Protecting Mosfet from overcurrent and shorts
Lightfoot
Posts: 228
I have a mosfet transistor that I want to protect from direct shorts and current above three amps. The IRF530 is an N channel fet that amplifies a pulse drive from a basic stamp. The circuit is a two-directional fan motor controller. The fan(s) connect to a terminal block which is why it needs to be protected.
Post Edited (Lightfoot) : 12/5/2005 10:27:08 PM GMT
Post Edited (Lightfoot) : 12/5/2005 10:27:08 PM GMT
Comments
I'd suggest using a fast-blow fuse.· A terminal block by nature should "prevent" accidental shorting, unless it's accidental shorting of the intentional variety, because·one shouldn't be able to lay a screwdriver across the terminals.
A fuse itself will not work because a transistor will fry faster then a fast acting fuse.
· There are lots of trade-offs.· The complexity of some current-monitoring scheme vs. replacing the FET in the event it's trashed.· Even crow-bar circuits aren't instantaneous.
· I don't know what current the motors runs at, but a fast-blow fuse may clear before the FET snuffs it.· The datasheet says the 530 can handle 49A for 10usec.· It's rated for 14A continuous.· I don't know the particulars of your project, but your power supply may not be able to supply all the current needed to fry your FET, either.· These are all factors that can work in your favor.
· [noparse][[/noparse] My hi-fi has fuses on its outputs, they must be there to protect the semiconductors in the event of a short-circuit. ]
Post Edited (PJ Allen) : 12/6/2005 1:54:07 AM GMT
Post Edited (Lightfoot) : 12/6/2005 4:05:30 AM GMT
You could 'derate' the FET by using one that is rated at a high amperage. This would make it more tolerant to abuse.
I suggest you read 'The Art of Electronics' for a more complete explanation.
If everything is hardwired and not going to be redesigned, the current limiting input may not be necessary. But, you should try to get the kind of FET that is specially designed for 5 volt switching as they turn on quickly and turn off quickly. They have ones that are rated at 3.3volts too. You use these in a low power, low voltage setup.
The only other real problem with FETs is that they cannot handle extremely high frequency switching because they have more internal capacitance than traditional transistors. I don't think you will have a problem with that.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)
······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan
Not so. One reason fets are used is because they CAN handle high frequency switching. Properly designed drivers will charge and discharge the gate capacitance with no problem. Look at the primary switching element in most modern motor drivers, switching supplies, switching amplifiers, RF power amps, etc.
Rick
· Before Diode, you were switching the FET on & off a lot -- and kicking its butt every time with that resulting inductive discharge.
· To run the motor in the reverse direction, and keep the damping diode, you need to re-wire the switch (see my dwg: dfw_2.jpg in my second post on this subject.)· I'd use a center-off type.
Are you refering to high frequency in comparison to mechanical relays or high frequencies in terms of radio frequencies??
I am interested in what you mean by 'properly designed drivers'.
I am just quoting 'The Art of Electronics'. I don't have much of a library here - everything is in Chinese.
It seems that the technology has moved ahead of that text in FETs.
Nonetheless, most of the switching for motor control is done at a rate that is below the gate capacitance issue.
And, YES - there is at least one way to configure them to avoid it.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)
······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan
{I need to take another look at the book to get into actual resistor values}
It won't hurt to try it.
If it is sucessful, you will never know because it protects from failures.
If it fails, at least you tried to protect the FET.
I manufactured goods these resistors are often absent because the FET is in a circuit that is not changing and needs no such protection.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"When all think alike, no one is thinking very much.' - Walter Lippmann (1889-1974)
······································································ Warm regards,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Truly Understand the Fundamentals and the Path will be so much easier...
Current limiting resistors on the gate, hadnt thought of that till reading this thread. That could serve 2 purposes first limiting available current to protect your micro if your fet shorts to ground. Second and I have no clue what the validity is here but could it do something to keep the stamp pin from sinking to much current off of circuit capacitance when it switches low and pulls the gate down or is that pointless? Personaly I havent used resistors to ground in the last few circuits because when a pin switches low it sinks anyway.
Chris
Note:
Careful of how high the resistance is on the gate resistor that you dont start to create a voltage drop on your gate drive, I wouldnt use anything higher than about 5k. Fets burn up quickly when turned partially on, the faster you make the transition from off to on the better.
Your fet's full power ratings are with a 10 volt gate drive not 5, so you'll have to decipher the data sheet and derate accordingly as your stamp pin is driving 5 volts not 10.
Motor start current can surge much higher than run current, and if reversed while running will surge even higher so may be pulling too much current for the underdrive on the fet. Startup surge can also be too short to get a good reading with an average meter.
Per Kramers suggestion:
International makes a nice line of ttl (5 volt drive) fets, the ttl version of yours is an IRL530 the L specifies logic level gate drive rating. In your application the IRL version should handle much more current since its driven further into saturation.
Post Edited (ChrisP) : 12/9/2005 4:06:36 AM GMT
Even with the 10k in there, the fet brings the full 12 volts through. No burn ups yet.
