mosfet heating up in motor driver circuit

[akshatanayak]

I am using a mosfet IRFZ48 for my motor driver ckt. The specifications of the mosfets and the requirements for the motors match, but the mosfets get heated up. Also, my seniors who originally designed the ckt for their project faced no such problems at that:roll:time. What could possibly be the problem? The input, ckt layout ,components, temperature ,motors everything remains the same then why do the mosfets heat up now?:roll:

 

[Tahmid]

A circuit diagram would help.

If you're using the same circuit, same PCB and same components, there is the possibility of a faulty component. Or you could have placed a component of different value than required, such as resistors, capacitors, diodes, etc.

What is the rating of the motor?
Did you use an anti-parallel diode across the motor?

Is it a 12V motor?
At what voltage are you driving the MOSFETs?

How hot does the MOSFET get? Depending on the current going through, it may be natural for the MOSFET to heat up. Did you use proper heatsinking?

 

[alexan_e]

Usually the heating problems are related to slow switching speed of the mosfets which increases the consumption while the mosfet turns on/off.
This is usually related to the mosfet driver circuit so unless you provide some schematic I don't think that anyone can help.

 

[goldsmith]

Hello my friends
I think the problem of that circuit is that he didn't create dead time at the PWM ? isn't it?
Best Wishes
Goldsmith

 

[varunkant2k]

Heating is natural for MOSFET switches that carries the very high current.
There are 3 kind of power dissipation: 1. Ontime ( I^2*R), 2. Offtime (Loss in freewheeling diode/body diode of the MOSFET), and last (not the least)
3. As said above, switching losses, mostly shoot through ( need to make zero by gate drivers) and parasitic capacitors charging and discharging losses.

If you have given sufficient dead time, then Shoot through may not be problem. Now you need to see the Gate to drain cap charging and discharging. If you do not provide, low impedance path (extra) in the gate, Possibility of turning on of the MOSFET.

 

[akshatanayak]

A circuit diagram would help.

If you're using the same circuit, same PCB and same components, there is the possibility of a faulty component. Or you could have placed a component of different value than required, such as resistors, capacitors, diodes, etc.

What is the rating of the motor?
Did you use an anti-parallel diode across the motor?

Is it a 12V motor?
At what voltage are you driving the MOSFETs?

How hot does the MOSFET get? Depending on the current going through, it may be natural for the MOSFET to heat up. Did you use proper heatsinking?

The components placed are of the same value so that possibility is ruled out. Even since the MOSFET heated up so much we changed the mosfets on the board and checked but the same problem arised. we used proper heat sinking . The excessive current passing through the MOSFET because of which it heats up has to ruled out because the MOSFETS on the board which the seniors had prepared weren't heating up at all. I cannot disclose the circuit diagram but the basic principle is that 89c51 controls the input of L293D and the output of L293D drives the MOSFETS. and is it necessary to use anti-parallel diode across the motor?the MOSFETS have parallel diodes connected to it .

 

[Tahmid]

It's necessary to connect diode across the motor. The MOSFETs have parallel diodes across them, not the motor. But it depends on how the motor is connected. If it is connected in H-bridge configuration, you don't need to connect anti-parallel diode across the motor. But if it is controlled by only one MOSFET (low-side configuration) with the motor as the load in the drain path, then the diode is needed.

Without the circuit diagram, we would just be guessing in the dark and I don't think, we can provide much help.

 

[goldsmith]

Hello my friends
Maybe some auxiliary circuits can help here , but without schematic , the best help , is Impossible.
Best Lucks
Goldsmith

 

[tpetar]

Where is circuit for this? We are not Ray Charles.

 

[FvM]

It's necessary to connect diode across the motor. The MOSFETs have parallel diodes across them, not the motor.

In a H-bridge circuit? How? And why, to short the motor? A single flyback diode would make sense in a one-transistor driver.

You're saying to use L293 as MOSFET driver and IRFZ48 (NMOS) as output transistors. I really wonder in which the circuit. L293 isn't designed as a H-bridge driver, it's a small H-bridge itself. I can imagine to drive a complementary H-bridge (N+PMOS) by the device, but not an all NMOS bridge.

What could possibly be the problem? The input, ckt layout ,components, temperature ,motors everything remains the same then why do the mosfets heat up now?

In other words, your claiming to use a known working circuit, asking us to prove that the assumption is wrong, without giving details. Funny!

 

[Tahmid]

Sorry, that was vague. I clarified this later:

If it is connected in H-bridge configuration, you don't need to connect anti-parallel diode across the motor. But if it is controlled by only one MOSFET (low-side configuration) with the motor as the load in the drain path, then the diode is needed.

Yes, the L293D is a small H-bridge itself. But he never said that he's using it to drive an N-MOSFET in a bridge configuration. So, that's why I said as above. He didn't clarify whether it's a one-MOSFET driver or an H-bridge.

 

[FvM]

But he never said that he's using it to drive an N-MOSFET in a bridge configuration.

What's been said is
- multiple IRFZ48 transistors
- L293 as gate driver
H-bridge output configuration was in fact my guess. But it's not been mentioned. So what could it be otherwise? May be an unipolar stepper motor with center tapped winding. Or multiple single transistor motor drivers.

In the latter case, you are right about the required diode. Without it, the transistor is driven into avalanche breakdown and burning the stored energy, effectively working with linear driver efficiency only.

I clarified this later.

Yes, I apologize for not reading carefully.

 

[goldsmith]

Dear friends
Again Hi
Per my experiences , the heating problem at these circuits are due to some things like : 1 - you didn't create enough dead time at your PWM signal . 2- you didn't use snubber networks. 3- you didn't use the mosfets with low Rdson . 4- if your current is very high , you didn't use the proper heat sinks . 5- it is possible that your mosfets can't switch at that frequency . so what is your PWM frequency? and did you do things that i said ?
Good luck
Goldsmith