Hie again. i am about to place an order for some n channel mosfets for my h bridge and there is a few things id like to verify.
i will be making a 4wd robot with skid steer configuration using 4 motors, so there will be 2 h bridges, one controlling 2 motors.
Since my motors stall at 42A each, will a 120A for each h bridge be enough? or is 80A enough since these motors are quite impossible to stall?
here is my chosen Fet
To be safe & reliable I'd recommend designing it to handle the stall current plus some safety margin. And, the starting (and reversing) current may be the same as (or close to ) the stall current (depending on inertia and the mechanical load).
4wd robot with skid steer configuration using 4 motors, so there will be 2 h bridges, one controlling 2 motors.
Since my motors stall at 42A each, will a 120A for each h bridge be enough? or is 80A enough since these motors are quite impossible to stall?
If you are running two 42 Amp motors in parallel from one bridge, that's 84 Amps. So, I'd go with the 120A part.
In addition to the current rating, there is power dissipation (heat). So I would assume you'll need a hefty heatsink.
If you need more current capability, just parallel a handful of the MOSFETs.
or is 80A enough since these motors are quite impossible to stall?
When a motor starts up, it briefly draws close to the stall current. If it happens to be rotating and the direction of rotation is suddenly changed, the motor can briefly draw TWICE the stall current.
Although the current rating of the mosfet needs to be greater than the motor stall current, what really matters is the Rds(on). If it isn't low enough then you will struggle to keep the mosfet cool. I presume you will be using high/low side mosfet driver chips that provide at least 10V gate drive. So look for a mosfet with a low Rds(on) quoted when Vgs = 10V and Id = 42A or more.
The mosfet you linked to looks has Rds(on) of 2.5 milliohms max @ Vgs=10V and Ids = 100A. When passing 42A it will dissipate 42 * 42 * 0.0025 = 4.41W of power, which isn't impossible to get rid of with a good heatsink. However, if you use 2 of them in parallel (each with its own gate resistor of about 10 ohms), then each will dissipate about 1.1W of power, and you may not need any heatsinks at all if air can flow freely around the mosfets.
jremington:
When a motor starts up, it briefly draws close to the stall current. If it happens to be rotating and the direction of rotation is suddenly changed, the motor can briefly draw TWICE the stall current.
IMO it's not worth trying to build an 84A H-bridge, it's much easier to design the software so that it is impossible to abruptly change the drive from full PWM in one direction to full PWM in the other direction.
Even so at those current levels a chunky heatsink is a requirement (assume stall
current will flow if only briefly and make sure the system has enough thermal mass
to tolerate this for several seconds)
You should be monitoring current levels in software and backing off the drive if the
current is too high - in effect limiting the continuous current flow to prevent overheating.
Concur about the MOSFET ratings - make sure Vds is at least twice the load (inductive
spikes) and the Rds(on) is low enough - the max voltage across the MOSFET when on should
be a small fraction of your supply voltage, and definitely less than a couple of volts
(or you'll risk taking the device out of fully-on state - the drain must be well below the
gate voltage when on).
Calculate the power dissipation with I-squared-Ron when selecting a heatsink. If you
detect over-current then the heatsink can be rated just for the max continuous current,
but a heavier heatsink will tolerate overcurrent for longer. You can mount a temperature
sensor on the heatsink too if you want.