As an Amazon Associate, Akermin earns from qualifying purchases. Product links on this page are affiliate links.
A battery tester works by applying a load, a small AC signal or a brief high-current pulse, measuring the voltage response, and computing an inferred quantity such as internal resistance, conductance or state of health. A load tester such as the TOPDON BT100 draws real current and watches voltage sag; a four-wire milliohm meter such as the YAOREA YR1035+ injects a 1 kHz AC signal and measures internal resistance directly; a monitor such as the MORNING GROUP DC 0-300V 100A unit does neither, and instead integrates voltage and current over time. None of them measures remaining capacity directly. They measure a proxy, and the proxy is only as good as the model behind it.
The three physical principles a battery tester can use
Every battery tester on the market reduces to one of three measurement principles, sometimes combined. The first is the load test. A known resistance or a controlled current draw is connected across the battery for a few seconds, and the tester records how far terminal voltage falls. A healthy lead-acid battery has low internal resistance, so its terminal voltage sags only slightly under a heavy load. A worn battery has high internal resistance, so the same load pulls the voltage down sharply. The tester compares the sag with the battery’s rated cold cranking amps and reports a pass or a state of health.
The second principle is AC internal resistance measurement. A small alternating current, typically at 1 kHz, is injected through the battery and the resulting AC voltage across the terminals is measured. Dividing one by the other gives internal resistance in milliohms. This is the principle behind a four-wire milliohm meter such as the YAOREA YR1035+, which the listing states measures internal resistance and voltage simultaneously using a true four-wire system. The four-wire arrangement matters because it separates the current-carrying leads from the voltage-sensing leads, so the resistance of the test leads and contacts is not included in the reading. On a battery whose internal resistance is a few milliohms, lead resistance of even a fraction of an ohm would swamp the result.
The third principle is coulomb counting, which is not really a test at all. A monitor measures current flowing in and out of the battery continuously and integrates it over time. The MORNING GROUP DC 0-300V 100A monitor is this type: the listing states voltage 0-300V, current 0-100A, power to 90000W and power consumption to 999kWh, with a built-in shunt. It tells you what has flowed, not what the battery can still deliver. That is a different question, and it is the one that matters for a solar or off-grid installation, which is why the choice of instrument depends on the question being asked rather than on the instrument being better or worse.
What state of health actually is, and why the number is a model output
State of health is not a directly measurable quantity. It is a ratio: the battery’s current capacity divided by its rated capacity, expressed as a percentage. No tester can measure that ratio directly without discharging the battery fully, which is slow and inconvenient. Instead the tester measures something fast, such as internal resistance or voltage sag under load, and maps it onto a health figure using a model calibrated for a particular battery chemistry and rating.
That mapping is why the battery type setting matters. A tester that offers settings for flooded, AGM flat plate, AGM spiral, EFB and GEL, as the TOPDON BT100 listing states, is applying a different model for each. The same internal resistance means different things on a spiral-wound AGM and on a flooded battery of the same cold cranking rating. If you leave the tester on the wrong setting, the health number is not merely imprecise; it is answering a question about a different battery.
The rating range matters for the same reason. The TOPDON listing states compatibility with 100-2000 CCA. Below or above that window, the tester has no calibration point and the health figure should not be trusted. The listing does not publish an accuracy figure for the state of health estimate. Not stated in the listing bullets. That is normal for this class of instrument, and it is worth understanding that the number is a modelled inference rather than a measurement with a published tolerance.
Voltage, current and resistance: which one the tester is actually reading
A common confusion is to treat all battery testers as variations on a multimeter. They are not. A multimeter measures voltage, current and resistance passively, using a small excitation current that the battery itself supplies. A battery tester imposes a condition on the battery and reads the response. The distinction is practical: a multimeter reading of 12.6 volts tells you the battery is fully charged at rest, and nothing about whether it can start an engine. A load test that pulls 100 amps and watches the voltage hold at 10.5 volts tells you the battery can deliver cranking current. The second reading is the one that predicts a no-start.
The table below sets out what each measurement principle can and cannot establish. The columns are deliberately blunt, because the most expensive mistakes in this category come from using an instrument outside the question it was built to answer.
| Principle | What is applied | What is read | What it can establish | What it cannot establish |
|---|---|---|---|---|
| Load test | Heavy current for a few seconds | Terminal voltage sag | Cranking capability, gross failure | Remaining capacity, small capacity loss |
| AC internal resistance | Small 1 kHz signal | Milliohm resistance and voltage | Trend in resistance, cell matching, screening | Absolute capacity without a baseline |
| Coulomb counting | Nothing; it observes | Current integrated over time | Energy in and out, consumption | Health, cranking ability, internal resistance |
The first two rows are tests. The third is monitoring. A monitor such as the MORNING GROUP unit is useful precisely because it does not test: it records what actually happened, including the parasitic drain that a one-off test would miss. If you are chasing a battery that goes flat overnight, the relevant procedure is covered in measuring parasitic battery drain, and a monitor is often the instrument that finds the culprit.
