Battery Module & Pack Testing Systems

Battery Module and Pack Testing from 20 V to 1,500 V

Arbin manufactures linear and regenerative testers for battery module and pack testing, from 20 V modules to 1,500 V packs at up to 1.8 MW. Tests are written and run in MITS, which integrates the whole test facility: chamber control, BMS communication over CAN, and results written straight to your database.

arbin product lineup hero web

Tests built in MITS

30+ programmable control types, 90+ meta variables and 127 reusable sub-schedules, with drive-cycle profiles built in the Simulation Editor.

CAN and BMS communication

CAN 2.0 and CAN-FD with DBC file import, plus UDS and SMBus. BMS signals are logged alongside voltage and current on one timebase.

Open APIs when you need them

ArbinCTI over TCP/IP with Python and C# clients, and ArbinDriver for direct control from a PLC or your own application.

Energy back to your facility

Up to 92% of discharge energy returned to the grid at 0.99 power factor, with anti-islanding disconnection in roughly 40 ms. Air-cooled, no plant water.

Find the Right Module & Pack Testing Platform

Three systems, grouped by technology (linear vs regenerative) and application (module vs pack). All three run MITS, build schedules and drive-cycle profiles the same way, and share the same CAN and API interfaces - moving test programs from module testing to pack testing does not mean relearning the software.

Module Testers

arbin lbts cell 1

All-Purpose · Linear

LBTS-Module

20 V–100 V · 0.1 A–80 A · up to 32 ch

Arbin's linear module tester, built for module R&D and end-of-line work. Four auto-switching current ranges per channel, up to 960 A with channels paralleled, in a compact air-cooled chassis that runs on a standard three-phase supply.

View LBTS-Module →

arbin rbt4 cell

Regenerative

RBT-Module

30 V–200 V · 10 A–300 A/ch · to 1,800 A

High-current module cycling with true bipolar circuitry and no switching time across zero, so a drive cycle is followed rather than approximated. Up to 96 kW charge power per module, with up to 90% of the discharge returned to the grid

View RBT-Module →

Pack Testers

ARBIN-RBT-High Power-right

Regenerative · High Voltage

RBT-Pack

300 V–1,500 V · 50 A–200 A/ch · to 360 kW/chassis

Pack-level cycling, end-of-line and BMS validation at full pack voltage. Dual auto-selecting voltage ranges per channel, Gen3 SiC power stages returning up to 92% of the discharge, and the standard drive cycles running natively.

View RBT-Pack →

Beyond one channel

Series and parallel, to 1.8 MW

Channels combine in series to raise voltage and in parallel to raise current and power, and whole chassis parallel beyond that. Paralleled channels gain a combined full-scale current range and keep their lower native ranges as well.

1,500 V
Combined, in series
3,000 A
Combined, in parallel
1.8 MW
Combined output power

Testing cells rather than modules?

The cell range runs from 10 µA on a benchtop unit to 1,600 A paralleled, including per-cell thermal control and high-throughput incoming QC.

Making the Choice

Which module or pack platform fits your work?

The comparison table below has the full numbers. This is the reasoning behind which system suits which job, and what the software and the facility need to provide alongside it.

01

Module Testers

Choose LBTS-Module for module R&D and end-of-line work where channel count matters more than power. Four auto-switching current ranges cover 0.1 A to 80 A on the same channel, at up to 32 channels per chassis.

Choose RBT-Module above roughly 100 A, or where a test runs for months and the power and cooling bill matters. Its regenerative stages return up to 90% of the discharge energy to the grid instead of dissipating it as heat in the room.

Best for
Module R&D End-of-line Lifecycle testing High-rate cycling
02

Pack Testers

RBT-Pack is specified by the pack's voltage window, 300 V to 1,500 V. Each channel carries two auto-selecting voltage ranges, so a 400 V pack runs on the 500 V range rather than at full scale.

Choose it for pack-level work: standard drive cycles, BMS validation over CAN, and closed-loop testing against a real controller. Channels combine to 3,000 A and 1.8 MW where one channel is not enough.

Best for
Pack lifecycle BMS validation Drive cycles EOL & incoming QC
03

Software and Integration

All three systems are written and run in MITS. Schedules, drive-cycle profiles, safety limits and data logging are all set up there, and MITS also controls third-party chambers and chillers, logs BMS traffic over CAN, and writes results straight into SQL.

Where a rig has to be part of something larger, the same hardware opens up: ArbinCTI and ArbinDriver let your own software, a PLC or an HIL setup command the channels, and control values can be exchanged with the BMS over CAN during a test.

Worth settling early
MITS or external control CAN / BMS interface Where data lands AC supply & grid-tie

Not sure which applies? Send the module or pack voltage window, peak current and power, and the channel count you need. An application engineer will come back with a specific configuration and what it asks of the room.

