I am supplying a TC4420 gate driver (https://bit.ly/2RiKnXZ) with a pulse from a signal generator at 1,440 hz. The problem is I can't get the gate driver much over 5v without it getting quite warm (105^F). The current draw from the bench supply at 5.25v is 115mA. The gate driver is rated at max 18v and 6A so I though I would be able to put more power into it. I'm connecting the circuit as shown in the attached diagram.

 

The problem is I can't get the gate driver much over 5v without it getting quite warm (105^F). The current draw from the bench supply at 5.25v is 115mA. The gate driver is rated at max 18v and 6A so I though I would be able to put more power into it.

MOSFET driver chips are designed to deliver high currents, but only in extremely short bursts (a few hundred nanoseconds at most) to charge and discharge MOSFET gates quickly. So even though the TC4420 is rated at 6 amps maximum output current, it isn't designed to deliver that much current continuously because of its typical output resistance of 2.5 Ω. Since the load you're attempting to drive is only a few ohms, the chip will dissipate significant power and that's why it's getting warm.

Your best course of action would probably be to use the TC4420 to drive the gate of an N-channel power MOSFET configured as a low-side switch, and let the MOSFET drive the motor.

 

MOSFET driver chips are designed to deliver high currents, but only in extremely short bursts (a few hundred nanoseconds at most) to charge and discharge MOSFET gates quickly. So even though the TC4420 is rated at 6 amps maximum output current, it isn't designed to deliver that much current continuously because of its typical output resistance of 2.5 Ω. Since the load you're attempting to drive is only a few ohms, the chip will dissipate significant power and that's why it's getting warm.

Your best course of action would probably be to use the TC4420 to drive the gate of an N-channel power MOSFET configured as a low-side switch, and let the MOSFET drive the motor.

Yup. And with a thermal resistance of 125°C/W (DIP version), it's not going to take a lot of power to heat it up significantly.

 

So even though the TC4420 is rated at 6 amps maximum output current, it isn't designed to deliver that much current continuously because of its typical output resistance of 2.5 Ω. Since the load you're attempting to drive is only a few ohms, the chip will dissipate significant power and that's why it's getting warm.

So I presume this is the case even at a 50% duty cycle? Frequency is quite low at 1.4khz.

 

So I presume this is the case even at a 50% duty cycle?

The power dissipated by the chip will be

Pd(avg) = duty cycle * (load current)^2 * chip output resistance​

and the chip internal temperature will be

Tj = ambient temperature + Pd(avg) * θja​

so the junction temperature can be calculated and compared with the part's rated maximum (150 °C). If you're below that, you're good to go-- even though, at that temperature, the chip will be EXTREMELY hot. For reliability and long life, being a conservative sort my choice would be to stick to no more than 100 °C.

Note that the thermal resistance from junction to ambient, θja, is different for different IC packages.

Frequency is quite low at 1.4khz.

True; the heating you're observing isn't being caused by switching frequency, but rather by simple I^2R dissipation.

 

Thanks for your input. I'll add a mosfet.
I'm not good at reading datasheets... just curious what the other temperature rance letters (C,I,E,V) refer to?

 

I'm not good at reading datasheets... just curious what the other temperature rance letters (C,I,E,V) refer to?

They refer to the four different grades (operating temperature ranges) the device is available in. See page 19 of the datasheet:

 

Ah right, thanks for pointing that out.
Cheers.