through the cell. The ACR technology has relatively low accuracy (such as ±0.5% FSR) and low current capability due to a complicated procedure that includes isolation, chopping, rectifying, amplification and filtration to determine the AC signal. ACIR technology has an advantage over direct DCIR because it applies a much lower current.
ΔDCIR should have an advantage over ACIR, because it requires similar low current level but has much higher accuracy such as ±0.02% FSR. The ΔDCIR method may provide the same or a better accuracy level as ACR for ohmic resistance measurement. The ΔDCIR method is simpler, cheaper, has a higher current capability and much better accuracy than the ACIR method.
As an example, the following table compares the precision of Ro and Rp with different technologies, ACR, DCIR and ΔDCIR. From this example, one can see that DCIR technology may get comparable precision (0.1 mΩ) to ACR technology only at a fairly large DC current (±5 A), when ACR's AC current is within 1 A or less. But ΔDCIR technology may get 10 times better precision (0.01 mΩ) to ACR at a comparable DC current (±0.5 A).
Difference measurement of open circuit voltage and of internal resistance of a battery cell based on a reference voltage can provide much more accurate determinations for ΔOCV and ΔDCIR. A difference voltmeter ΔOCV/ΔDCIR and equipment to implement it can have 10˜100 times better accuracy compared to a voltmeter OCV/DCIR. The difference voltmeter is based on using the same Vref for a set of related voltage measurements for comparison purposes. The Vref values, as well as the error of Vref, are canceled in the comparison, and hence, even a relatively low accuracy of Vref does not affect the accuracy of the measurement for ΔV, ΔOCV and ΔDCIR. Although the accuracy of Vref is not substantially important, the stability and constancy of Vref are important. The Vref should have very low ripple and/or noise and should be fixed for a test.
For example, the SD and SDR measurements are for comparing the OCV at different points in time for grading and sorting battery cells. The Vref values are canceled in ΔOCV measurement for the SD and SDR, where SD=ΔOCV=OCV2−OCV1=ΔOCV2−ΔOCV1. The Vref values are also canceled when comparing SD and SDR of a group of cells. As another example, DCIR measurement is to compare the Vcell at different current and the Vref are canceled when employing the ΔDCIR technology disclosed herein, where ΔV/ΔI=((V2−Vref)−(V1−Vref))/(I2−I1)=(ΔV2−ΔV1)/(I2−I1).
In one embodiment, the present invention includes a battery testing module that comprises a cabinet having front, back, left and right sides and at least one pair of guide rails, wherein one of the pair of guide rails is received in the cabinet on the left side and the other is received in the cabinet on the right side, and wherein the pair of guide rails is configured to receive a battery tray that holds a plurality of batteries. The battery tray can be a same battery tray used during formation and testing of the batteries. The battery tray comprises: a printed circuit board (PCB) having opposing upper and lower sides and an edge; a plurality of battery holders on the upper side of the PCB; electronic devices or components operatively engaged with the plurality of battery holders through traces on or in the PCB for measuring parameters of the batteries during formation and testing; and an edge connector on the edge of the PCB. where the edge connector is operatively engaged with the plurality of battery holders and with the electronic devices or components through traces on or in the PCB. Each battery holder preferably has two or more upright spring fingers for holding and engaging a battery, which also function as a negative terminal, and a positive terminal for contact with one end of a battery. The edge connector preferably has first and second opposing sides, with each positive terminal connected to the first side, and each negative terminal is connected to the second side. Batteries are preferably received in the battery holders in an upright position with the lower end of the battery touching the positive terminal and with the spring fingers touching the cylindrical case of the battery, which is the negative terminal.
The battery testing module can be incorporated in a battery formation rack or in a battery aging rack or can be a standalone cabinet. The battery testing module includes a measurement module located in the cabinet near the back and adjacent to the pair of guide rails for receiving and connecting to the battery tray. The measurement module is designed and configured to include the reference voltage, the difference voltmeter and the difference DCIR described herein. The measurement module is designed and configured to determine the OCV, the IR and the SDR of each battery in the battery tray(s) using the difference measurement principles described herein. Each battery is preferably uniquely identified, such as by affixing a bar code to each battery. The measurement module is preferably designed and configured to identify groups of batteries having the same or very similar values for OCV, IR and SDR, as determined using the difference measurement principles described herein.
The present invention can be employed in and includes a battery sorting module, which may be a standalone module or which may be incorporated with the battery testing module. The battery testing module preferably includes automation such as a robotic arm. The battery sorting module uses the information determined by the battery testing module and physically removes each battery from the battery tray and places each of the batteries into bins, where one bin is designated for one group of batteries. Some batteries may be considered unfit for use and are sorted into a bin for batteries to be reworked, recycled or discarded. Other bins will receive batteries having reasonably consistent values for OCV, IR and SDR. A battery pack can be made from the batteries in one of these bins so that all of the batteries in the battery pack have reasonably consistent values for OCV, IR and SDR.
In another embodiment, the present invention provides a battery grading and sorting machine that includes the following components: a battery tray for receiving and holding a plurality of batteries, preferably a double-sided, smart battery tray such as described in International Patent Application No. PCT/US2020/062548, Publication No. WO 2021/113161 A2; a battery testing cabinet; one or more measurement modules, such as described with reference to
Embodiments of the present invention include the following.
An embodiment 1 of the present invention is a method for forming and testing a plurality of batteries. The method comprises the steps of:
An embodiment 2 of the present invention is the embodiment 1, further comprising determining the internal resistance (IR) of each battery in the set of raw batteries by:
The embodiments described herein are merely examples for the sake of clarity and are not intended to limit the scope of the present invention. Other variations or modifications may be made by those skilled in the field of the above-described technology. There is no need and no way to describe all possible implementations of the principles of the difference measurement technology described herein. Obvious changes or variations resulting therefrom are still within the scope of the invention.
This application claims priority to U.S. Provisional Patent Application Ser. Nos. 63/277,143 filed on Nov. 8, 2021, and 63/236,023 filed on Aug. 23, 2021, which are incorporated by reference. International Patent Application No. PCT/US2020/062548, Publication No. WO 2021/113161 A2, is a related application, which is incorporated by reference.
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/US2022/021643 | 3/24/2022 | WO |
| Number | Date | Country | |
|---|---|---|---|
| 63236023 | Aug 2021 | US | |
| 63277143 | Nov 2021 | US |
