Testing is designed to tell us things we want to know aboutindividual cells and batteries.
Some typical questions are:
Is it fully charged ?
How much charge is left in the battery ?
Does it meet the manufacturer's specification ?
Has there been any deterioration in performance since it was new?
How long will it last ?
Do the safety devices all work ?
Does it generate interference or electrical noise ?
Is it affected by interference or electrical noise ?
The answers are not always straightforward.
Although all of the cell parameters the design engineer may wish tomeasure can be quantified by direct measurement, this is not alwaysconvenient or possible . For example the amount of charge left inthe battery, the State of Charge (SOC) can be determined by fullydischarging the battery and measuring the energy output. This takestime, it wastes energy, each test cycle shortens the battery lifeand it may not be practical if the battery is in use. It would alsobe pointless for a primary cell. For more detailed information ofhow this is done see the State of Charge page.
Similarly, the remaining life of a secondary cell can be determinedby continuously cycling it until it fails, but there's no point inknowing the cell life expectation if you have to destroy it to findout. This is known as the State of Health (SOH) of thebattery.
What is needed are simple tests or measurements which can be usedas an approximation to, or indirect measure of, the desiredparameter. For more information see the State of Health page
In all of the following tests, and testing in general, the testconditions must be specified so that repeatable results can beobtained, and meaningful comparisons can be made. This includesfactors such as method, temperature, DOD, load and duty cycle. Forinstance the cell capacity and cycle life, two key performanceindicators could vary by 50% or more depending on the temperatureand the discharge rate at which the tests were carried out. Seealso cell Performance Characteristics.
Battery specifications should always include the test conditions to avoidambiguity.
Qualification testing is designed to determine whether a cell orbattery is fit for the purpose for which it was intended before itis approved for use in the product. This is particularly importantif the cell is to be used in a "mission critical" application.These are comprehensive tests carried out initially on a smallnumber of cells including testing some of them to destruction ifnecessary. As a second stage, qualification also includes testingfinished battery packs before the product is approved for releaseto the customer. The tests are usually carried out to verify thatthe cells meet the manufacturer's specification but they could alsobe used to test the cells to arbitrary limits set by theapplications engineer to determine how long the cells survive underadverse conditions or unusual loads, to determine failure modes orsafety factors.
The battery packs should also be tested with the chargerrecommended for the application to ensure compatibility. Inparticular the potential user patterns must be evaluated to ensurethat the batteries do not become inadvertently overcharged. Seealso the section on Chargers.
Shake and Bake
Typical tests are included in the safety standards below. Theyinclude simple tests for dimensional accuracy to dynamic testing toverify that the product can survive any static and dynamicmechanical stresses to which it may be subject.
Typical tests are included in the safety standards below. They aredesigned to exercise the product through all the environmentalconditions likely to be encountered by the product during itslifetime.
The purpose of abuse testing is to verify that the battery is not adanger to the user or to itself either by accidental or deliberateabuse under any conceivable conditions of use. Designing foolproofbatteries is ever more difficult because as we know, fools are soingenious.
Abuse testing (always interesting to witness) is usually specifiedas part of the Safety Testing (below). Recent accidents withLithium cells have highlighted the potential dangers and stricterbattery design rules and a wider range of tests are being appliedas well as stricter Transport Regulations for shipping the products.
Consumer products normally have to comply with national orinternational Safety Standards required by the safety organisations of the countries in which theproducts are sold. Examples are UL, ANSI, CSA and IECstandards.
Strength, rigidity and flammability
Mould stress (Temperature)
Electrolyte not under pressure
No explosion or fire risk
Protection from or tolerance to
Power output - Load test
Instructions for use
Nail penetration tests
Exposure to fire
The published safety standards specify the method of testing andthe limits with which the product must comply.
Cells used in military applications usually have to meet morestringent requirements than those used in consumer products.
This is perhaps the most important of the qualification tests.Cells are subjected to repeated charge - discharge cycles to verifythat the cells meet or exceed the manufacturer's claimed cyclelife. Cycle life is usually defined as the number of charge -discharge cycles a battery can perform before its nominal capacityfalls below 80% of its initial rated capacity. These tests areneeded to verify that the battery performance is in line with theend product reliability and lifetime expectations and will notresult in excessive guarantee or warranty claims.
