# High side N channel gate driver

So I realized with the suggestion in the comments that there is a much better solution to driving a higher voltage H bridge than using transistors or secondary mosfets and zener voltage. I have revamped the question. I am still pretty green to all of this stuff so maybe if anyone has a moment they can check over the schematic I have drawn for any other suggestions or problems with it. I'll post data sheet links to the parts below if you want to look at them. I realize that some of these parts like the 600v diode rectifiers are way over kill, but I have most of them parts on hand except the drivers.

The question(s) I have now is as follows.

1. the datasheet schem I started with showed a half bridge without the flyback diodes but has the capacitor 5 in the schematic I have drawn. I do believe that the datasheet schem was not an inductive load so I am guessing that is why they did not have the flyback diodes. So do I need this capacitor in my case if I have the flyback diodes, or is it a good idea to have it there anyway for reasons I do not yet understand with my wee bit of knowledge?

2. As far as operating this bridge, my plan was to pwm the high side with arduino uno, and just put the corresponding low side either high or low. Example: pin 3 pwm value, pin 2 high, pin 5 & 4 low for reverse. And then pin 5 pwm value, pin 4 high, pin 3 & 2 low for forward. In my research since I discovered the high side boost drivers yesterday I found some discussion on the capacitors not recovering quick enough but the circuit was a much higher frequency switching circuit like over 100KHZ. My question on this is, should I pwm the low side MOSFETS instead, or will the capacitors recover quick enough with a 480HZ or max 980HZ pwm cycle of the arduino pin depending on the pin.

3. My last question, should I run both channels on the driver at the same time so 3 and 2 would be on driver 1, and 5 and 4 on driver 2? Or is it fine to run the high on one driver, and the low on the other driver like I have drawn in the schem.?

200V capacitors: https://www.mouser.com/ds/2/293/e-uvz-1219460.pdf

             https://www.mouser.com/ds/2/212/KEM_A4018_ELH-1104321.pdf

• This is why half bridges tend to be Nch with high side boost regulator from low side PWM – Tony Stewart Sunnyskyguy EE75 Mar 1 '18 at 22:27
• In #1 you are putting 60V across a 12V zener with no current limiting. – τεκ Mar 2 '18 at 0:15
• In #2 the zener does not do anything. Vgs is 60 volts and will punch through the gate breaking the MOSFET. – τεκ Mar 2 '18 at 0:17
• In #3 turn off time is slow because all the gate charge has to go through the 1k resistor. – τεκ Mar 2 '18 at 0:18
• In #3 the source potential is a lot higher (60V) than the maximum gate potential (12V), so the high-side MOSFETs won't be able to turn off at all. – Turbo J Mar 2 '18 at 15:39

I do believe that the datasheet schem was not an inductive load so I am guessing that is why they did not have the flyback diodes.

The MOSFETs have 'body' diodes built in, so external flyback diodes are not required. They can be added if you want to reduce heating in the FETs (use Schottky diodes for lower voltage drop).

capacitor 5... do I need this capacitor ?

Yes. It reduces voltage ripple on the power rail caused by the current pulses drawn by the motor. It is especially important if the wires between the power supply and controller are long, as their inductance could cause the voltage to spike high enough to blow up the MOSFETs.

I found some discussion on the capacitors not recovering quick enough... should I pwm the low side MOSFETS instead ?

With this scheme you must PWM the high-side MOSFETs - otherwise they will not turn on fully because with no PWM pulses the bootstrap capacitors will discharge. You must keep the PWM ratio sufficiently below 100% to ensure that the capacitors remain charged whenever the upper FETs are turned on.

If you want a full 0-100% PWM range then you will need drivers with internal voltage boosters (eg. LTC7004), or use isolated DC/DC converters to provide 12-15V to the high-side drivers. Then you could turn the upper FETs on continuously and PWM the lower FETs.

will the capacitors recover quick enough with a 480HZ or max 980HZ pwm ?

At 480-960Hz the capacitors should have no trouble staying charged up because the average charge transferred to the FET Gates is small. However at this low frequency motor inductance probably won't be enough to maintain continuous current flow through the windings, which will cause increased vibration and motor heating. Most motors need a PWM frequency >=3kHz for efficient operation.

should I run both channels on the driver at the same time so 3 and 2 would be on driver 1, and 5 and 4 on driver 2? Or is it fine to run the high on one driver, and the low on the other driver like I have drawn in the schem.?

To drive the motor you need to turn on M1 and M4 in one direction, or M2 and M3 in the other direction. If two FETs on the same side are turned on at the same time then current will 'shoot through' them and short out the power supply. Shoot-through should be avoided at all costs because it will eventually blow up the FETs.

When testing your firmware you should use a current limited power supply, or insert a current limiting device (eg. incandescent light bulb) between the power supply and the controller. This will save your FETs if a software glitch causes them to be turned on at the wrong time.

To brake the motor turn on both lower FETs. This puts a short cross the motor so it generates maximum current and braking torque.

• Thanks, that pretty much clarifies everything I wasn't quite sure on. I maybe should have worded that last question a little different, but my question was does it make a difference if i run M1 and M4 on the same driver or if I run M1 on one driver and M4 on the other driver and in either situation M1 would obviously be the high side out of the driver. – LonnyM Mar 3 '18 at 15:02