Back to Knowledge Base
Technology

How DC Fast Charging Works: Power Modules and Thermal Design

12 min read

Most DC charger failures trace back to heat and to the modules that generate it. The internal chain explains why two cabinets with identical ratings age very differently in the field.

A DC fast charger is a power conversion plant in a weatherproof box. It takes three phase AC from the grid, rectifies it, regulates the output to match whatever voltage the vehicle battery presents, and manages a high current connection to a pack whose demand changes continuously through a session.

This guide follows the power path from grid connection to connector, and identifies where the decisions that set service life are made.

The Chain From Grid to Battery

Two paths run in parallel through the cabinet. Power moves from the AC input through filtering and rectification to a DC bus, then through an output stage that adjusts voltage to the pack. Signals run from the charge controller to the vehicle over the connector, and from the controller to the network over OCPP.

Protection sits between them: contactors that isolate, sensors that verify, and interlocks that halt the sequence when a check fails. A charger that never starts has a signal or protection problem. One that aborts mid session usually has a power path problem.

Three Phase Input, Filtering, and Precharge

Commercial DC chargers draw from a three phase supply, commonly 380 to 480 volts depending on region, through a dedicated breaker and often a surge protection device. An EMI filter sits at the input to keep switching noise out of the grid, since the rectifier generates high frequency harmonics that would otherwise fail EMC limits.

Precharge circuitry limits inrush current when the DC bus capacitors first charge. Without it, closing the input contactor on an empty bus draws a spike that shortens contactor life and can trip upstream protection. Precharge resistors, a bypass contactor, and a bus voltage check are the standard arrangement, and a failed precharge resistor is a common cause of startup breaker trips.

AC to DC: Power Modules in Parallel

Rectification happens in power modules. A module typically contains an active front end that converts AC to a regulated DC bus, followed by an isolated DC-DC converter that produces the output. Modules commonly run between 15 kW and 40 kW each, so a 120 kW cabinet holds several of them, as in a DC-DC ultra fast charging unit.

Parallel modules introduce the central engineering problem: current sharing. A module delivering more than its share runs hotter and ages faster, and the failure cascades. Designs use droop control or a master-slave arrangement to balance load, but that balance drifts as modules age, and redundancy hides a single weak unit until a second module fails.

The DC Bus, the Output Stage, and Vehicle Voltage

The internal DC bus runs at roughly 700 to 800 volts, high enough to serve an 800 volt platform directly and to step down to a 400 volt pack. Output voltage follows the pack, because the battery decides its own terminal voltage as a function of state of charge, moving through a range of several hundred volts within one session.

Output current limit, not the power rating, is therefore often the real constraint. Power is voltage multiplied by current, so a 400 volt pack at a 500 amp limit reaches about 200 kW, while an 800 volt pack on the same limit reaches about 400 kW. No firmware setting changes that arithmetic.

The reverse case matters too: a charger whose maximum output voltage sits below the pack voltage may not charge an 800 volt vehicle at all.

SubsystemFunctionCommon failure point
AC input and EMI filterThree phase supply, harmonics suppression, surge protectionFilter capacitor aging, surge device degradation after repeated strikes
Precharge and input contactorLimit inrush current, connect and isolate the rectifier stageWelded or worn contacts, open precharge resistor
Power modulesAC to DC conversion and isolated output regulationSemiconductor wear from thermal cycling, internal fan failure
DC bus and output stageVoltage regulation across the full pack voltage rangeBus capacitor aging, current imbalance between parallel modules
MeteringMeasure delivered energy for billing and reportingCalibration drift, accuracy loss after a current sensor fault
Insulation monitoringVerify isolation before and during a session, detect leakageSensor drift in humid conditions, nuisance trips after water ingress
Cooling systemRemove heat from modules, busbars, and cablesClogged filters on air units, coolant loss or pump failure on liquid units

Metering, Insulation Monitoring, and Protection

Delivered energy is measured on the DC side by a meter holding a defined accuracy class, because in most markets that reading is the basis for billing. Before the output contactors close, the charger runs an insulation resistance test on cable and connector. During the session, insulation monitoring watches for leakage to earth, and residual current monitoring detects DC fault currents a breaker cannot see.