Temperature, charge/discharge rates and the Depth of Discharge eachhave a major influence on the cycle life of the cells (See the pageon Cycle Life) Depending on the purpose of the tests, the temperature and theDOD should be controlled at an agreed reference level in order tohave repeatable results which can be compared with a standard.Alternatively the tests can be used to simulate operatingconditions in which the temperature is allowed to rise, or the DODrestricted, to determine how the cycle life will beaffected.
Similarly cycle life is affected by over charging and overdischarging and it is vital to set the correct voltage and currentlimits if the manufacturer's specification is to beverified.
Cycle testing is usually carried out banks of cells using multichannel testers which can create different charge and dischargeprofiles including pulsed inputs and loads. At the same timevarious cell performance parameters such as temperature, capacity,impedance, power output and discharge time can be monitored andrecorded. Typically it takes about 5 hours for a controlled fullcharge discharge cycle. This means testing to 1000 cycles will take208 days assuming working 7 days per week 24 hours per day. Thus ittakes a long time to verify the effect of any ongoing improvementsmade to the cells. Because the ageing process is continuous andfairly linear, it is possible to predict the lifetime of a cellfrom a smaller number of cycles. However to prove it conclusivelyin order to guarantee a product lifetime would require a largenumber of cells and a long time. For high power batteries thiscould be very expensive.
Load testing is used to verify that the battery can deliver itsspecified power when needed.
The load is usually designed to be representative of the expectedconditions in which the battery may be used. It may be a constantload at the C rate or pulsed loads at higher current rates or inthe case of automotive batteries, the load may be designed tosimulate a typical driving pattern. Low power testing is usuallycarried out with resistive loads. For very high power testing withvariable loads other techniques may be required. A Ward-Leonardcontroller may be used to provide the variable load profile withthe battery power being returned to the mains supply rather thanbeing dissipated in a load.
Note that the battery may appear to have a greater capacity when itis discharged intermittently than it may have when it is dischargedcontinuously. This is because the battery is able to recover duringthe idle periods between heavy intermittent current drains. Thustesting a battery capacity with a continuous high current drainwill not necessarily give results which represent the capacityachievable with the actual usage profile.
Load testing is often required to be carried out with variable loadlevels. This may simply be pulsed loads or it could be more complexhigh power load profiles such as those required for electricvehicle batteries. Standard load profiles such as the Federal UrbanDriving Schedule (FUDS) and the Dynamic Stress Test (DST) specifiedby the United Sates Advanced Battery Consortium (USABC), in theUSA, and the United Nations Economic Commission for Europespecification (ECE-15) and the Extra Urban Driving Cycle (EUDC) inEurope have been developed to simulate driving conditions andseveral manufacturers have incorporated these profiles into theirtest equipment.
ECE-15 Simulated Driving Cycle
While these standard usage cycles have been developed to provide abasis for comparison, it should be noted that the typical userdoesn't necessarily drive according to these cycles and is likelyto accelerate at least twice as fast as the allowed for in thestandards.
Battery thermal management is critical for high-power battery packs.Obtaining accurate heat generation data from battery modules isessential for designing battery thermal management systems. Acalorimeter is used to quantify the total amount of heat generatedby the battery while it is cycled through its charge/dischargecycles. This is essentially an insulated box into which the batteryis placed which captures and measures the heat generated thebattery during cycling. The system is calibrated by comparing theheat generated by the battery with the heat generated by a knownheat source.
Thermal imaging is used to check for "hot spots" which wouldindicate points of high thermal stress in the cell or the batterypack. It is a photographic technique which records the intensity ofthe infra red radiation emitted by a subject using a specialcamera. The image on the left is of a lithium ion pouch cell aftera prolonged discharge at 4C. In this case the temperature is evenlydistributed within the cell and the cell terminals are runningcool. These tests can help to identify problems such asoverheating, inadequate heat sinking or air flow, undersizedcurrent conductors and interference from neighbouring cells ordevices. The images can also be used to determine the best locationfor the temperature sensors used in protection circuits.
Electromagnetic Compatibility (EMC) testing
Electromagnetic compatibility (EMC) is the ability of electronicand electrical equipment and systems to operate without adverselyaffecting other electrical or electronic equipment OR beingaffected by other sources of interference such as power linetransients, radio frequency (RF) signals, digital pulses,electrical machinery, lightning, or other influences.