Integration interfaces

What the system connects to alongside MITS, and how it opens up when a rig has to be driven from outside.

BMS & vehicle buses

CAN, SMBus and UDS

CAN 2.0 and CAN-FD, with DBC file import so your existing message database is used as-is
UDS for diagnostic services, and SMBus 2.0 / I²C for smart-battery registers
BMS signals logged in real time, on the same timebase as voltage and current
Control values sent and received over CAN, so the BMS can steer the test step
Safety limits can be driven from BMS-reported values, not only tester measurements
The API layer

ArbinCTI & ArbinClient

TCP/IP API with Python and C# clients, for third-party software to monitor and control tests
Both command-response and publish-subscribe architectures supported
Multi-dimensional lookup tables generate control values live from measured voltage, temperature, capacity or auxiliary inputs
Custom Test Instructions in C#/.NET for logic beyond the standard command set
Connects to LIMS, MES, digital twins and external analytics platforms
Direct, no MITS

ArbinDriver

A portable DLL loaded straight into your own application
Talks directly to the per-channel microcontrollers, with no MITS in the loop
Intended for high-speed control pipelines and PLC integration
The right layer when the cycler is a subsystem of a larger rig rather than a test station
Hardware safety limits and interlocks remain active regardless of what is commanding the channel
Lab-wide

I/O, chambers and data

Analog and digital I/O with PID or open-loop control, relay and TTL, for pumps, fans, valves and interlocks
Third-party chambers and chillers driven and logged from inside the schedule via MTCI
Data straight into Microsoft SQL Server or PostgreSQL, streamed via Apache Kafka, or auto-exported as CSV on a schedule
Auxiliary temperature, voltage and pressure inputs logged on the test timebase
UPS option for controlled shutdown and restart through a power interruption
Simulation, Drive Cycles & BMS

Test the pack against real load profiles

Module and pack testing is run against load profiles rather than constant currents. Profiles are built in the MITS Simulation Editor and called from a schedule like any other step, or driven over CAN while the test runs. The same channels also emulate a battery, so a charger, inverter or BMS can be tested closed-loop before a real pack exists.

10 ms
Minimum set-point interval on a simulation profile
≤2 ms
Current rise and fall, 10% to 90% of full scale, single channel
Zero
Switching time between charge and discharge, from true bipolar circuitry
1 kHz
Maximum data acquisition rate, from 10 kHz ADC sampling
Drive cycles & load profiles

Profiles are built in the Simulation Editor and called from a schedule like any other step, or driven dynamically over CAN while the test runs.

FUDS, HPPC, DST, WLTP and custom cycles
Time vs. current, time vs. power and time vs. load
Uploaded as a file, or streamed live over CAN
Adaptive logging raises the acquisition rate through transients
Battery simulation & HIL

The channel can emulate a battery rather than test one, using an Rint model driven by your own lookup tables — V = OCV(SOC) − I·R(SOC).

SOC–OCV–DCR tables from your own test data or the manufacturer's spec
Closed-loop testing of external devices — chargers, inverters, controllers
BMS emulation where the real controller is not available yet
Lua scripting for interpolation and custom real-time control algorithms
Validation & production

The same channels and the same schedules cover the work from first prototype module to the end of a production line.

BMS validation against live CAN traffic
Module and pack end-of-line testing, with PLC handshaking over digital I/O
Incoming quality control and capacity grading
Lifecycle, DCIR and pulse characterization, second-life assessment
Unattended and remote operation, with fail-safe recovery after a power interruption

Control is built from 30+ programmable control types, 90+ meta variables, 16 user-defined variables, up to 9 nested loops and 127 reusable sub-schedules, with nested AND/OR termination logic — so a validated drive-cycle procedure is called rather than rebuilt. MITS runs on Windows, macOS and Linux from one codebase, and MITS 11, 10 and 8 remain supported.

Side by Side

Compare Arbin Module & Pack Testing Systems

Compare each platform by application, voltage window, channel count, and how it integrates into a wider test system. Current and power figures are per channel; combined series and parallel ceilings are given on their own line and never folded into the headline range.