Post Edited (Lightfoot) : 12/9/2005 4:22:16 AM GMT
"The switch is initially closed and current is flowing through the inductor [noparse][[/noparse]coil.] When the switch is opened, the inductor 'tries' to keep current flowing as it had been. [noparse][[/noparse]This means that the low side of the coil goes positive relative to its high side.] In a case like this it may go 1000V positive before the switch contact 'blows over.' This shortens the life of the switch... [noparse][[/noparse]and if it] happens to be a transistor, it would be an understatement to say its life is shortened; its life is ended!
"When the switch (semiconductor) is on, the diode is back-biased (from the DC drop across the inductor's winding resistance.) At turn-off the diode goes into conduction, putting the switch terminal a diode drop above the supply voltage. The diode must be able to handle the initial diode current, which equals the steady current idle that had been flowing through the inductor [noparse][[/noparse]relay coil]..."
And from Electronics Now, 7/94, "All About Relays", by Ray Marston:
"Because relay coils are inherently inductive, they can generate large back electromotive forces (EMFs) or voltages if their coil-current conductors are suddenly opened.
"These high unwanted voltages can easily damage switch contacts or electronic components connected in the coil circuit.
"To prevent damage from back EMF, it is advisable to damp relay coils with protective diodes... often called a snubber diode."
Now, both these references were describing relay circuits, however a motor coil (windings) situation is just the same.· Hope somebody appreciates this exhaustive research and will find it again (and again) by using the SEARCH button.·
Post Edited (PJ Allen) : 12/10/2005 11:12:29 PM GMT
Chris
· In his article, "All About Relays," Marston continues, "In situations where extra damping would prove beneficial, a pair of protective diodes can be installed."· And, he adds, "A diode pair is recommended in circuits where switch S1 has been replaced by a transistor or other solid-state device."
· Most every circuit I've come across, using a BJT, has only the diode across the coil, few if any with both.· I don't think I've ever seen a situation where there was no diode across the coil, but did have a diode across the "switch" (No.)
· The IGFETs you've described are typically used in switching power supplies.
· * BJT = bipolar junction transistor (NPN, PNP)
"All manufacturers of power MOSFETS seem to connect the body internally to the Source. Because the body forms a diode with the channel, this means that there is an effective diode from Drain to Source; some manufacturers even draw the diode explicitly in their MOSFET symbol so that you won't forget."
· Now, if you look at the datasheet for the MOSFET (IGFET) of Lightfoot's concern, the genesis of this topic, the IRF530, you'll see that the effective diode is depicted as a Zener.·http://www.jameco.com/wcsstore/Jameco/Products/ProdDS/210500.pdf )
· The difference between a Zener and a rectifier diode is that the Zener conducts in both directions (when its·VCC·>·VZ, then there will be Zener current, [noparse][[/noparse] IZ·= (VCC - VZ) / RZ.]· Also, the reverse voltage recovery characteristics of Zeners and rectifier diodes are quite different.
· Given that: a FET is not protected by that internal diode alone.· If it's switching an inductive load or has an inductor in series, then a snubber diode across that inductance is imperative.
Post Edited (Lightfoot) : 12/12/2005 8:28:40 AM GMT
Of course, one wonders why you'd want or need·to reverse the direction of this fan motor -- but, I guess that's another topic in itself and I sure ain't gonna go there.
▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
When you get 1st Place in the "Darwin Awards", you're a Winner & a Loser.
lightfoot said: I just put a 1N5007 3 amp diode from the drain to VDD (reverse biased of course) and the performence increased a lot. Why is this?
If by that you mean the fan runs more smoothly or faster, I think it is because the diode acts as a current path for the motor when the fet is off.
When the fet is on, the current in the motor builds up to some value that depends on the motor inductance, supply voltage and the time the fet is on. The lower the inductance, the higher the current at the end of the on time.
When the fet turns off, the inductance wants to keep that current flowing. This is caused by the collapse of the magnetic field around the windings of the motor. To keep the current flowing in the same direction, the voltage across the inductance has to switch polarity to the opposite from the voltage that charged it. When that happens, the end of the motor attached to the fet will swing positive, putting that voltage plus your battery voltage across the fet. Before, the body diode in the FET was breaking down from the high voltage, and dissapating the power stored in the fan. This would cause the energy stored in the inductance of the fan to be quickly dissapated and the current will rapidly drop to zero. Also will heat up your FET. [noparse]:)[/noparse]
With the diode, that current now has a path to take and the voltage across your fet after it turns off is now the forward voltage of the diode, maybe 1 volt, depending on the current through it, plus the battery voltage.
The current in the fan keeps going for a while, and when the mosfet turns on again, it isn't starting at zero current through the fan.
If you didn't turn the fet on again, the current will drop off to zero, but now more slowly than it built up becaue the voltage across the motor is only the voltage of the diode, not the full battery voltage. Lower voltage across an inductance causes a slower change in the current.
I'm not sure if this is completely accurate for the case of a motor. If it is an average DC motor with commutator etc, there is back EMF caused by the motor acting as a generator. That probably complicates things a bit. Depending on the inductance of the windings and the speed of the fan, all that current ramping and falling might be happening in the time one winding is connected, before the next winding swings around to contact the brushes.
MRC