Note that EMC concerns both the emission of electromagneticinterference (EMI or radio frequency interference RFI) by a productor device and the product's susceptibility to EMI emitted fromother sources. The interference may be conducted through power orsignal cables or the chassis of the equipment, it may be propagatedthrough inductive or capacitive coupling or it may be radiatedthrough the atmosphere.
Just because batteries are DC devices we can not assume that theyare immune from EMC problems. At MPower we have seen the batteryprotection circuitry in a two way radio disabled by RF interferencefrom the handset's transmitter. Similar problems are possible inautomotive applications where the power cabling is notoriouslynoisy due to interference from the ignition systems and transientsfrom electric motors and switches. While the battery itself may notemit RF interference, the same can not be said of the charger. Manychargers use switch mode regulators which are also notorious foremitting electrical noise. Radiated EMI can be critical to suchapplications as heart pacemakers, medical instrumentation,communications equipment and military applications.
As with many problems prevention is better than cure and it is wiseto start considering EMC at the earliest possible stage of thedesign to avoid costly design changes when the project is submittedfor final approval. This may involve system design choices such asoperating frequencies, circuit layouts and enclosure design and theavoidance of designs with high transient currents.
Various techniques are used to minimise the effects EMI. Sensitiveparts of the circuit may be physically separated from sources ofinterference, the equipment may be enclosed in a sealed metal box,individual parts of the circuits may be shielded with metal foil,filters can be added to cables to filter out the noise,
EMC testing involves specialised test equipment and facilities.Testing must be carried out in an environment free from othersources of EMI. This usually means an anechoic chamber or a Faradaycage. Special wide range signal sources and sensitive receivers areneeded to generate and measure the interference.
Some examples of EMC requirements are give in the section onStandards
Conducting a process audit of the cell manufacturer's productionfacilities is further way of gaining confidence in the cells underconsideration however this option is usually available only tomajor purchasers of high volume or high cost cells. Unless you areone of these you will have to rely on your friendly pack maker whopossibly qualifies for special treatment.
The process audit involves verifying that the cell maker hasappropriate quality systems in place and that these are being fullyimplemented at every stage of the manufacturing process. To beeffective this task needs to be conducted by a team with specialistindustry knowledge. Again this is a job best left to your packmaker who should have the necessary experience and credibility withthe cell makers.
Inspection and Production Testing
The purpose of inspection production testing is to verify that thecells which have been purchased and the products built with themconform to agreed specifications. These tend to be short testscarried out on 100% of the throughput or on representative samples.The composition of the materials from which the components are madeshould not be overlooked. We have seen examples of unscrupuloussuppliers plating connectors with a gold coloured alloy rather thanthe gold specified and using cheap plastics which buckle in theheat rather than the high quality plastics required.
Typical tests include both mechanical and electrical tests. Thecomponents are checked for dimensional accuracy and samplesubassemblies are subject to weld strength testing of theinterconnections. Electrical parameters measured include theinternal impedance and the output voltage of the cell or batterypack with or without a load. The battery is also submitted to shortduration charging and discharging pulses of about 2 milliseconds tocheck that the unit accepts and can deliver charge.
Battery packs are normally subjected to more comprehensive testing toensure that the electronics are functioning correctly. Theprotection circuit is checked by applying a short circuit acrossthe battery terminals for 1 or 2 seconds and checking that thecurrent path is cut within the prescribed period and that thebattery recovers afterwards. The output of the fuel gauge ischecked and if the battery has built in memory, the data such ascell chemistry code, date and serial number are read out andrecorded to permit traceability.
Charge conditioning orFormation
This is normally carried out by the cell manufacturer but in somecircumstances it could be the responsibility of the battery packassembler. In any case the cells must be tested to ensure that theyare ready to deliver current.
Performance monitoring is used to verify whether the cell iscontinuing to perform as required once it is in use in theapplication for which it was specified. These are individual testsspecified by t he user.
There are no simple direct measurements, such as placing avoltmeter across the terminals, to determine the condition of thebattery. The voltmeter reading may tell us something about thestate of charge (with an enormous margin of error), but it cannottell us how well the battery will deliver current whendemanded.