Product Best For Voltage & Current Channels Integration & Control
Module Testers
LBTS-Module All-Purpose Linear module cycler View Product → Module R&D and end-of-line testing where channel count matters more than power, and the tester sits on a production line rather than a grid connection. 20 V – 100 V Six range options. 0.1 A – 80 A per channel across four auto-switching ranges; up to 960 A paralleled. Rise time under 1 ms. 4 – 32 4 or 8 per module, up to 32 per chassis MITS platform, shared Third-party chamber and automation-software integration, real-world profile simulation, and data into MS SQL, PostgreSQL or Kafka.
RBT-Module Regenerative High-current regenerative module cycler View Product → High-current module cycling, drive-cycle simulation and BMS-in-the-loop work, and long-running tests where facility power and cooling are a live constraint. 30 V – 200 V Five range options, all operating down to 8 V. 10 A – 300 A per channel across up to three auto-switching ranges; 1,800 A paralleled per chassis. 24–96 kW charge power per module. 2 – 24 2, 4 or 8 per module; up to 3 modules per chassis Full stack CAN 2.0/CAN-FD with DBC import, SMBus, UDS; ArbinCTI and ArbinDriver; analog and digital I/O; MTCI chamber and chiller control; auxiliary voltage, temperature and pressure. Regenerates up to 90%.
Pack Testers
RBT-Pack Regenerative High-voltage regenerative pack cycler View Product → Pack-level lifecycle and end-of-line testing, BMS validation, and the standard drive cycles — FUDS, HPPC, DST, WLTP — at full pack voltage and power. 300 V – 1,500 V Two auto-selecting ranges per channel; charge from 0 V, discharge floor 20 V to 60 V by range. 50 A – 200 A and 60–200 kW per channel; 180–360 kW per chassis. Combined: 1,500 V, 3,000 A, 1.8 MW. 2 – 6 Per chassis, voltage dependent; chassis parallel for more Full stack Same interfaces as RBT-Module — CAN 2.0/CAN-FD with DBC import, SMBus, UDS, ArbinCTI, ArbinDriver, I/O and MTCI — plus Rint-model battery simulation for HIL. Regenerates up to 92%, anti-islanding in ~40 ms.

Product-family ranges are summarized for comparison. Available voltage, current, channel-count and power combinations depend on the selected configuration, and chassis output power is customizable in increments. Testing cells rather than modules? See LBTS-Cell, RBT-Cell, HPS and the rest of the cell range.

Regenerative Battery Testing Systems

Configure Your RBT System

Start by selecting the system type, then configure it step by step. Both series use regenerative circuitry that returns discharge energy to the facility grid.

Configure Your RBT-Module

Choose a channel module, voltage range, and chassis density. The system specification updates as you select.

1 Channel module

2 Voltage range

All ranges support 8 V minimum operation. Discharge to 0 V available upon request.

3 Modules per chassis

4 AC/DC power supplies per module

Each supply adds +12 kW charge and -11 kW discharge power. Available supply counts depend on the voltage range and chassis density.

RBT42082H
Channels per chassis
8
8 per module
Voltage per channel
8 V to 60 V
4-wire Kelvin sensing
Current ranges (±)
75 A, 10 A
Auto-switching
Charge power per module
96 kW
8 supplies × +12 kW
Discharge power per module
88 kW
8 supplies × -11 kW
Chassis charge power
96 kW
1 module installed
Regenerative efficiency
Up to 82%
Energy returned to grid
Max parallel current
Up to 600 A
All 8 channels paralleled

Part number: · Reference configuration:

* Charge and discharge power are calculated from the number of AC/DC supplies selected, at +12 kW charge and -11 kW discharge per supply. Parallel current assumes all channels in the chassis are paralleled, up to the 1,800 A chassis limit. Contact Arbin to confirm the final power allocation for a specific configuration.

Configure Your RBT-Pack

Choose a model, facility AC input power, voltage range, channel count, and total output power. The system specification updates as you select. Channels can be paralleled to increase current and power output, and RBT43012 and RBT44012 channels can also be configured in series to increase voltage.

1 Model

2 AC input power

Every system requires 2 facility power connections: the selected three-phase feed plus a 200 to 240 VAC single-phase feed. The input power also sets the top voltage range on RBT44012 (700 V or 750 V) and RBT46012 (1,400 V or 1,500 V).

3 Voltage range

4 Channels

Each channel is served by its own channel module.

5 Total output power

Shared power is one pool distributed across all channels; each channel can draw up to its per-channel maximum, and the pool limits the combined output. Dedicated power reserves the per-channel maximum for every channel simultaneously.

6 Chassis in parallel

Identical chassis can operate in parallel to reach up to 3,000 A and 1.8 MW combined. Paralleled chassis must share the same voltage range and power type, and each chassis requires its own three-phase and single-phase feed.

RBT45012
Voltage ranges
40 V to 1,000 V
Second range: 40 V to 500 V, auto-selected
Current ranges per channel (±)
100 A, 50 A
Auto-switching
Channels
2
Parallel capable
Max power per channel
100 kW
Charge or discharge
Total output power
100 kW
Shared between both channels
Facility power connections
440 to 520 VAC
3-Phase, plus 1-Phase 200 to 240 VAC
Max parallel current
Up to 200 A
All 2 channels at ±100 A
Full current available
Up to 500 V
Power-limited above this voltage
Regenerative efficiency
Up to 92%
Charge efficiency up to 94%

Part number: · Reference configuration:

* Calculated as total output power divided by the combined current of all channels in parallel. Above this voltage, output current is limited by the total output power rather than the channel current rating. Contact Arbin to confirm the final configuration.

