In the choice between monolithic and gate drivers with external MOSFETs to drive motors, cost, size and thermal issues are key considerations.
When designing a motor control circuit, it’s vital to determine how to deliver the high current required to drive the motor. Designers must choose whether to use monolithic integrated circuits (ICs) that have internal power devices, or gate driver ICs and discrete external power MOSFETs.
This article discusses the advantages and disadvantages of each approach and provides guidance on when to choose either solution.
The first option is to use a monolithic driver IC to drive a motor. An integrated IC is comprised of one silicon die in a package; this die integrates logic, support, and protection circuitry, as well as the power devices like MOSFETs that drive the current through the motor.
Because the MOSFETs in a monolithic solution are fabricated on the same die as the control circuitry, these solutions provide the benefit of accurate current measurement. Monolithic ICs also provide robust protection features, such as over-current protection (OCP) and over-temperature protection (OTP), since this circuits can be placed in close proximity to the MOSFETs on the silicon.
Integrated drivers are limited to voltage and current ratings that are compatible with IC processes, which means that the highest available voltage rating is between 80V and 100V. In addition, these drivers can drive up to about 15A.
Monolithic drivers are used almost exclusively in high-volume applications such as printers, where supply voltages are typically below 35V and motor currents are below 5A.
An example of an integrated driver is MPQ6541, an automotive-specified, 3-channel power stage device. It’s rated at a supply voltage of up to 45V and a continuous load current of 8 A, or a peak current of 15A per channel. This motor drive integrates six MOSFETs that have an RDS(ON) of 15mΩ each. It comes in a TQFN-26, 6-by-5mm flip-chip package.
Figure 1 The block diagram highlights the key building blocks of an integrated driver. Source: Monolithic Power Systems
Figure 1 shows the MPQ6541’s block diagram. Note that it integrates current measurement for each channel. This eliminates the need for large and expensive current-sensing resistors or current-sense amplifiers.
The second option uses discrete power MOSFETs—or in some cases, other power devices—to drive the current through the motor, and the MOSFETs are controlled via a gate driver IC, pre-driver, or multiple gate drivers.
Monolithic solutions do not exist for applications that require high voltages exceeding 100 V or very high currents. In these cases, a gate driver, plus discrete MOSFETs, are required.
Since multiple devices are required in this scenario—sometimes as many as three gate drivers and six power MOSFETs—the solution size, which is the PCB area occupied by the motor driver, is much larger than what’s required by a monolithic driver.
An example of a highly integrated gate driver is the MPQ6533, a 3-channel gate driver IC with integrated features like slew rate control and internal diagnostic functions. This device is available in a 5-by-5mm QFN-32 package.
Figure 2 The block diagram highlights the key building blocks of a gate driver. Source: Monolithic Power Systems
Figure 2 shows the MPQ6533’s block diagram. Note that this solution requires six power MOSFETs. Generally, three dual MOSFETs (two MOSFETs packaged together into one IC package) are used.
Analog and mixed-signal IC processes are much more complex than dedicated discrete MOSFET processes. Since fabricating low RDS(ON) MOSFETs in an IC process takes a large area of silicon, the cost of a device with the same RDS(ON) and voltage in a MOSFET process is usually higher than it would be to fabricate a similar device in a dedicated discrete MOSFET process.
For lower-current and/or lower-voltage motor drivers, the penalty for fabricating the MOSFETs in the IC process is small. Since the control and protection functions take up a big part of the die, the added area for the MOSFETs does not increase the cost as much as using external MOSFETs.
For high-current applications, however, the cost of the MOSFETs in an IC process starts to dominate the cost of the device. Even though there are monolithic motor drivers that can support a 15-A motor current, they typically are more expensive than an implementation using a gate driver plus discrete MOSFETs.
There are cases where the small size of a monolithic part is valued so highly that it justifies a more expensive solution. For example, some systems require an integrated driver inside a motor, where there is little space available. In these scenarios, a solution using a gate driver plus MOSFETs simply may not fit in the constrained space.
To get a rough idea of the relative cost of monolithic solutions vs. a gate driver solution, we can compare the cost of a monolithic IC plus a gate driver IC with three dual MOSFETs and three current-sense resistors. Other supporting components, such as bypass capacitors, have similar prices between both solutions. Note that these costs are based on low quantity catalog prices; actual volume production prices are typically much lower.
Table 1 A cost comparison is shown between a dedicated monolithic IC and gate driver IC. Source: Monolithic Power Systems
Table 1 shows the cost comparison between a dedicated monolithic IC and gate driver IC with discrete MOSFETs.
Monolithic drivers are almost always smaller than the equivalent solution using gate drivers and discrete MOSFETs.
As an example, we can compare the PCB area occupied by the MPQ6541 to the MPQ6533 with additional power MOSFETs (Figure 3). Both parts significantly differ in size, with the MPQ6541 occupying 130mm2 and the MPQ6533 occupying 520mm2, which is four times larger. Note that the gate driver solution shown here uses dual MOSFETs in small packages; in other cases, the MOSFETs can be much larger, which further increases the solution’s PCB area.
Figure 3 A size comparison is shown between a dedicated monolithic IC and gate driver IC. Source: Monolithic Power Systems
To effectively dissipate the heat that is generated in the power MOSFETs, the PCB is typically used as a heatsink. Larger packages typically have better thermal conductivity to the PCB, which means that bigger solutions are better from a thermal dissipation standpoint. This can work in favor of solutions that use gate drivers, since the power MOSFETs are typically large. Low RDS(ON) power MOSFETs are readily available, so in some cases—especially with applications that need to operate in harsh environments—thermal considerations may preclude the use of a monolithic driver.
Monolithic drivers come in smaller packages. To compensate for the higher thermal resistance in these packages, the RDS(ON) for a given current must be lower than it would be for a comparable solution using a discrete MOSFET.
Consider the MPQ6541 monolithic driver and its smaller size. If the PCB is designed correctly, a significant current can be driven by this part. Figure 4 shows the MPQ6541’s temperature on a 5 cm x 5 cm, 2-layer PCB, while delivering 6 A of current to a 3-phase brushless motor. The case temperature measured was 38°C above the ambient temperature. A 4-layer PCB with internal planes would further lower the temperature rise.
Figure 4 The thermal image of a monolithic driver highlights temperature conditions. Source: Monolithic Power Systems
Carefully consider tradeoffs
Table 2 summarizes the main differences between integrated and gate driver solutions.
Table 2 A part comparison highlights major differences between a dedicated monolithic IC and gate driver IC. Source: Monolithic Power Systems
The choice between monolithic motor drivers and gate drivers with external MOSFET solutions to drive motors is complex. Tradeoffs between cost, solution size, and thermal characteristics must be considered.
For very small motors, monolithic drivers are the best solution. Likewise, for very high-power motors, a solution using gate drivers and discrete MOSFETs should be used. However, there is a large overlap between both solutions, so designers should consider the specifications of their application when making their choice.
This article was originally published on EDN.
Pete Millet is staff technical marketing engineer at Monolithic Power Systems (MPS).