Diode or no diode?

This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
Post Reply
Posts: 14
Joined: Wed May 01, 2002 1:01 am

Diode or no diode?

Post by doctorb5 » Mon May 05, 2003 11:03 am

Hello,<p>I am building a DC speed controller for a 24 volt motor. The controller is based on a PIC 16F84 that drives a couple of IRLZ44N FET's. The speed variation is achieved with the PWM method. The frequency of the PWM is about 2000 Hz, and the motor draws about 5 to 10 amps.<p>My question has to do with diode protection. I am getting conflicting information about the need for a protection diode in between the motor leads. With regular bipolar transistors, the diode is always needed. But I've been told that I don't need a diode with a MOSFET because they have one built in. (It is part of their design.) Yet, I have seen several DC speed controllers with MOSFET's *and* diodes!<p>So what am I to believe??? :confused:

Posts: 26
Joined: Thu Apr 17, 2003 1:01 am

Re: Diode or no diode?

Post by bruinbear714 » Mon May 05, 2003 11:07 am

I think you're a little confused as to where to place the diode... It needs to go in parallel with the motor to prevent voltage transients during turn on/off of motor from damaging the rest of the electronics.<p>The majority of the FETS have diodes built in, but that is different from putting a diode in parallel with the motor.

Posts: 14
Joined: Wed May 01, 2002 1:01 am

Re: Diode or no diode?

Post by doctorb5 » Mon May 05, 2003 11:43 am

Thanks Lien.<p>You are correct, the diode is in parallel with the motor and prevents current from reversing during the "cut-off" period of the PWM. But, correct me if I'm wrong, a MOSFET has a diode between Source and Drain that would prevent any of those transient currents from coming back. (I'm not sure I'm making sense, but I think I know what I mean... :D ).

Ron H
Posts: 360
Joined: Tue Dec 18, 2001 1:01 am
Location: Boise, ID

Re: Diode or no diode?

Post by Ron H » Mon May 05, 2003 2:51 pm

The motor is an inductor. When the current is interrupted, a large flyback voltage spike is generated by the collapsing magnetic field. The freewheeling diode across the motor provides a path for the current to decay gradually, which prevents the voltage spike from occurring. The diode in the MOSFET is in the wrong location to provide this function. You need the diode across the motor to prevent damage to your MOSFETs.

User avatar
Posts: 1862
Joined: Fri Feb 01, 2002 1:01 am
Location: Mideast USA

Re: Diode or no diode?

Post by Externet » Mon May 05, 2003 3:30 pm

Hi RonH.<p>There is something that bothers me with the diode shorting the spike at the magnetic field collapse.
We had this subject before in maybe another forum a while ago.<p>The energy lost as heat by the shorting of the spike into windings may be insignificant at the moment of the mosfet turning off, but; if happens ~20,000 times every second on a pulse widht DC motor speed control, this energy waste HAS to be significant.<p>¿ How would you handle such pulses to return the energy to the storage battery instead of shorted to waste ? <p>It would be a similar case for switching power supplies, how would you route that EMF energy back towards the storage (or holding/filter capacitor)?<p>-Unless the shorting diode does not really wastes energy but actually feeds it back in the SAME polarity to the motor softening the mosfet shutdown- I do not have that part clear.<p>It's called BACK emf; but the current is NOT reversed in polarity, is it ? The inductor is pushing current on the SAME direction it had when energized !<p>Thanks,
- Abolish the deciBel ! -

Posts: 1286
Joined: Tue Mar 04, 2003 1:01 am

Re: Diode or no diode?

