IllustrationA battery emulator is a two-quadrant supply that can both source and sink current and adds a programmable internal resistance. Your device then sees the voltage sag, discharge curve and charge path of a real battery. Use it for repeatable battery-life, brownout and charger tests, then confirm the results on real cells.
A bench supply holds its voltage during a radio burst where a real cell sags, and it gives a charger nothing to charge. This guide covers what a battery emulator does for a small device, how internal resistance and charge current change a test, how fast it must respond to load steps, and when an SMU, a bench supply or a real cell fits better, with a decision table and a test checklist.
This is an engineering summary based on vendor documentation and application notes. Specs change, so check the linked pages before you buy or plan a test. All sources were read on November 15, 2026.
What does a battery emulator do?
A battery emulator is a power supply that acts like a battery. Rohde & Schwarz says the linear two-quadrant design of its NGM200 supplies lets them operate as a source and sink, and that in battery simulation mode the supply behaves like a predefined battery and can simulate discharging and charging a rechargeable battery (Rohde & Schwarz application card, as of November 2026).
- Two quadrants. As soon as the externally applied voltage exceeds the set nominal voltage, current flows into the power supply, which switches from source mode to sink mode on its own (R&S NGM200 page, as of November 2026).
- Internal resistance. Qoitech says Otii adjusts the output voltage depending on the load, just like a real battery with internal resistance (Otii battery emulator, as of November 2026).
- A battery model. Rohde & Schwarz lists capacity, open circuit voltage (Voc) and equivalent series resistance (ESR) as key characteristics that depend on state of charge (SoC), and its NGM-K106 option simulates batteries from that data (R&S NGM200 page, as of November 2026).
For how the battery, charger and protection parts fit together on the board, see battery management systems for small devices.
How does internal resistance change what your device sees?
Voltage sag during radio bursts
TI studied pulsed loads on CR2032 coin cells for Bluetooth low energy sensors. Its test profile used a 30 mA or 15 mA RX/TX peak for 1 ms and a 0.1 mA sleep for 200 ms, and it counted a cell as at end of life when the voltage dropped below 2.0V (TI SWRA349, as of November 2026).
TI found that the voltage drop caused by internal resistance during peak load limits the effective capacity, and that internal resistance increases rapidly as capacity is used. It gives the limit for the 30 mA peak as "approximately 30ohm" (TI SWRA349, as of November 2026). Engineering reasoning: a device can stop working with charge still in the cell, because the peaks pull the voltage under its brownout threshold.
A bench supply hides this. An emulator lets you raise the internal resistance step by step and find where the device fails. TI also found that a capacitor in parallel with the cell gave more than 40% improvement with poor quality CR2032s (TI SWRA349, as of November 2026). An emulator lets you check such a fix at high internal resistance without waiting for a cell to age. For the radio side, see what Bluetooth Low Energy is.
How much resistance can you set?
The NGM200 data sheet lists an adjustable output impedance up to 100 Ω in 1 mΩ increments (R&S NGM200 data sheet, as of November 2026). Qoitech lists an ESR range up to 5 kohm, with resolution down to 1 mohm, for the Otii Arc Pro, and notes that the toolbox emulates the total ESR (Otii Battery Toolbox, as of November 2026). Engineering reasoning: total ESR includes the test leads, so keep them short.
Replaying a discharge curve
A discharge curve tests the whole battery life, not one moment. Qoitech shows two curves for a chosen battery, unloaded voltage over used capacity and internal resistance. Its Fixed setting holds a constant used capacity, and Follow emulates a discharge over the time you are recording (Otii battery emulator, as of November 2026).
The model is only as good as its data. Qoitech's profiler discharges a real cell at a high and a low level and uses the difference to calculate internal resistance. It also advises profiling the battery at the temperatures it will see in real use (Otii battery profiler, as of November 2026).
Why does the emulator need to sink current?
A charger needs somewhere to push current. A single-quadrant supply cannot absorb it, so the charger never sees a battery. Rohde & Schwarz says the charging behavior of a battery can also be simulated, which matters when designing battery chargers, and that the NGM200 is used in sink mode for this (R&S NGM200 data sheet, as of November 2026).
Size the sink for your charger. TI's BQ25120A, a charger for wearables, supports charge currents from 5 mA to 300 mA and a configurable termination current down to 500 µA (TI BQ25120A data sheet, as of November 2026). The Otii Arc Pro lists a sink current up to 2.5 A with a setting resolution of 39 µA, and the sink requires an Otii Toolbox (Otii Arc Pro, as of November 2026). Engineering reasoning: the sink must cover the fast-charge current, and the measurement must resolve the small current at the end of charge, or you cannot see whether termination works.
How fast must it respond when the load jumps from sleep to a radio burst?
Rohde & Schwarz notes that battery devices require very little power in sleep mode, but the current increases abruptly as soon as the device switches to transmit mode. A supply for these devices must handle load changes from a few nA to the ampere range without creating voltage drops or overshoots (R&S NGU page, as of November 2026). Engineering reasoning: a slow control loop means you measure the instrument instead of the battery.
Logging speed matters too. The NGM200 FastLog function reaches up to 500 ksample/s, so voltage and current results are available every 2 µs (R&S NGM200 page, as of November 2026). The Otii Arc Pro lists 4 ksps in its ±19 mA range and an analog bandwidth of 400 Hz (Otii Arc Pro, as of November 2026).
Qoitech notes that an LDO typically draws energy in constant current mode and a DC/DC in constant power mode (Otii battery profiler, as of November 2026). Engineering reasoning: behind a buck converter, battery current rises as battery voltage falls, so the same burst sags harder near empty. See LDO vs buck converter.
