Battery testing does not always require a long discharge test or a high-current load. Matrix-based Multi-Model Electrochemical Impedance Spectroscopy (EIS) evaluates a battery by comparing its measured electrical response with reference data from batteries in known performance states. The comparison can be used to assess or estimate capacity, cold-cranking ability, state of charge, and other characteristics.
The approach described by Battery University is intended for both lead-acid and lithium-ion batteries. Documentation associated with the technology describes testing in about 30 seconds, although the actual result depends on the instrument, battery condition, test configuration, and matrix used. The method is often described as non-invasive, while the current source also describes EIS Direct Drive as slightly invasive. That distinction matters: the test does not normally require opening the battery or removing electrolyte, but it does electrically stimulate the battery and measure its response.
How Matrix-Based Multi-Model EIS Battery Testing Works
A conventional load test applies a substantial current and observes voltage behavior. A capacity test discharges the battery under controlled conditions until a defined endpoint. Matrix-based EIS takes a different route. It applies a set of electrical signals, measures the battery’s response over multiple frequencies, and compares the resulting signature with a reference matrix.
The matrix acts as a set of performance anchors. It is built from batteries belonging to the same model or battery group but having different levels of capacity, aging, and operating condition. A measured battery is then positioned against those reference signatures. The output may be a broad pass/fail decision or, where the matrix is sufficiently specific and well populated, a numeric estimate such as capacity or another battery-health value.
The method can support measurements or estimates related to:
- capacity;
- cold-cranking ability, commonly expressed as CCA for starter batteries;
- state of charge;
- battery condition and degradation; and
- other characteristics represented by the selected matrix.
The result is not the same as a direct capacity measurement. A direct capacity test establishes how much charge or energy a battery delivers during a controlled discharge. Matrix-based EIS infers performance from electrical behavior and the quality of the comparison set. Its usefulness therefore depends heavily on the relevance and integrity of the matrix.
EIS Direct Drive, Frequency Scanning, and Test Range
The described Spectro technology uses EIS Direct Drive to extend testing to batteries rated up to 48 V nominal and 300 Ah. Direct Drive applies the scanning frequencies to the battery and produces the ripple voltage needed for the measurement. In the stated implementation, this avoids the need for a separate decoupling capacitor.
Removing that separate component can reduce circuit size and avoid the cooling requirements associated with handling the energy and heat that a larger interface circuit may produce. These are design benefits claimed for the technology, not universal properties of every EIS battery tester.
A typical frequency scan runs from approximately 2,000 Hz down to 4 Hz. Very large batteries may require a lower frequency range because their electrical response and the practical measurement conditions differ from those of smaller batteries. The source also reports a minimum battery-test resistance of 300 micro-ohms. A low resistance value is relevant because the instrument must distinguish small changes in the battery’s electrical response from the resistance of cables, contacts, switching components, and the measurement circuit itself.
In practical use, test range should not be confused with guaranteed diagnostic accuracy. A tester may be electrically capable of connecting to a battery of a given voltage or capacity while still requiring the correct matrix, battery temperature, state of charge, connections, and operating procedure for a meaningful result.
Matrix Integrity and the Three Reference-Matrix Types
Matrix Integrity Level (MIL) describes the quality and usefulness of the reference matrix used to interpret a test. A strong matrix contains representative batteries from the same model or battery group across a broad range of performance and degradation states. It should provide enough distinct reference points to separate healthy, aging, and weak batteries rather than merely distinguish a new battery from a completely failed one.
An useful analogy is a bridge supported by several pillars. Each pillar represents a region of battery performance. If the reference set includes only failed batteries, the matrix has little support at the healthy end. If it includes only new batteries, it cannot reliably describe intermediate degradation. More batteries within each performance region make the reference structure more articulate and can improve the resolution of the readings.
The three matrix categories are:
| Matrix type | Primary use | Typical output |
|---|---|---|
| Generic matrix | Screening a broad class of batteries, such as starter batteries | Broad pass/fail or condition classification |
| Custom matrix | Testing a selected battery group that is not well represented by a generic matrix | Improved interpretation for that group |
| Model-specific matrix | Testing a defined battery model with a sufficiently developed reference set | Numeric readings, such as capacity or another model-specific health value, where supported |
A generic matrix is useful when the objective is rapid screening across a broad family of batteries. It generally provides less precision than a matrix built for one exact model, but it can support practical service decisions.
A custom matrix is developed for a specialty battery or an uncommon group. It is more closely matched to the batteries being tested, but its reliability depends on how representative the samples are and how well the data are collected.
A model-specific matrix is the most targeted of the three. With sufficient reference data, it may provide numeric capacity or health-related readings instead of only a general pass/fail result. Such values remain estimates derived from the matrix; they should not automatically be treated as equivalent to a controlled discharge capacity test.
Starter-Battery Screening: Capacity, Cranking Performance, and Service Decisions
Generic matrices can be applied to starter-battery screening, where technicians often need to decide whether a battery can remain in service, should be monitored, or should be replaced. A pass result indicates that the measured signature falls within the acceptable region represented by the selected matrix. Depending on the test system and service policy, that may support continued use or provide time for planned service rather than immediate replacement.
A fail result indicates that the battery falls outside the matrix’s acceptable reference range. Replacement may be justified even when a separate cranking measurement still appears acceptable, particularly if the matrix indicates inadequate capacity or another important weakness.