These are safety layers, not features to trade for cost. A charger that shortcuts the insulation test may appear to work and fail on a wet day or a vehicle with a marginal pack.

Control, Communication, and Session Management

A charge controller coordinates the sequence: vehicle handshake over the connector, insulation test, contactor closure, the power request loop, and orderly shutdown. Vehicle communication follows IEC 61851-23 for the safety sequence and ISO 15118 for higher level functions such as plug and charge, carried over powerline communication in the CCS family described by CharIN.

The controller also exposes the station to a backend over OCPP, where remote start, load limits, firmware updates, and fault reporting live. Controller firmware is the least visible part of a charger specification and one of the most expensive to get wrong, because failures there affect every session.

CCS vs NACS: EV fast charging infrastructure explained (The Kilowatts)

The video covers how the North American connector transition affects site design, which complements the hardware discussion above: connector decisions change cable assemblies, cooling requirements, and post layout.

Thermal Design: Air Cooling Versus Liquid Cooling

Conduction losses in modules, busbars, and cables turn into heat, and removing it separates a charger that lasts a decade from one that needs module replacement in year three. Air cooled cabinets use fans and filters, which are cheap to build and simple to service, but filters clog in dusty environments and fan power rises with ambient temperature.

Liquid cooled designs circulate coolant through cold plates on the modules and, in high current units, through the charging cables. Cooled cables carry far more current for the same conductor cross section, which is why very high power sessions use a cable a driver can lift. The trade is complexity: pumps, hoses, and coolant become service items, as covered in the liquid cooled ultra fast charging overview.

Derating behavior is the practical test of a thermal design. Every charger reduces output as internal temperature climbs, but a good unit holds rated power across the expected ambient range. Ask for the derating curve against ambient temperature, and check whether the brochure figure is quoted at 25 degrees Celsius or at 50. A unit rated at 25 degrees may deliver two thirds of that on a hot afternoon.

Constant Current and Constant Voltage

The battery management system chooses the curve, not the charger. The vehicle requests a voltage and current, and the charger delivers within its own limits. In the constant current phase the vehicle asks for a fixed current and power rises as pack voltage climbs. In the constant voltage phase it holds voltage at the pack maximum and tapers current down.

That taper is why power falls away sharply above roughly 60 to 80 percent state of charge, and why the charger must regulate smoothly across a wide output range. A cabinet that oscillates at low current produces aborted sessions on vehicles that taper early.

What This Means for Procurement

Rated power is the least informative number on a datasheet. The details that predict service life are:

  • The derating curve plotted against ambient temperature
  • The maximum output voltage and current window
  • The module count, redundancy strategy, and access for replacement
  • The cooling method and the service items it introduces

Ask for those four in writing rather than comparing headline figures. A DC fast charging station specified for a wide output voltage window serves both platform generations, while a split configuration lets one cabinet serve terminals matched to the traffic. The wider DC range is listed on the product pages.

What are the main components of a DC fast charger?+

An AC input stage with filtering and precharge, paralleled power modules that rectify and regulate, a DC bus and output stage with contactors, a DC energy meter, insulation monitoring, a charge controller, and a cooling system.

Are power modules the most common failure point?+

They are among the most common, along with cooling components. Modules fail from thermal cycling and internal fan wear, and redundancy masks their failures until a second module goes. Count and access for replacement matter more than nominal rating.

Is liquid cooling better than air cooling?+

Liquid cooling holds rated power better in high ambient temperatures and allows higher current through thinner cables. Air cooling is cheaper and simpler to service where air quality is good. Ambient temperature, dust, and required cable current decide.

Why does a 350 kW charger charge my car at 90 kW?+

The vehicle requests the power. Battery voltage, current acceptance, state of charge, and pack temperature all cap what a session can draw. A 400 volt pack on a current limited charger often cannot reach the headline figure.

Need Help Specifying a Charger?

Tell us your power requirements, site type, and target market. Our engineering team will come back with a configuration proposal and a quote.