It is necessary to know the internal resistance of the cell inorder to calculate the Joule heat generation or I2Rpower loss in the cell, however a simple measurement with anohmmeter is not possible because the current generated by the cellitself interferes with the measurement.
To determine the internal resistance, first it is necessary tomeasure the open circuit voltage of the cell. Then a load should beconnected across the cell causing a current to flow. This willreduce the cell voltage due to the IR voltage drop across the cellwhich corresponds to the cell's internal resistance. The cellvoltage should then be measured again when the current is flowing.The resistance is calculated by ohms law from the voltagedifference between the two measurements and the current which isflowing through the cell.
Open Circuit Voltage OCV
Measuring a battery's open circuit voltage is not a reliablemeasure of its ability to deliver current. As a battery ages, itsinternal resistance builds up. This will reduce the battery'sability to accept and to hold charge, but the open circuit voltagewill still appear normal despite the reduced capacity of thebattery. Comparing the actual internal resistance with theresistance of a new battery will provide an indication of anydeterioration in battery performance.
State Of Charge (SOC)
For many applications the user needs to know how much energy isleft in a battery. The SOC is also a fundamental parameter whichmust be monitored and controlled in Battery Management Systems. The methods of estimating the SOC are explained in the sectionon State Of Charge.
State Of Health (SOH)
The State of Health is a measure of a battery's ability to deliverthe specified current when called upon to do so. It is an importantfactor for monitoring battery performance once it has entered intouse. This is treated briefly in the section below and more fully inthe section on State Of Health.
Impedance and Conductance Testing
The discussion about the battery equivalent circuit in the sectionon Performance Characteristics shows that we can expect the battery impedance to increase withage.
Battery manufacturers have their own definitions and conventions forImpedance and Conductance based on the test method used. Though notstrictly correct they serve their purpose.
The test method involves applying a small AC voltage "E" of knownfrequency and amplitude across the cell and measuring the in phaseAC current "I" that flows in response to it.
The Impedance "Z " is calculated by Ohm's Law to be Z=E/I
The Conductance "C" is similarly calculated as C=I/E (thereciprocal of the impedance)
Note that the impedance increases as the battery deteriorates whilethe conductance decreases. Thus C correlates directly with thebattery's ability to produce current whereas Z gives an inversecorrelation. The conductance of the cell therefore provides anindirect approximation to the State of Health of the cell. Thismeasurement can be refined by taking other factors into account.These are outlined in the page about State of Health.
In addition to impedance and conductance these tests will obviouslydetect cell defects such as shorts, and open circuits.
These test methods can be used with different cell chemistrieshowever different calibration factors must be built into the testequipment to take into account differences in the aging profiles ofthe different chemistries.
Impedance and conductance testing are reliable, safe, accurate,fast and they don't affect the battery performance. They can becarried out while the battery is in use or they can be used tocontinuously monitor the battery performance, avoiding the need forload testing or discharge testing.
Note that DC measurements do not recognise capacitance changes andtherefore measurements of the internal resistance of the cell donot correlate so well with the SOH of the cell.
Using a conventional ohmmeter for measuring the resistance of thecables, contacts and inter-cell links is not satisfactory becausethe resistance is very low and the resistance of the instrumentleads and the contacts causes significant errors. More accuracy canbe achieved by using a Kelvin Bridge which separates the voltagemeasuring leads from the current source leads and thus avoids theerror caused by the volt drop along the current source leads. Seealso charger voltage sensing.
Battery analysers are designed to provide an quick indication of theState of Health (SOH) of the battery. Some analysers also have the dual function ofreconditioning the battery.
There are no industry standards for this equipment, mainly becausethere is no standard definition of State of Health. Each equipmentmanufacturer has their own favourite way defining and measuring it,from a simple conductance measurement to a weighted average ofseveral measured parameters and the test equipment is designed toprovide the corresponding answer. This should not be a problem ifthe same equipment is used consistently, however it does causeproblems if equipment from different manufacturers is used to carryout the tests.
Cell failure analysis is best carried out by the cellmanufacturers. Only they will have the detailed specifications ofthe cell mechanical and chemical components and it normallyrequires access to expensive analytical equipment such as electronmicroscopes and mass spectrometers which they should be expected tohave. More information see Why Batteries Fail and Lithium Battery Failures