Battery Emulation Explained
Battery emulation makes a programmable supply behave electrically like a real cell or pack, including how its voltage sags under load and shifts with state of charge. Done well, the device under test cannot tell it is connected to an emulator rather than a battery.
What a battery looks like electrically
To a connected device, a battery is not a fixed voltage. Its terminal voltage is set by three things working together:
- Open-circuit voltage (OCV): the rest voltage, which varies with state of charge along a characteristic curve.
- State of charge (SoC): the remaining capacity, which rises while charging and falls while discharging.
- Internal resistance (Ri): the series resistance that drops voltage under load — terminal voltage sags when current is drawn and rises when current is pushed in.
An emulator computes terminal voltage as OCV(SoC) minus the current times Ri, continuously integrating current to track SoC. The result is a source that behaves like the real chemistry, not a flat bench voltage.
Why bidirectional operation is mandatory
A real battery both supplies and accepts current. If a device under test — say an inverter or a DC-DC converter — ever pushes current back into the emulated battery, the emulator must absorb it and raise the modelled SoC, exactly as a real cell would charge. A source-only supply cannot do this; it would just clamp or fault.
That is why credible battery emulation requires a bidirectional supply. Sourcing covers discharge, sinking covers charge, and the seamless crossover between them lets the emulated SoC move up and down naturally as the device under test draws or returns current. The same loop characteristics that give a clean source/sink crossover give a believable emulated battery.
Model libraries versus measured curves
There are two ways to get the OCV/SoC/Ri behaviour into the emulator. The NS81000 battery-emulation function on the N35500 supports both:
- Model libraries: seven built-in model libraries cover common chemistries and configurations, so you can start emulating quickly with representative parameters.
- Custom measured curves: you load real OCV-versus-SoC and internal-resistance data measured from your own cells, so the emulation matches the exact battery your product will see in the field.
Libraries are fast for early development; measured curves are essential for validation, where small differences in voltage sag change how the device under test behaves. Ultra Power Systems provides the NS81000 emulation function for the N35500 platform.
Emulator versus simulator
The terms are often blurred but the distinction is real. A battery simulator in the loose sense may just output a programmed voltage. A battery emulator models the dynamic relationship between current, SoC and terminal voltage, so it reacts the way a cell would — voltage falls under load, recovers at rest, and tracks charge over time. For testing chargers, BMS units and converters, that dynamic behaviour is the whole point: it exercises the device under realistic source conditions without the safety hazards, slow turnaround and aging of physical cells.
Frequently asked questions
What is the difference between a battery simulator and a battery emulator?
Why does battery emulation need a bidirectional supply?
Should I use a model library or measured curves?
Specifying a bidirectional DC supply?
The articles here describe the N35500 platform from Ultra Power Systems.