I need to design a H-Bridge circuit to drive a 24 volt brush style motor. The stall current is under 12 amps. I would like to use to-220 case style N channel parts as they are easily heat-sink mountable. I am having trouble finding mosfets that have a large enough Vgs-max yet are switchable with 3.3 volt logic. Any part number suggestions? Additionally would be open to suggestions on a gate drive IC, especially if it was I2c or SPI driven. I only need to switch the motor on for 45 seconds every half hour at full speed.

Thanks Much

 

I would like to use to-220 case style N channel parts

If you use N channel MOSFETs for the high side of the bridge, then you need a high-side driver to generate the required gate voltage, which must be higher than the supply voltage by the required Vgs to fully turn on the MOSFET.
Such drivers use some type of bootstrap (switched-capacitor) circuit to generate the required voltage.
I don't know of any offhand that work from I2C or SPI.

To turn on the bottom N-MOSFETs with 3.3V, then you need logic-level type MOSFETs with a maximum Vgs(th) of ≤2V.

 

I2C or SPI not fast enough for gate drivers - that's more appropriate for a motor-control IC that can drive external MOSFETs.

large enough Vgs yet are switchable with 3.3 volt logic

I didn't understand this comment - large enough Vgs for what reason? Or did you mean Vds? If you're thinking of driving a power MOSFET direct from an MCU GPIO port @ 3.3v I'd think again. Most MCU don't have enough drive capability to turn a MOSFET under that load on or off fast enough - you will incur significant switching losses. Gate drives exist for that reason. The venerable but still useful IR2110 is cheap & effective..

 

I2C or SPI not fast enough for gate drivers - that's more appropriate for a motor-control IC that can drive external MOSFETs.


I didn't understand this comment - large enough Vgs for what reason? Or did you mean Vds? If you're thinking of driving a power MOSFET direct from an MCU GPIO port @ 3.3v I'd think again. Most MCU don't have enough drive capability to turn a MOSFET under that load on or off fast enough - you will incur significant switching losses. Gate drives exist for that reason. The venerable but still useful IR2110 is cheap & effective..

Sorry, I edited my post to reflect that I meant Vgs(Max). Since the motor is 24Volts I need a Vgs(Max) rated above 24 correct?

 

Since the motor is 24Volts I need a Vgs(Max) rated above 24 correct?

You can use 10 or 12 volt zener diodes across the gate-source to clamp the drive below 24 volts.

 

I only need to switch the motor on for 45 seconds every half hour at full speed.

Why do you need a H bridge, are you reversing the motor?

 

. I only need to switch the motor on for 45 seconds every half hour at full speed.

If that's the case, and you don't mind the - side of the motor not being grounded and no reverse requirement, then only one MOSFET needed. However, finding a MOSFET in a TO220/TO262/TO247 case that will be sufficiently turned on at 3v and has a Vds>30v is tricky. Possible candidates are (wattage in brackets is estimated dissipation @ 3v based on spice model):
Vishay SUP50020EL (1.2W)
Infineon IRLB3036PbF (5W)
STM STP160N4LF6 (3.2W)




A better solution is to use a low-cost gate driver, eg the Microchip TC4420, which will enable a much wider range of MOSFETs and reduce the dissipation in Q1 to <1W for many devices, meaning no heatsink required.



or if you prefer a discrete option: