How to build a H-Bridge Driver with 4 N-MOSFET?

Thread Starter

chrislee84

Joined Dec 8, 2007
14
HI,

I need to build a 2 H-bridge driver with 4 N-MOSFET to drive a PMDC motor.
The motor have rated voltage 24V and rated current 8.5A.
So, I choose the IRFP 260 N-MOSFET because it can sustain the current untill 45A in order to protect the MOSFET from over current.
But the problems I face now are
1) I don't know how to control or turn on the High side of the H-bridge. The
battery I use is 24V.
2) What is the protection circuit to protect the H-bridge, the motor and the
the controller which produce the PWM to control the motor speed?
As i know, we can use an optocoupler, but how to use it?
3) What others protection do I have to consider in order to protect my
controller and the motor?

Can someone post a schematic circuit of the H-Bridge driver which is suitable for my project?
Thanks.
 

Seidleroni

Joined Dec 8, 2007
18
Hi Chris,

I actually ran into a very similar problem recently. The short answer is that I had to use 2 N channel and 2 P channel MOSFET's. Then you can control the circuit easily with the proper voltages. One issue I had is "shoot-through" which means that if both signals are on, you short circuit the power supply which is very bad. What I did to get over this was to have the two MOSFETs on the left side of the H-Bridge controlled by one signal, and the right side by the other signal. IF you do the truth tables for this, you will see that there is no condition where "shoot through" exists. I do not have a schematic, but it should be fairly simple.
 

nanovate

Joined May 7, 2007
666
I choose the IRFP 260 N-MOSFET because it can sustain the current untill 45A
Wow how big is your heatsink? You are dissipating around 80W per FET.


1) I don't know how to control or turn on the High side of the H-bridge. The battery I use is 24V.
Take a look at the HIP4081 from Intersil. Their FAQs have recommendations on protection circuits also. I am assuming you are interfacing from a logic level device like a microcontroller. You can also look for MOSFET driver circuits which bootstrap the voltage to the high side gates.

As Seidleroni mentioned, be careful about how you sequence and handle the turn on/off times. PCH FETs are not as fast as NCH ones (and they have higher RDS_on).
 

Thread Starter

chrislee84

Joined Dec 8, 2007
14
Actually I had test the IRFP260N, but I face a problem which make me very curious about it, that is, when I supply a voltage at drain pin, and the source pin connected to a LED, the LED is light without I apply voltage to the gate pin. Why this happen?
Why the drain and source pin seen short circuited?
 

Thread Starter

chrislee84

Joined Dec 8, 2007
14
Hi SgtWookie, why the MOSFET is ON by default?
To ON a N-MOSFET, we shall apply a voltage MORE positive than Vs at Vg rite?
Because the N-MOSFET will be on at Vgs =10V as i look at the datasheet.
 

nanovate

Joined May 7, 2007
666
Did you connect the MOSFET backwards? There is a body diode. Also if the MOSFET is bad then it will look like a short.


You should have voltage connected through a resistor to the anode of the LED and the cathode of the LED connected to the Drain of the FET and the Source grounded. The Gate should be pulled down to ground via a resistor.

JFETs and "depletion-mode" MOSFETs are ON by default. The IRFP260N is an "enhancement-mode" type.
 

Thread Starter

chrislee84

Joined Dec 8, 2007
14
Hi nanovate,

"You should have voltage connected through a resistor to the anode of the LED and the cathode of the LED connected to the Drain of the FET and the Source grounded."

Yes, I do what you mention before, but when I turn on the MOSFET, it will stay ON although I remove the apply gate voltage. I apply a 8V to the gate.

"The Gate should be pulled down to ground via a resistor."
What is this sentence mean? Why the gate have to be pulled down to ground? If the gate is grounded, then the MOSFET is not ON.
 

hgmjr

Joined Jan 28, 2005
9,029
In order to answer your question about the MOSFET that will not turn off once it is turned off, I think we are going to need a sketch of the circuit hook up you are using when you encounter the problem you describe.

hgmjr
 

nanovate

Joined May 7, 2007
666
What is this sentence mean? Why the gate have to be pulled down to ground? If the gate is grounded, then the MOSFET is not ON.
See attached jpeg



When you take the 8V away is the gate floating?