Post by rshayes » Tue May 06, 2003 1:28 am

Motors are fairly complex beasts. There are several things that happen when the current to a motor is abruptly interrupted. Energy is stored in two forms in a motor. There is a magnetic circuit with an air gap that has a definite inductance and stores energy in electrical form when current is flowing through it. The motor was probably also turning, in which case, there is kinetic energy stored due to the angular momentum of the armature and any inertia associated with its load.<p>The torque in the motor is proportional to current. When the current is interrupted, the torque drops to zero, but the motor keeps turning until the kinetic energy in the motor is either dissipated by bearing and windage losses or transferred to the mechanical load. This may take tenths of seconds to several seconds (or possibly minutes in a gyroscope). While the motor is turning, it generates a voltage of the same polarity as the voltage originally applied to it, but decreasing in value as the motor slows down. This is often referred to as the "back EMF" of the motor, and is a voltage generated by the motor that is proportional to its speed. Some speed controllers use this voltage as a measure of the motors speed.<p>One characteristic of an inductor is that it tends to oppose abrupt changes in current. The inductance in motors tends to be associated with a time constand on the order of tens of microseconds to milliseconds, depending on the size of the motor. When the current is abruptly interrupted, the voltage across the inductance reverses polarity, and jumps to whatever value is necessary for the current to continue flowing. Often, this is enough to breakdown the switching device and destroy it.<p>The diode across the motor is there to give this current a path where the motor current can flow with only about a volt of reverse voltage across the motor. Since the current continues to flow in the same direction, the motor continues to produce torque, and converts the stored electrical energy into mechanical energy and transfers it to the mechanical load of the motor. Usually, this energy is much less than the kinetic energy in the system and has an insignificant effect.<p>In practice, you will want an additional inductor in series with the motor to keep the current flow through the motor continuous. Charging and dicharging the stray capacity of the motor windings will result in high peak currents which stress the switching device. Often an additional capacitance is placed directly across the motor to reduce interference due to commutator noise. Changing the motor current at a rapid rate also causes eddy currents in the iron parts of the motor and increases the heat which the motor has to dissipate. The sharp voltage transitions due to switching should also be filtered or they will tend to cause radiation from the leads between thre controller and the motor.<p>You will probably need a switching device (such as a MOSFET), a diode, an inductor, and a capacitor in the controller. At the motor end, you will probably need an additional shunt capacitor of a lower value, to reduce the noise generated by the motor.

User avatar
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey

Re: Diode or no diode?

Post by MrAl » Tue May 06, 2003 8:44 am

Hello there,<p>To try to quickly address some of the questions that came
up in this thread...<p>[1]
Some MOSFET's have internal reverse diodes, and some dont, and still
some have reverse diodes that are slower then you would want so you
might add higher speed reverse diodes if the forward voltage is lower
then the internal diodes forward voltage.
If your MOSFET's have internal diodes, then they are protected for
reverse voltage to some degree, but not for a high positive spike.
Usually you will use either two transistors in a half bridge or
four transistors in a full bridge (which acheives reverse motor
operation as well as forward).<p>[2]
If you are using a bridge configuration (4 transistors with 4 reverse
diodes) then when the voltage across the motor reverses the 4 diodes
act as a bridge rectifier, routing the current back to the input
storage device, and this raises efficiency.<p>In addition to the reverse diodes:
sometimes snubber circuits are added to absorb the positive spike
that can appear across the transistor collector emitter.
The spike is usually fairly quick, because one of the reverse diodes
eventually takes over the current flow. The signal path is fairly
long during this time though, and the inductance of the leads delays
the forward conduction of the diode. During this time the snubber
absorbs the spike. After the diode begins to conduct, the excess
energy absorbed by the snubber capacitor is dissipated in a
parallel resistor.<p>Also, in some cases, but not all, an inductor is wired in series with
each transistor half bridge. The inductor keeps the current low while
one of the diodes turns off, otherwise a very high current pulse might
flow through the transistors.<p>Good luck with your circuits,
LEDs vs Bulbs, LEDs are winning.

Posts: 14
Joined: Wed May 01, 2002 1:01 am

Re: Diode or no diode?

Post by doctorb5 » Tue May 06, 2003 10:49 am

Whoa! Have you ever had a day when you wanted to ask what you though was a simple question and ended up discovering a whole new world? Well, after reading all this very instructive stuff, I feel just like that... :) <p>I think I understand most of what was said, and I will definitely use the diodes. By the way, is there an easy way to size them? Also, I have seen circuits using Schottky diodes. What's the advantage?<p>Finally, just to add a bit more info, I am not building an H bridge. There is only one switching stage and a simple relay handles the reversing of the motor.

L. Daniel Rosa
Posts: 400
Joined: Tue Dec 25, 2001 1:01 am
Location: Bellingham, WA

Re: Diode or no diode?

Post by L. Daniel Rosa » Tue May 06, 2003 11:30 am

If the motor is ever to be reversed, then don't put one (other options exist) diode directly accross it. At 2000Hz, nearly any rectifier will be fast enough to give you some benefit, and as such a bridge rectifier with the AC terminals on the motor and the + output terminal (yes, the cathodes) to the positive supply and the - terminal to the negative rail may be a viable option. This will do the same as the protection diodes on a full H-bridge.

Bert Russell
Posts: 7
Joined: Tue May 06, 2003 1:01 am
Location: Bellingham, WA.

Re: Diode or no diode?

Post by Bert Russell » Tue May 06, 2003 12:10 pm

Not being an expert on those motor drivers, I would be inclined to include the diodes. If there is already a diode in the MosFet, they would be in parallel, so it probably wouldn't cause a problem, whereas, without any diode, the motor may be in serious trouble.

Post Reply

Who is online

Users browsing this forum: Google [Bot] and 47 guests