Inside a programmable DC electronic load, four regulation modes share one hardware platform but run different control loops: CC forces a fixed sink current via a current-shunt feedback path, CV clamps the input terminal voltage with a voltage-divider feedback path, CR divides the sensed input voltage by a programmed conductance to draw I = V/R, and CP multiplies the sensed V and I and regulates their product to a fixed wattage setpoint [S2][S4].
The mode you pick decides what your device under test (DUT) actually sees at its terminals, and that is the single most important decision when a bench engineer, battery lab, or production line programs a discharge profile [S3].
Constant Current (CC): the default discharge mode
CC mode sinks a programmed current regardless of the source voltage, up to the load's maximum current rating, which is why the CC loop is the go-to choice for battery discharge testing, load regulation checks of regulated supplies, and constant-current stress of DC-DC converters [S2][S3].
Because the current stays flat while the source voltage sags (a battery from 4.2 V to 3.0 V, for example), CC cleanly exposes the source's voltage regulation under fixed demand, and it pairs naturally with a defined discharge C-rate on Li-ion cells [S3]. Programmable loads also use CC as the basis for dynamic profiles, where the current setpoint is stepped or pulsed to simulate CPU wake events, motor inrush, or inverter switching [S3][S4]. The practical limit is the load's Safe Operating Area (SOA): a load rated 1000 W at 80 V cannot draw 1000 W at 5 V even if the current channel supports it, so CC at low voltage can become power-limited, not current-limited [S3].
Constant Voltage (CV): battery simulator for charger test
CV mode commands the load to sink whatever current is required to keep its input terminals at a programmed voltage, effectively turning the load into a programmable, current-limited battery [S2][S4]. This is the standard method for verifying an onboard charger (OBC), USB-PD source, or any current-limited supply under a controlled back-EMF [S3].
CV behaves as a closed voltage clamp: the load's internal amplifier compares the sensed terminal voltage to a reference and increases sink current until the two match, so any DUT that tries to push the terminal above the setpoint is held in place, an important feature when discharging through a regenerative source or a solar array simulator [S4]. The mode has a hard ceiling, set by the load's maximum sink current, and the load will drop out of regulation if the DUT cannot deliver the current demanded by Ohm's law at the setpoint, a condition visible as the load switching to CC or hitting its current limit [S7]. Engineers should size the load's current rating for at least 1.2× the maximum expected DUT current when in CV, to avoid spurious mode changes at the boundary [S3].
Constant Resistance (CR): the controllable rheostat
CR mode uses a control amplifier that directly compares the measured input voltage against a programmed conductance, sinking I = V/R in real time so the load emulates a fixed resistor across the operating range [S4][S6]. It is the mode of choice for measuring inrush and start-up behaviour of electronic devices, characterising battery capacity at a near-constant impedance, and replicating resistive cable drops without swapping physical resistors [S2][S4].
Unlike a real resistor, the electronic load's CR setting is digitally programmable in fine steps, can be swept during a test, and avoids the heating drift of a wire-wound rheostat, but it still obeys V = I·R, so at very low terminal voltages the current floor becomes too small to resolve and the mode becomes unstable, a limit most users hit around 0.5–1% of the load's full-scale current [S3]. CR is also the basis for many automotive load-dump and brown-out tests, where the source behaviour into a fixed R is part of the compliance pass/fail [S4]. When the device under test is a DC power supply being qualified for constant-impedance loads, CR is the most representative mode available in bench form.
Constant Power (CP): the constant-power load profile
CP mode (also marked CP/CW on some instruments) sinks whatever current is needed to keep the product V·I at a programmed wattage, so the load emulates a class of real-world devices that draw more current as their supply voltage sags, notably battery chargers near constant-power, telecom rectifier stages, and LED drivers operating in their regulated region [S1][S2][S9].
The control loop multiplies sensed V and I and integrates the error against the setpoint, so the current draw is inversely proportional to terminal voltage: a 300 W CP setpoint at 60 V draws 5 A, at 30 V it draws 10 A, and at 15 V it draws 20 A, which is exactly the kind of profile a fuel cell or a regulated solar array sees from its downstream converter [S4][S9]. CP is bounded by the same SOA as CC, so the minimum voltage on the load sets an upper bound on the current it must sink, and a poorly chosen CP setpoint at low voltage can still hit the current channel's limit [S3]. Engineers who routinely test converters used in DC power supply front ends and battery simulators find CP indispensable because it reproduces the dynamic I(V) curve of the real load better than CC or CR alone [S4][S9].
