How much does a driver heat up and how is the heat loss calculated?
Conduction loss follows P = I² x RDS(on): for 15 A and 10 milliohms that is 2.25 W per switch. Double the current and the loss becomes four times as large.
In a MOSFET driver the heat comes from two sources: conduction loss and switching loss. Conduction loss is calculated as P = I² x RDS(on); for 15 A and 10 mΩ that gives 15² x 0.010 = 2.25 W per switch. At the same resistance the loss climbs to 9 W at 30 A: twice the current, four times the loss.
The second item: switching loss
At the transition moments when the switch turns on and off, both voltage and current are present across it; a small amount of energy turns into heat at every transition. That energy is multiplied by how many times a second the switching happens, so switching loss grows directly with frequency. A high PWM frequency makes the motor quieter but extracts a price in heat; this trade between quietness and temperature is the reason not to raise the frequency arbitrarily.
The most common mistake: using the 25 °C value
The RDS(on) figure that stands out in datasheets is for a junction temperature of 25 °C. RDS(on) rises noticeably with temperature; as the driver heats up the resistance grows, the loss grows, and the temperature climbs further still. The correct approach is to read the value at the operating temperature from the datasheet's normalised RDS(on) versus junction temperature curve. A calculation made with the room-temperature value understates both the loss and the junction temperature.
The heat has to leave the body
The calculation gives you the loss, not the temperature. What sets the temperature is how easily the heat produced is passed to the surroundings: the mounting surface, air circulation and, where needed, a heatsink. A driver squeezed inside a closed box runs far hotter at the same current than one sitting in the open. Whether the driver needs a separate cooler depends less on the current than on this mounting condition.
Ambient temperature enters the calculation directly as well: if it is hot outside, less temperature difference is left over to shed the same heat. Reducing the current capacity at high ambient temperature is therefore the rule, not the exception.
In the field you usually do not have the RDS(on) data to hand. In that case observation works better than calculation: run the driver on the heaviest duty cycle and watch the body temperature. If the temperature rises over the first few minutes and then settles at a value, thermal equilibrium has been reached; if it keeps rising, there is no equilibrium and the mounting condition is inadequate. If you are going to compare, repeat the measurement at the same ambient temperature and with the same load.
No numerical loss figure can be given for the AXI modules
An important note: the MOSFET type, RDS(on) and efficiency figures for the ST42, KS250 and PT500 are not found in the product data, and for that reason no numerical loss figure can be attributed to these modules. The 15 A and 10 mΩ example above is there to show the method; it is not a measured value for these products. On these modules the only reliable numbers are the continuous and peak currents given in the catalogue; do the sizing over those.
Steps that lower the heat in the field
- Run the driver in the 70-80% band of its continuous current rating; working at the limit leaves no thermal margin.
- Leave air circulation at the mounting location, and do not bury the driver in a tight, insulating cavity.
- Tighten the terminals to their torque; a loose connection is a source of heating in its own right.
- Do not choose a higher PWM frequency than you need.
- If the driver gets too hot to touch, measure the current; the problem is usually not in the calculation but in the mechanics or the sizing.
The full topic: DC Motor Driver Selection Guide