I also am attaching an article that you (and others) might find helpful. It has some good info on designing an H-Bridge circuit.
 

Attachments

thingmaker3

Joined May 16, 2005
5,084
Yes, I do what you mention before, but when I turn on the MOSFET, it will stay ON although I remove the apply gate voltage. I apply a 8V to the gate.

"The Gate should be pulled down to ground via a resistor."
What is this sentence mean? Why the gate have to be pulled down to ground? If the gate is grounded, then the MOSFET is not ON.
The gate should be pulled down to ground via a resistor. When you apply the 8v, it will be across the resistor to turn the MOSFET on. When you remove the 8v, the MOSFET gate will discharge through the resistor and turn off.
 

ashokcp

Joined Mar 8, 2007
50
MOSFET input impedance is very high, in the order of Giga Ohms! But the gate is (gate source) is a cap which get charged when you apply a voltage, and, untill this cap (of the order of few thousand pF depending on the MOSFET) is discharged, the MOSFET remains On! MOSFET input impedance is very high, in the order of Giga Ohms, and, does not discharge this voltage soon enough! A 10k between the gate and source discharges this cap (let us say 1000 pF) with a time constant of about 10 uS.
 

Thread Starter

chrislee84

Joined Dec 8, 2007
14
Hi everybody, who can help me solve my problem? Recently very headache!!!

I have try testing with a simple H-bridge that i build. But after the problem i face now is how i turn on the high side easily?

I try with the half bridge with the motor connected in between of the two mosfet, and turn on it by a pulled down resistor 10k.

It can work but the voltage to the motor is just 3.5V instead of 12 volt(Battery supply) when is turn ON and the high side mosfet is heat up very fast. The motor i use to test is the 12V power window motor.Why this problem happen? How i have to overcome this problem.
 

Thread Starter

chrislee84

Joined Dec 8, 2007
14
Hi nanovate,
The high side MOSFET i use is the N-Channel.
If u mention that i have to luse the MOSFET Driver, which one or how i choose the most suitable one?
And, did it had others solution instead using the MOSFET Driver?
Thanks.
 

nanovate

Joined May 7, 2007
666
You shouldn't use an N-CH MOSFET on the high side unless you can supply the proper gate voltage. You need to pull the gate to 10V above the source. If you are supplying 12V then you need approx 12V + Vgs --> 22V.

If you look back to Seidleroni's post you'll see he recommended using a PCH MOSFET for the high-side. For a PCH you need to pull the gate below the source by 10V (assuming Vsg >=10V for saturation). So just pull the gate to GND to turn it on.

Intersil makes the HIP4080/HIP4081 which is a full function H-Bridge controller. I posted an article which if you read should show you one way to do it.

You can also look at the MAX620 or MAX621 for HS drivers.
 

bbroach

Joined Dec 12, 2007
3
Hi everybody, who can help me solve my problem? Recently very headache!!!

I have try testing with a simple H-bridge that i build. But after the problem i face now is how i turn on the high side easily?

I try with the half bridge with the motor connected in between of the two mosfet, and turn on it by a pulled down resistor 10k.

It can work but the voltage to the motor is just 3.5V instead of 12 volt(Battery supply) when is turn ON and the high side mosfet is heat up very fast. The motor i use to test is the 12V power window motor.Why this problem happen? How i have to overcome this problem.
This IC:

http://www.national.com/ds.cgi/LM/LMD18200.pdf

will simplify your motor driver application considerably. The Mil/Aero version is used on the Mars Rovers for motor control and, in the commercial package (11 pin TO220), it's a very popular hobbyist solution. The DS includes several applications circuits, including torque & speed based control loops, which were extensively bench tested prior to release.

regards, billb....
 

thingmaker3

Joined May 16, 2005
5,084
It is indeed possible to use N-channel MOSFETs for both low and high-side. Its a little "clunky," though.

nchhbr.jpg

I got one of these to run at more than half a megaHertz before I blew up my MOSFETs.
 
Top