Discharge to 0 V and limited negative voltage operation are supported with appropriate multi-channel configurations and control modes. Systems can be combined to reach up to 1,500 V, 3,000 A, and 1.8 MW.

Beyond the Channel

The complete test cell, not just the cycler

At module and pack scale the cycler is the smallest part of the decision. How the pack is connected, how the heat leaves the room, what the BMS is told, what commands the test, and what the building can supply all shape the result as much as the channel driving it. Arbin specifies all six layers together against your pack and test plan, rather than shipping a chassis and leaving the rest to the installation.

01
Electrical

Cycler & test channels

Channel selection follows the pack, not the catalogue: the voltage window, peak current, peak power and pulse requirement together decide the module type, how many fit in a chassis, and whether channels need combining in series or parallel to reach the duty.

Voltage window, peak current and peak power per channel
Series and parallel combination to 1,500 V, 3,000 A, 1.8 MW
Bipolar output, with zero switching time between charge and discharge
02
Connection

Busbars, cabling & fixtures

At hundreds of amps the thermal design of the connection stops being incidental, and contact resistance enters the measurement directly. Every channel senses 4-wire Kelvin, and the high-current interface is specified around the pack's own terminal layout.

4-wire Kelvin sensing on every channel
High-current terminal interface and cable routing
Pre-charge circuitry matches bus to pack voltage before the contactor closes
03
Thermal

Chamber & chiller integration

Modules and packs are conditioned rather than isolated, so the real question is how tightly the chamber and the coolant loop are tied to the schedule. Both are driven from inside the test and logged alongside the electrical data, so a thermal excursion appears in the same file as the current that caused it.

Third-party chamber control and synchronization via MTCI
Chiller and coolant-loop control through analog and digital I/O
Temperature-driven step control and conditional safety limits
04
Signals

BMS traffic & auxiliary measurement

A pack result that cannot be traced back to a cell group or a BMS state is hard to act on. CAN and SMBus traffic lands on the same timebase as voltage and current, and isolated auxiliary inputs add the in-pack electrical, thermal and mechanical picture alongside it.

CAN 2.0 / CAN-FD, SMBus and UDS, with DBC file import
Isolated auxiliary voltage for in-pack cell and group monitoring
Thermocouple, PT100, PT10k and pressure or force transducers
05
Control & data

What commands the test, and where data lands

Tests are written and run in MITS, and the same platform covers a 20 V module and a 1,500 V pack, so a validated schedule transfers without rework. Where a rig has to be part of something larger, ArbinCTI, ArbinDriver, Lua and lookup tables let your own software, a PLC or a model command the channel directly.

MITS Pro schedules, Simulation Editor for drive cycles
ArbinCTI over TCP/IP, ArbinDriver DLL, Lua, Custom Test Instructions
Direct SQL, Kafka streaming, CSV service, CDS and CMCS across the fleet
06
Facility

Power, safety & commissioning

At pack energies the tester is a piece of facility plant, and the safety chain has to work independently of whatever is commanding the channel. Limits are checked outside the test control loop, the E-Stop can be tripped from a facility dry contact, and regeneration disconnects itself if the mains misbehaves.

Three-phase 340–520 VAC, grid-tied regeneration, anti-islanding
Hardware E-Stop with all-pole disconnection, channel interlock, isolation contactors
Installation, commissioning, training, and ISO/IEC 17025-accredited calibration

Every layer is specified against your pack and your test system, not sold as a bundle. Send us what you are testing and how you intend to drive it, and an application engineer will come back with a configuration matched to both.

Contact Us
Next Step

Tell us what you are testing

Most configurations are settled in a single conversation, because the four things below narrow the options far faster than a catalogue does. Send them across and a sales engineer will come back with the system that fits, the interfaces it needs, and what it asks of your facility.

01

The module or pack

Nominal and maximum voltage, capacity, cell configuration, and how the terminals are laid out. If it has a BMS, tell us which bus it speaks on.

02

Current, power and profile

Peak current and peak power, whether they are continuous or pulsed, and which profiles you run — a WLTP cycle asks something different of a channel than a constant-current soak.

03

Software and data

Whether MITS runs the test on its own or has to work with your own software, a PLC or an HIL rig — plus which chambers need to be in the loop and where the data has to land.

04

Facility

AC supply and available breaker capacity, floor space and clearance, whether the site permits grid-tied regeneration, and how the test area is conditioned.

Scroll to Top