How the four modes compare on the criteria that matter
Lining the four modes up against four common decision criteria makes the trade-off visible at a glance for specifiers and lab managers [S2][S3][S4]:
Primary controlled parameter: CC controls current, CV controls voltage, CR controls equivalent resistance (V/I), and CP controls the V·I product.
Best fit for battery discharge: CC dominates because C-rate is a current spec; CR is used for capacity-at-impedance; CV is the wrong tool (it would force a flat terminal voltage rather than allow natural sag).
Best fit for charger / source test: CV dominates because it replicates a battery as a load; CP matches the constant-power brick profile; CC is used only for steady-current stress.
Settling / loop stability: CC and CV are the fastest and most stable because the feedback path is single-variable; CR is slightly slower due to division in the loop; CP is the slowest because the loop depends on the multiplication of two sensed values and can oscillate near the SOA edge [S4][S6].
On a single bench instrument, the four modes are firmware-selectable on the same hardware, so the choice is always application-driven, not hardware-driven [S2][S3].
When each mode is the wrong pick
CC is wrong for testing a constant-voltage bus where the regulation point itself is what you want to characterise, because forcing flat current will not exercise the cross-regulation of the source [S3]. CV is wrong for measuring a battery's capacity, because the load will keep the terminal at the setpoint and never let the cell's natural discharge curve develop [S2]. CR is wrong for characterising a converter's constant-power region, because R stays flat while real loads increase current as V drops, so CR will under-stress the converter at low line [S4]. CP is wrong for measuring pure inrush into a capacitive load, because the V·I loop can chase its own tail and produce an over-current trip rather than a clean inrush capture, so CC or CR with a current limit is preferred for that case [S3][S4].
The general rule, drawn from OEM guidance, is: match the mode to the physical load the DUT will see in service, not to the parameter you want to plot on a graph, because most compliance curves (efficiency, hold-up, regulation) are defined against a specific load class [S3][S4].
Beyond the four: short-mode, dynamic, and battery-test profiles
Modern programmable loads add a short-circuit mode (zero-ohm or zero-volt transient), a transient or dynamic mode (alternating between two CC levels at a programmed slew rate, used to test regulation recovery), and a battery-test mode that automates a CC or CP or CR discharge to a cut-off voltage, recording Ah and Wh [S2][S3][S9]. These are not separate control loops, they are sequencers built on top of the four basic modes, which is why a spec sheet that only lists CC/CV/CR/CP still implies support for those higher-level tests through firmware [S3][S9].
The under-recognised constant-power profile, in particular, is increasingly specified for fuel-cell and PV-inverter testing where the source's V·I curve is non-linear and neither CC nor CR reproduces it faithfully, so the CP loop has moved from a niche feature to a required line on most 2025-2026 spec sheets for bench electronic loads above 1 kW [S3][S4].
Sourcing, standards, and what to verify before you buy
Selection criteria worth checking on the datasheet, not on marketing copy: stated accuracy in each mode (CC is usually the tightest, CP the loosest, often ±0.5% to ±1% of set plus offset), minimum operating voltage at full current (often 1-2 V for a MOSFET-based load, higher for older IGBT designs), SOA curves for each mode, and protection setpoints (OCP, OPP, OVP) that are independent of the regulation setpoint [S3][S4]. The standard interface for remote programming is typically USB, LAN, or GPIB; SCPI command sets are common but not identical across vendors, which matters when you port a test sequence between instruments [S3].
For application notes on how constant-power load profiles interact with regulated converters and on the boundary between CV and CC as a source sags, OEM application notes from the load's own maker remain the most reliable primary source [S4][S6]. A useful cross-reference for the constant-resistance case in flow control is the way a fixed R on a hydraulic line behaves like a CR electronic load, a comparison laid out in the Full Bore vs Reduced Bore Ball Valve: Pressure Drop, Cv and Selection piece, where a constant Cv behaves the same way at the pipe level. Engineers choosing a bench load for converter validation will also want to cross-read guidance on the source side, the DC power supply and DC-DC converter reference pages, before locking in a mode strategy for a new test plan.