I'm wanting to drive a high current load using a 12V PWM controlled by an Arduino. Since the load requires commonality with ground, I'm forced to use a P-Channel mosfet to drive it. In order to eliminate acoustic rattling, the PWM frequency has to be quite high. This means that I have to use a Gate Driver to reduce the phase transition time so the fet doesn't get too hot due to the time it spends in ohmic mode switching on and off. I think a suitable driver would be the IR4426 which is an inverting gate driver.
On the data sheet it shows an example schematic (it's for a non inverting application using N fets but the principle is the same). It has resistors between the FET gates and the IR442X outputs. Why is this? Wouldn't putting resistors there totally negate the point of using the driver in the first place as it would introduce a delay for the gate to charge/discharge? So naturally I'm a little confused now. Could anyone clear it up as to why those resistors have to be there?
Now that is what I'm not getting. The output from the driver is inverted, so when I apply 0V to the input, I should get VCC out of the output (open circuit low side is off) which turns off the mosfet, and when I apply a logic voltage to the input I should get 0V at the output (low side is on so connects to ground) which turns the mosfet on. Or am I getting all this back to front?
I have bread boarded this without using the resistors (output to gate directly) and it does seem to work.
I may be wrong in my assumption, so please correct me if so. A low side driver switches a an open voltage to ground and a high side switches an open ground to a voltage?
Basically the gate driver has to do a better job than this
The resistors to the Gates of the MOSFETs are there to limit current, because the Gate of a MOSFET has a very high "capacitance". More correctly stated, the Gate takes a large amount of charge. So without current limiting somewhere (in the driver or external to it), the current can spike very high.
Those resistors can also help prevent ringing or oscillation.
I do understand how the circuit works, I'm just not up on the jargon. I thought low-side meant the switch (Q2) was on the low (ground) side of the load and inverted because the gate of the mosfet is at an inverted logic to the base of the transistor. It's one thing to understand circuits and another to know the terminology.I guess, it seems very counter-intuitive to me. So High Side would be clearer to understand if they said the LOAD is on the High Side rather than the SWITCH being on the high side as the switch by definition is actually on BOTH, the load can only be on one. I still don't get why it's non-inverting though, as the logic is being reversed by the transistor. Is the term is relative to the device it's driving? If I used the same circuit to drive an N-Channel would it then become an inverting high side?
I'm still however a little confused, now I understand the resistors are there to protect from these high currents, but again, wouldn't they mess up the driver's function? The gate would then have to charge and discharge through those resistors and introduce a delay making it no better than using the transistor circuit above? Also if the spiking current is high wouldn't those resistors burn out if they were regular 0.25W ones?
polymorph:
High side: On the "high" (upper) end of the load.
The resistors to the Gates of the MOSFETs are there to limit current, because the Gate of a MOSFET has a very high "capacitance". More correctly stated, the Gate takes a large amount of charge. So without current limiting somewhere (in the driver or external to it), the current can spike very high.
Those resistors can also help prevent ringing or oscillation.
Lets clear this up. Gate resistors are not normally required when using
a MOSFET driver driving one MOSFET. You can select the driver that
outputs enough current to switch your MOSFET as fast as you want.
You normally have to add gate resistors if driving two or more MOSFETs from the
same driver, because there is a differential oscillation mode in this configuration
that can occur and needs damping down. Resistors also have the role of allowing
devices with different plateau voltages to switch in unison (plateau and threshold
voltage spread can be as much as +/-1V).
You can add a resistor to slow down switching time in order to reduce
high frequency EMI from the switching circuit, its basically a compromise
between switching efficiency and EMI reduction.
You can add a resistor and a diode in order to make switching asymmetrical
(typically the diode conducts to discharge, causing a faster switch-off than
switch-on, as a safeguard against shoot-through).
Many common MOSFET drivers have an output resistance of a few ohms to a few
dozen ohms and typically need no resistors if driving only a single MOSFET and EMI
is not a concern.
If you do add gate resistors because of paralleled devices, the resistor should
to be non-inductive (not wire-wound) and right next to the gates.
I am indeed only driving one Mosfet so I won't bother using a resistor. From what I can tell an inverting low side mosfet driver will do the same job as a non-inverting high side. Essentially, they both output VCC when input is low and the output becomes 0V when input is high yes?
I thought low-side meant the switch (Q2) was on the low (ground) side of the load
That -is- what a low-side driver is. The switch is on the low side. What you drew is a high side driver.
If the switch is between V+ and the Load, it is a High Side switch. If the switch is between the Load and Ground, it is a Low Side switch.
and inverted because the gate of the mosfet is at an inverted logic to the base of the transistor.
All that matters: When the Input is ON, the load current is ON, then it is non-inverting. If the Input is ON and the load current is OFF, it is inverting.
The word "load" is ambiguous to me. There's 2 loads in that circuit, R2 and R1. The switch cannot be between R1 and anything because R1 is not part of that side of the circuit and R2, well the switch is between R2 and ground. See why I'm confused?
Forget it, the terminology is too ambiguous and counter-intuitive for me. I'll stick to programming, it makes more sense.
This is the problem in this field for beginners. Everyone automatically assumes the beginner is going to understand every bit of jargon and terminology. I can tell you on a physical quantum scale why a mosfet works, I understand doping etc.. but because I don't know the definition of "high side" or "low side" and/or mis-interpret it, I'm buggered because no where there is a real unambiguous definition. Like in this example, the SWITCH is the mosfet, where as I thought it meant the transistor, the load is the final load whereas I thought it meant the pullup resistor. That kind of detail is never covered just the terms "switch" and "load". I mis-interpretted because when I imagine a driver, I don't see the final load, all I see is the transistor and pull-up, that to me IS the driver and so I took the parts of the driver itself as what switch and load referred to.
But anyway, thank you both for clearing it up. Now I actually do understand.
That's life. The pullup resistor -is- a load on Q1, but -the- load of interest is R1. That's the point of this whole exercise, right? How to take an Arduino output and drive a 1 ohm load at 12V.
polymorph:
That's the point of this whole exercise, right?
No need to get snarky, I thanked you for your help. You have explained what I wanted to know and cleared up my misunderstandings. There is no need to re-assert the points which I previously stated I now understood.
But again, many thanks for your help and to all who posted. It's appreciated. And sorry for "Reading My Attitude" into your message there polymorph. I'll see to it it never happens again. Of course you weren't being "snarky" as I accused you of, that was poor judgement on my part. I must have been feeling snarky at the time and misunderstood you.
I'm trying to learn programming, having been an electronics tech for most of my life. It takes a whole new way of looking at things. Learning electronics is probably a bit like having to learn a lot about nearly every programming language, at the same time.