This reflects an important difference between capacity and cranking performance. Cold-cranking ability describes the battery’s ability to deliver a high current for a short starting event under specified conditions. Capacity describes the amount of charge or energy the battery can provide over a longer discharge. A battery may still deliver enough current to start an engine while having lost a significant portion of its reserve capacity.
For that reason, strong cranking performance alone does not establish that a starter battery is healthy in every respect. Capacity loss can reduce reserve operation, increase the risk of failure after repeated starting events, and leave less margin when electrical loads are present while the engine is stopped or operating at low speed.
The source discusses comparative failure proportions attributed to capacity fade and low cold-cranking performance. However, the supplied material does not identify the underlying study population, battery types, test method, or publication date sufficiently to treat those percentages as independently verified. They should not be used as universal failure statistics. The general diagnostic point remains useful: evaluating only cranking current can miss capacity-related deterioration.
Building Custom Matrices from Representative Battery Samples
Custom matrices are intended for specialty batteries and battery groups that are not adequately represented by a generic reference. Examples may include uncommon industrial batteries, unusual form factors, or battery models with performance characteristics that differ from the broader population.
The described collection process uses a Data Acquisition Unit (DAU) as a portable field-lab instrument for scanning batteries. Retired batteries can supply samples with degraded performance, while new and lightly used batteries help represent the higher-capacity portion of the distribution. The source describes the DAU as a portable lab sent to interested parties as suitable retired batteries become available.
A useful custom matrix cannot be built from failed batteries alone. The sample pool should cover the performance range that the tester is expected to encounter:
- new or nearly new batteries;
- lightly used batteries;
- batteries with intermediate degradation; and
- retired or clearly degraded batteries.
The supplied source states that each custom matrix requires at least 12 diverse batteries. That figure should be treated as a stated technology requirement rather than a universal statistical rule for all battery-test systems. More important than the count alone is whether the samples are genuinely diverse, correctly identified, and measured under consistent conditions.
Matrix development should also account for practical data controls. Samples need traceable model and condition information, reliable electrical connections, suitable state-of-charge preparation, and consistent test procedures. Instrument calibration and repeatability checks are important because errors in the reference data can become errors in every later diagnosis. If representative batteries are unavailable, the resulting matrix may have limited coverage and should not be expected to produce precise readings outside its demonstrated range.
Cloud Matrix Libraries and Deep Battery Diagnostics
Reference matrices may be stored in a matrix library and accessed through cloud analytics where the supporting system is configured for that workflow. This creates the possibility of distributing matrices to compatible instruments and retaining test results as batteries move through service, use, and retirement.
The same approach can extend beyond a simple pass/fail screen. Custom matrices for unique battery groups are described as enabling Deep Battery Diagnostics, in which the system uses a more detailed reference model to analyze condition and performance. The technology is also described as estimating remaining useful life through indicators such as Minimum Operational Reserve Energy (MORE).
These are technology-specific diagnostic claims, not universal measurements produced by every EIS instrument. Remaining useful life is especially dependent on operating profile, temperature, charging practice, load history, maintenance, and the definition of end of service. A matrix-based estimate should therefore be validated against field outcomes and, where necessary, direct capacity or endurance testing.
Cloud tracking can still be valuable even when the result is treated as an estimate. Repeated measurements on the same battery can reveal changes in its electrical signature, help identify deterioration trends, and support maintenance planning. The quality of those trends depends on consistent test conditions and correct battery identification.
Matrix-based Multi-Model EIS is best understood as a comparison and estimation method. It can make battery screening faster and can expose condition changes that a simple voltage or cranking test may miss. It does not remove the need to understand the battery application, verify the matrix’s coverage, and use direct capacity or controlled discharge testing when a definitive capacity value is required.
References
- Battery University | BU-919: Building a Matrix to test Batteries. (n.d.). http://www.batteryuniversity.com/article/bu-919-building-a-matrix-to-test-batteries
- Battery University | Battery Test Methods. (n.d.). http://www.batteryuniversity.com/article/battery-test-methods
- [PDF] Battery Testing, Analysis and Design - Department of Energy. (n.d.). https://www.energy.gov/sites/prod/files/2014/05/f15/APR13_Energy_Storage_e_IV_Battery_Tstg_Design_2.pdf
- Battery Testing Protocol | Definition and Examples. (n.d.). https://www.pem-motion.com/battery-glossary/battery-testing-protocol
- BU-921: Testing Batteries by Multi-Model EIS. (n.d.). http://www.batteryuniversity.com/article/bu-921-testing-batteries-by-multi-model-eis
- Performance Matrix Analysis Method of Power Battery .... (n.d.). https://asmedigitalcollection.asme.org/electrochemical/article/18/2/020902/1087510/Performance-Matrix-Analysis-Method-of-Power
- Stationary battery testing. (n.d.). https://www.transcat.com/media/pdf/megger-Stationary-battery-testing-brochure.pdf
- BU-920: Matrix Library. (n.d.). http://www.batteryuniversity.com/article/bu-920-matrix-library
- How to test batteries - Electrical - Chief Delphi. (n.d.). https://www.chiefdelphi.com/t/how-to-test-batteries/474129
- Battery Performance Testing – Battery Benchmark Testing | Element. (n.d.). https://www.element.com/connected-technologies/battery-testing-services/battery-performance-testing