Mosfet gate voltage clamp... need advice!

Thread Starter

johndavid400

Joined Jun 28, 2009
5
I built an H-bridge using P-channel mosfets for the high-side switches. The mosfet's maximum gate-to-source voltage is +/-25vdc.... My design uses pull-up resistors to pull the gate voltage up to the source voltage to turn them off. This works fine at 12v and 18v.... But if I use a +24v power supply, this will create a gate-to-source voltage that is very close the max Vgs for the mosfet.

My question is how to limit the voltage difference between the gate and source to keep it well under 25v even though each gate is pulled up to source. Since the mosfet only needs 10v to turn on, will a 15v zener diode work between the gate and source to keep the Vgs within range? If so, how would I go about connecting it?

thanks,
jd
 

MikeML

Joined Oct 2, 2009
5,444
Since you are turning off the PFet by letting a resistor (the pullup from gate to source) discharge the gate capacitance, you must not be worried about switching speed or dissipation in the PFet. Maybe, you should be...

Slow switching speed causes the PFET to take a long time turning off, during which time it is neither on or off, which means it dissipates power during the transition. Turning it off quickly gets it through the transition so that the dissipation is reduced.

What is the driver that is pulling the gate low (turning on the PFet)? Is it an NPN transistor with the emitter grounded? If so, you could connect a resistor from the collector to the gate of the PFET, roughly equal to the pull-up that is already there. This will create a voltage divider which will limit the gate voltage swing to about half the power supply range. However, this slows the turn on of the PFet, while keeping the turn off unchanged, making the dissipation hit twice as bad, because now you have both a slow turn on as well as turn off.
 

SgtWookie

Joined Jul 17, 2007
22,230
Try this:



R1 limits Q1's base current
R2 limits max sink current to around 50mA.
D1, a 12v Zener, keeps the gate from getting too far away from the source.
R3 keeps Q7 turned on when Q1 is off.
D2 provides a discharge path for Q2's gate, but keeps it from turning on Q7 when charged to Vcc.
R4 keeps Q2 turned off if Q7 or D2 fail.
 

Attachments

Thread Starter

johndavid400

Joined Jun 28, 2009
5
I am using the n-channel mosfet equivalent of a 3904 bjt, the 2n7000, to turn on the high side fets. I am currently using an Arduino microcontroller with the pwm frequency at 1kHz. It works great at lower voltages, but I need to make sure I don't fry anything if I use 24v.

I'm gonna try Sgt. Wookie's diagram on a breadboard...

Here is the schematic I came up with for my current setup, each mosfet has a 47ohm gate resistor and a 4.7k pull-up/down resistor, mosfets are FQP47P06 and FQP50N06L (RdsOn=.026ohm, 47amp), sorry I used LED's for the motor. Any other ideas are greatly appreciated:
 
I have designed a small mosfet amplifier's gate clamp that clamps the gate source area to 6 Volts, should the gate exceed the maximum of 7 amps. Relying on the specs of the 2sk1058 of mutual conductance of 1 amp per volt. Using a ZVN2110A Nchan minifet is ideal because the zener can be connected to the drain, buffering the zener's nonlinearities, whilst the input 100k and 150k minifet bias resistors prevent loading, hence lowering the distortion out of the picture. The whole circuit is then connected to the source through 1N4148 signal diode to prevent reverse conduction. The circuit is fed from the previous stage with a 240 Ohm resistor. After the clamp connect to the 2sk1058 gate with 62 Ohm R's

Of course, under 6 Volts the circuit has no effect at all.

AND the same circuit may be used with the 2sj162 fets with the same results. Just remember that the voltage are the other way round so the board has reverse connections. Mutual Conduction of this fet is also 1 Amp per Volt.
 
Top