Why On-Board Chargers Derate — and How to Specify Around It
Derating is the difference between the power printed on the front of the datasheet and the power the charger will actually deliver in your vehicle on a hot afternoon. It is not a defect and it is not dishonesty: it is the thermal and electrical reality of a converter that has to survive its worst case. The problem is that most specifications quote one number and never say which conditions produced it.
Key takeaways
- Ask for the derating curve against ambient and coolant temperature, not just the headline power.
- Input voltage matters: at 90 V AC the input current for the same output power is roughly 2.9× the current at 264 V AC, and conduction losses follow.
- Continuous, 30-minute and peak ratings are three different products.
- Altitude above 2,000 m reduces air cooling capacity; a fanless design derates earlier.
The four variables that move the number
Ambient and coolant temperature
Every converter has a maximum internal junction temperature. As the surrounding air or coolant gets hotter, the heatsink has less temperature difference to work with, so the controller reduces output to hold the junction inside its limit. A typical curve is flat to 45 °C, then falls roughly linearly to about 50–60% of rated power at 65 °C. Two suppliers quoting “6.6 kW” can be quoting the flat part and the knee respectively.
Input voltage
Power factor correction stages draw a current proportional to 1/Vin. At 230 V AC a 6.6 kW charger pulls about 29 A; at 90 V AC the same power needs about 73 A. Conduction losses scale with current squared, so an input at the bottom of the window is a thermally harder operating point than the nominal one — even though the output is identical.
Duty cycle
A 30-minute rating assumes the unit starts at ambient and is allowed to cool afterwards. A vehicle that charges overnight for eight hours is a continuous application and must be specified against the continuous curve. Mixing the two is the most common cause of “it works in testing and throttles in the field”.
Altitude and enclosure
Above roughly 2,000 m the air is thin enough that convection and fan cooling both lose capacity. Inside a sealed enclosure with no airflow, the local ambient the charger sees can be 15–20 °C above the vehicle ambient — which moves the operating point along the curve without anyone changing a specification.
| Condition | What to request from the supplier | Why it matters |
|---|---|---|
| Ambient 45 °C, still air | Continuous output power, and the junction temperature margin | This is the number that decides whether the vehicle charges at the advertised rate in summer |
| Coolant 65 °C | Derating curve against coolant temperature | Liquid-cooled units inherit the vehicle’s coolant loop; a hot loop is the normal case, not the exception |
| Input 90 V AC | Efficiency and thermal behaviour at the bottom of the window | Weak grids and long extension leads put real installations at the low end |
| 8-hour charge | Continuous rating, not 30-minute | Overnight charging is the dominant use case for low-speed and commercial EVs |
| 2,000 m+ altitude | Altitude derating factor | Air-cooled units lose capacity; the label does not mention it |
Why fanless designs derate differently
A fan-cooled unit has a flat curve until the fan reaches its limit, then falls quickly. A fanless unit depends entirely on conduction into its mounting surface and natural convection from its fins, so its curve is gentler but starts earlier, and it is far more sensitive to how the unit is mounted. A fanless charger bolted to a painted panel with thermal paste omitted behaves like a much smaller charger. This is why the installation instructions are part of the thermal specification, not an afterthought.
Practical rule: size the charger for the power you need at the temperature you will actually see, then add 15–20% headroom. Sizing at the headline number and hoping is how programmes end up with a vehicle that charges at 3.3 kW on a 6.6 kW label.
Reading a derating claim critically
Three questions expose most weak claims:
“At what ambient?” If the answer is 25 °C, the number describes a laboratory, not a vehicle.
“For how long?” If the answer is “peak”, ask for the continuous figure.
“Mounted how?” Free-standing on a bench, or bolted to a plate inside an enclosure? The second is the real case and it is usually 10–20% lower.
Grapcharger publishes the operating window rather than a single point: the OBC range runs from 700 W to 6.6 kW on a single-phase 90–264 V AC input, and the 3.3 kW unit exists precisely because many platforms need the derated margin more than they need the headline 6.6 kW. The thermal management note explains the conduction path in more detail.
Frequently asked
Why does my charger only deliver 4 kW when it is rated 6.6 kW?
Almost always one of three things: the ambient is above the flat part of the curve, the input voltage is at the low end of the window, or the mounting surface is not conducting heat away. Check the input voltage and the mounting first — they are the cheapest to fix.
Is a 30-minute rating acceptable for a vehicle charger?
Only if the vehicle genuinely charges in 30-minute bursts. Overnight charging is continuous, and a 30-minute rating will throttle part-way through.
Does liquid cooling remove derating entirely?
No. It moves the limit from air temperature to coolant temperature, and vehicle coolant loops run hot. The curve still exists, it is just drawn against a different axis.
Specifying a platform?
Send the pack voltage and the target power.
We reply with a model, a derating curve and the compliance documents that go with it.
Keep reading
Grapcharger designs and manufactures on-board chargers, DC-DC converters and integrated charging units from 700 W to 6.6 kW for low-speed EVs, commercial vehicles, marine and industrial platforms, and supplies vehicle manufacturers and distributors worldwide.