Why a four-wire measurement is not the same as a two-wire one
Internal resistance of a healthy lead-acid battery is on the order of a few milliohms. A test lead and its contact with the terminal can easily contribute ten times that. A two-wire measurement, where the same leads carry the excitation current and sense the voltage, includes the lead and contact resistance in the result. On a battery, that error is large enough to make the reading meaningless.
The four-wire, or Kelvin, connection solves this by using separate leads for current and voltage. The current leads carry the excitation; the voltage leads sense the drop across the battery alone, and because almost no current flows in the voltage leads, their resistance does not appear in the result. The YAOREA YR1035+ and the T-king YR1035 set both state a true four-wire system. The T-king listing goes further and states a 1 kHz AC sinusoidal excitation and a range from 0.00001 ohm to 200 ohm. That resolution figure is a display and measurement resolution, not an accuracy specification; the listing does not publish an accuracy figure. Not stated in the listing bullets.
This is also where a listing claim should be read carefully. The T-king bullet text contains phrases that do not describe the instrument, including a reference to a high voltage battery internal resistance tester and to handmade appearance. That text reads as though it belongs to a different product listing. The product title and the four-wire, 1 kHz, 0.00001 to 200 ohm range are the parts that describe a battery internal resistance meter, and those are the parts to rely on. The rest should be treated as listing noise rather than specification.
Where the whole category stops being the right answer
A battery tester answers one question well: is this battery, of this type and rating, likely to deliver its rated current in the near future? It does not answer how much energy is left, how long a solar installation will run overnight, or whether a battery is safe. Those are different questions with different instruments.
For capacity in amp-hours, a tester that reports state of health is reporting a modelled ratio, not a measured capacity. The honest procedure is a controlled discharge test, which is slow and requires a load bank, and the principles are set out in choosing a battery capacity tester. For a solar installation, the question is usually energy balance over days, which is a monitoring question rather than a testing one; choosing a battery monitor for an off-grid system covers the trade-offs, and choosing a solar PV tester covers the array side, which is a separate instrument again.
For a single voltage check before condemning a battery, a multimeter is enough, and the procedure is in testing a 12V battery with a multimeter. What a tester adds is the loaded or AC-excited response, which is what separates a charged but worn battery from a healthy one. What no tester adds is an alarm. A tester is not a monitoring system, and a monitor is not a tester. If the requirement is to know that a battery is failing before it fails, the instrument is a monitor with logging, not a hand-held tester used once a season.
The reading is a proxy, and the proxy has a domain
A battery tester works by imposing a known condition and reading the response, then mapping that response onto a health figure using a model tied to battery type and rating. The mapping is useful inside its domain and misleading outside it. A load test tells you about cranking; an AC resistance measurement tells you about trend and cell matching; a monitor tells you what actually flowed. Choose the instrument that matches the question, set the battery type correctly, and treat any health percentage as an inference rather than a measurement.
Frequently Asked Questions
Can a battery tester tell me how many amp-hours are left in my battery?
No. A tester measures a proxy such as voltage sag under load or internal resistance, and infers state of health. Amp-hour capacity requires a controlled discharge over time, which is a different procedure. A monitor that counts current in and out can estimate state of charge, but that is bookkeeping, not a test, and it drifts if the battery is not periodically fully charged.
Why does my tester give a different health figure from my mechanic's tester?
Because the two instruments are applying different models. Battery type setting, rating range, temperature and the exact excitation used all change the result. A tester set to AGM on a flooded battery will not agree with one set correctly. The absolute number is less useful than the trend on the same instrument with the same settings.
Is a four-wire internal resistance meter better than a load tester?
They answer different questions. A four-wire meter measures internal resistance directly and is excellent for screening cells and tracking resistance trend. A load tester applies real cranking current and tells you whether the battery can start the engine. For a car, the load test is closer to the question. For matching cells in a pack, the four-wire meter is the right instrument.
Can I use a battery tester on a lithium battery?
Only if the tester explicitly supports that chemistry. The listings referenced here describe lead-acid types: flooded, AGM, EFB and GEL. Applying a lead-acid model to a lithium battery will produce a health figure that means nothing. Not stated in the listing bullets for any of the products above regarding lithium support.
Does a battery tester need the battery to be fully charged?
Yes, for a meaningful health reading. State of health models assume a rested, charged battery. Testing a discharged battery confounds state of charge with state of health, and the tester will often report a low health figure that recovers once the battery is charged. Charge, rest, then test.
Will a battery tester tell me if my alternator is charging?
Some do, as a separate function. The TOPDON BT100 listing states a charging test that checks charging current, rectifier diode and generator. That is a charging system check, not a battery health check, and it is worth running separately. A tester that does not offer it cannot answer the question, and a monitor such as the MORNING GROUP unit can show charging current directly.