Battery emulator vs SMU vs bench supply vs real cell
The rows are engineering reasoning. Sourced facts are in the sections around the table.
| Tool | Strength | Misses | Use it for |
|---|---|---|---|
| Battery emulator | Sources and sinks current, adds internal resistance, follows a discharge curve. Runs repeat exactly. | Cell-to-cell spread, aging and temperature, unless the model includes them. | Battery-life runs, brownout and charger tests. |
| Source-measure unit | Fine current measurement over a wide range. Sources and sinks. | Battery models are optional on some models. | Sleep current, leakage, single rails. |
| Bench supply | A stiff, stable voltage. Simple and cheap. | Usually cannot sink current. Hides the sag a real cell shows. | Bring-up and early firmware work. |
| Real cell | True chemistry, protection circuit, temperature and aging. | Every cell differs, and a drained cell takes hours to restore. | Final validation and cold tests. |
A source-measure unit puts measurement first. Rohde & Schwarz says its NGU units have two or four quadrant architecture to work as a source and a sink, and measure current drains from nA to A in one pass (R&S NGU page, as of November 2026). The data sheet gives resolutions down to 1 µV/100 pA and describes the NGU201 as mainly designed for extremely low sleep currents, such as Bluetooth low energy devices (R&S NGU data sheet, as of November 2026).
Check low-range accuracy before you trust a sleep reading. The NGM200 gives < 0.05 % + 15 μA in its 10 mA range (R&S NGM200 data sheet, as of November 2026). The Otii Arc Pro gives ±(0.1 % + 50 nA) for currents up to 19 mA in auto range (Otii Arc Pro, as of November 2026). Engineering reasoning: for microamp sleep currents, the offset term decides whether the reading means anything.
Nordic describes its Power Profiler Kit II as a tool to obtain real-time current measurements of your designs (Nordic Power Profiler app, as of November 2026). Before using any profiler as a battery stand-in, check that it can sink current.
When is a real cell still the right test?
An emulator repeats one model exactly. TI tested at least 12 batteries per vendor because battery-to-battery variation can be quite large. It found an average capacity loss of 9% between the 15 mA and 30 mA peaks, while some vendors only reached about 50% of rated capacity (TI SWRA349, as of November 2026). TI also notes that internal resistance increases with lowering temperatures (TI SWRA349, as of November 2026).
Use the emulator to find limits fast, and real cells to confirm them. A battery-life claim should rest on real-cell runs. See how to test battery life claims honestly.
What should a battery emulator test plan include?
Run these before the power design is frozen.
- Record the current profile first: sleep, wake, radio burst and any motor peaks.
- Load a model of the exact cell you will ship.
- Run the profile at full charge and near empty. Log the lowest voltage at the device pins during each burst.
- Raise the internal resistance until the device browns out. Compare that value with a real cell near empty.
- Check that firmware reads battery voltage outside radio bursts and warns before the brownout.
- Connect the charger and confirm the emulator sinks current through charge and termination.
- Plug and unplug the charger while the radio is active.
- Repeat key runs on real cells from more than one lot, including a cold run.
Power tests touch the schematic, firmware and charger at once. For how that work is organized on a custom device, see how the electronics work is organized.
Questions about battery emulators
Can a normal bench power supply stand in for a battery?
For bring-up, yes. Rohde & Schwarz notes that a power supply should have an output impedance as low as possible to suppress loading effects on the device (R&S NGU data sheet, as of November 2026). A real cell has internal resistance, so a stiff supply hides voltage sag. A single-quadrant supply also cannot absorb charger current.
Do I still need to test on real batteries?
Yes. TI tested CR2032 cells from six vendors and found that some only reach about 50% of rated capacity under a pulsed radio-style load (TI SWRA349, as of November 2026). An emulator repeats one model. Real cells from several lots show the spread your customers will see.
How can Duobay help with battery testing for a custom device?
Duobay develops custom devices with a Shenzhen engineering team that covers electronics, firmware, enclosure, app and test, and builds samples and small batches. Duobay is also developing an AI hardware debug box for this kind of bench work; it is in development, and details may change. Firmware source and the SBOM are delivered to you, and source files are assigned to you at every paid gate. Initial calls are free, and we do not publish prices. Tell us about your device.
If you are planning battery tests for a new device, tell us about your device.
Sources. All pages accessed November 15, 2026. Instrument specs and software change, so check the current version before you buy or plan a test.
- Rohde & Schwarz, Minimizing power consumption, application card (link)
- Rohde & Schwarz, R&S NGM200 power supply series, product page (link)
- Rohde & Schwarz, R&S NGM200 power supply series data sheet, version 04.00 (link)
- Rohde & Schwarz, R&S NGU source measure units, product page (link)
- Rohde & Schwarz, R&S NGU source measure units data sheet, version 02.01 (link)
- Qoitech, Otii user manual, Battery emulator (link)
- Qoitech, Otii user manual, Otii Arc Pro (link)
- Qoitech, Otii user manual, Battery profiler (link)
- Qoitech, Otii user manual, Battery Toolbox overview (link)
- Texas Instruments, Coin cells and peak current draw, white paper SWRA349 (link)
- Texas Instruments, bq25120A data sheet, SLUSD08A (link)
- Nordic Semiconductor, Power Profiler app repository on GitHub (link)
Planning a device like this? Tell us what it needs to do. A Duobay engineer replies within 1 business day. Send a project brief →