Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Deadband circuit or mosfet driver for P+N (with 100% duty)

Status
Not open for further replies.

ACharnley

Member
Hi,

I'm currently scratching my head over a problem and wondering if I'm missing a trick.

I have three of the attached circuits, all work at 5v. I'm using a PMOS on the high side although I do now have a 10-12v supply available so NMOS is an option.

The input is not PWM based but is a hysteric feedback loop. For deadband I feed the output from the comparator into a PIC16 and use three CWG peripherals to generate the driving signals with deadband.

I'd rather not use the MCU to do this as I've also a PIC32 and it's being troublesome to manage the communication overhead and other aspects.

All the MOSFET drivers I'm finding are for NMOS high side with a floating driver. My requirement is up to 100% duty.

Can anyone suggest;

1. a deadband circuit?
2. a driver which allows for a PMOS on the highside or an NMOS with a driving voltage input (no bootstrap capacitor for floating driver).

Cheers,
Andrew
 

Attachments

  • Screenshot_2019-10-24_21-41-30.png
    Screenshot_2019-10-24_21-41-30.png
    22.1 KB · Views: 154
Look for a half-bridge MOSFET driver with a charge-pump for the upper N-MOSFET that can operate at DC, such as this.
I believe they have a built-in deadband circuit.
 
Doesn't seem to be many around, plus a PIC16 is $0.74, that one is over $2 and I need three of them.
That may be.
But you wanted a DC high-side driver for the MOSFET and the PIC can't do that.
 
Not directly but the parts to do the driver shown in the first post cost $0.03 all in.

I found a deadband circuit using a 74 series AND+OR, but once factoring in the cost of those, cost of the driver IC's, plus the floorspace it made no sense to go with anything but the PIC even if I didn't use it's CPU.
 
Hi Andrew. YOu might be able to get away with not using a FET driver. It depends on the amount of current the FETs need to pass. If it is a high current (anything more than about 10A), the PIC might not have enough current drive to turn the FETs on quickly enough and the FETs will run hot. For low current output and 100% duty cycle, a high side PFET should be good enough with a logic circuit (and its prop delay) to give you the anti shoot through. The LTC4440 is an excellent FET driver if your FETs need to pass high currents. You can use this to drive an NFET on the low side (short out the diode and remove the capacitor as you dont need the internal charge pump). Use another one, similarly wired, to drive the PFET.
 
Look for a half-bridge MOSFET driver with a charge-pump for the upper N-MOSFET that can operate at DC, such as this.
I believe they have a built-in deadband circuit.
A charge pump based driver won't do 100% duty cycle since there's no low period to refill the pump capacitor.
 
I'm keeping the deadband rather than relying on FET charge/discharge characteristics as I've two 100F super capacitors which will nicely toast a FET if I get it wrong, or those timings change because of heat etc. I've managed to condense the design to use a single PIC32MM which comes with 4 CLCs (aka logic gates). As I need two drivers I've used the attached design to do the deadband which simulates well in LtSpice.
 

Attachments

  • Screenshot_2019-10-30_23-07-36.png
    Screenshot_2019-10-30_23-07-36.png
    17.3 KB · Views: 142
A charge pump based driver won't do 100% duty cycle since there's no low period to refill the pump capacitor.

There is little leakage is mosfet gate capacitance so there is no reason to recharge the charge pump capacitor.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top