DC Power Supply Glossary
Datasheets compress a lot of engineering into a few abbreviations. This glossary defines the terms that actually decide whether a programmable DC power supply fits your application, with the precise meaning behind each spec rather than the marketing gloss.
Topology and direction
| Bidirectional | A supply that can both source current (deliver power, like a charger) and sink current (absorb power, like a load) through the same terminals. Essential for battery charge-discharge cycling and inverter test. |
| Two-quadrant (dual-quadrant) | Operates in two of the four voltage/current quadrants, typically positive voltage with either sourcing or sinking current. Covers most charge-discharge work at a single polarity. |
| Four-quadrant | Sources and sinks at both voltage polarities, the full V-I plane. Needed where the device under test can swing the terminal voltage either way. |
| Regenerative load | A sink that returns absorbed energy to the AC grid instead of dissipating it as heat in resistors. Cuts energy cost and cooling load; recovery efficiency (e.g. up to 93%) is the figure of merit. |
| Dissipative load | The traditional alternative, which converts absorbed power to heat. Simpler, but expensive to run and cool at high power. |
Operating modes
| CC (constant current) | The supply holds output current at the setpoint and lets voltage float to whatever the load requires. Used for battery charging and current-limited tests. |
| CV (constant voltage) | Holds voltage fixed while current varies with load demand. The default for powering a device at a fixed rail. |
| CP (constant power) | Regulates the product of voltage and current to a fixed wattage, adjusting both as the load changes. Mirrors how converters and motors actually draw energy. |
| CR (constant resistance) | Behaves like a fixed resistor, drawing current proportional to terminal voltage. Useful for emulating resistive loads and inrush behaviour. |
Accuracy and dynamics
| F.S. (full-scale) accuracy | Error expressed as a percentage of the instrument's full range, not the reading. A 0.02% F.S. spec on a 2250 V supply means about ±0.45 V of uncertainty across the range. Always check whether a number is %F.S. or %reading. |
| Slew rate | How fast the output can change, in volts or amps per unit time. Governs how steep a ramp the supply can produce for transient and MPPT testing. |
| Response time | How quickly the output recovers and settles after a step or load change (e.g. ≤5 ms). Sets the usable bandwidth for dynamic profiles. |
| Readback | The instrument's own measurement of its output voltage and current, returned over the control interface. High readback resolution is what makes a supply double as a precise meter, critical for cell-level coulomb counting. |
| Resolution | The smallest increment the supply can set or read. Distinct from accuracy: a supply can be high-resolution but biased, or accurate but coarse. |
Simulation and emulation terms
| MPPT (maximum power point tracking) | The algorithm a PV inverter uses to keep its operating point at the peak of the solar array's power curve. Tested against an EN 50530 reference curve on a PV simulator. |
| SoC (state of charge) | The remaining capacity of a battery as a percentage of its full capacity. A core input to any battery emulation model. |
| OCV (open-circuit voltage) | A battery's terminal voltage with no load, which varies with SoC. The SoC-to-OCV relationship is the backbone of a battery model. |
| Ri (internal resistance) | The battery's effective series resistance, which produces voltage sag under load. Together with SoC and OCV it lets a source emulate a battery's dynamic terminal behaviour. |
| I-V / P-V curve | The current-voltage and power-voltage characteristic of a solar array. A PV simulator reproduces this curve at its terminals so an inverter sees a realistic source. |
For worked examples of these terms in real test setups, the Ultra Power Systems N35500 series documentation shows how the same specifications map onto battery, PV and ATE applications.
Frequently asked questions
Does 0.02% F.S. accuracy mean 0.02% of my reading?
No, and the difference is large at low output. Full-scale accuracy is 0.02% of the instrument's full range, so on a 2250 V supply it is a fixed band of roughly 0.45 V whether you set 2000 V or 50 V. At low setpoints, percent-of-reading specs look tighter, so always confirm which convention a datasheet uses.
What is the difference between two-quadrant and four-quadrant operation?
Two-quadrant operation typically sources and sinks current at one voltage polarity, enough for most single-polarity charge-discharge work. Four-quadrant adds both voltage polarities, covering the entire voltage-current plane, which is needed when the device under test can drive the terminal voltage negative.
Why does readback resolution matter so much for battery testing?
Capacity and coulombic efficiency are derived by integrating measured current over time. If the supply's own current readback is coarse or biased, that error propagates straight into the capacity figure. High readback resolution lets the source act as a precise meter, removing the need for a separate instrument.
Specifying a bidirectional DC supply?
The articles here describe the N35500 platform from Ultra Power Systems.