I am using a surface mount dual MOSFET (P and N Channel) to switch 12vdc (50ma) ON and OFF with a momentary switch. The P-channel MOSFET in this dual package gets destroyed as a result of an unrelated 10amp motor that gets switched ON then OFF.

Why is the MOSFET getting damaged and what can I do to protect it? See attached schematic.

Thanks,

Marty

 

you are exceeding rated specs of components and then you are surprised that they do not work?

http://www.irf.com/product-info/datasheets/data/irf7319pbf.pdf
http://www.datasheetcatalog.org/datasheet2/4/09ddj7ucy7raey5779qx3p9aht3y.pdf

note that Q1 and Q2 are not isolated, they are on a same die. exceeding spec for one may destroy another.

 

Is Q2 paired with Q1 or Q3 in the same package?

 

you are exceeding rated specs of components and then you are surprised that they do not work?

http://www.irf.com/product-info/datasheets/data/irf7319pbf.pdf
http://www.datasheetcatalog.org/datasheet2/4/09ddj7ucy7raey5779qx3p9aht3y.pdf

note that Q1 and Q2 are not isolated, they are on a same die. exceeding spec for one may destroy another.

Hello Panic Mode - Which rated spec do you think I have exceded? The pump motor and its MOSFET work fine all day long. It is the somewhat unrelated soft power circuit that is getting zapped when the motor cycles.

In testing the damaged MOSFETs, it appears that the P-channel (top on schematic) half is the one taking the hit.

 

Is Q2 paired with Q1 or Q3 in the same package?

Q2 and Q3 are in the same package. Q1 is in its own TO-200.

 

You may using a high current diode in series between Battery and Motor, and using a lower current diode in series between Battery and unswitched 12Vdc.

12V Battery → high current diode → Motor(and diode)
12V Battery → lower current diode → unswitched 12Vdc, if the voltage<12V and affective the function of cicuit, then you may cancel the lower current diode.

 

You may using a high current diode in series between Battery and Motor, and using a lower current diode in series between Battery and unswitched 12Vdc.

12V Battery → high current diode → Motor(and diode)
12V Battery → lower current diode → unswitched 12Vdc, if the voltage<12V and affective the function of cicuit, then you may cancel the lower current diode.

Hi Scott,

Not sure I understand. Under normal use of the motor, no diodes are conducting. They are there to protect Q1, the motor's MOSFET.

 

The switched circuit only switches 50mA and should be decoupled from the high power 12v circuit.

I would use a 10 ohm series resistor between the battery and the circuit, and take the power from the + battery terminal, keeping the high power section also powered directly from the + battery terminal.

Then after the series resistor add a large cap to ground, a 470uF 25v should be fine. That cap should be close to all the other parts of the circuit. (ie on the same small PCB).

A photo of your construction will tell a lot.

 

hi Marty,


did you even read the datasheets? they always start with the MAXIMUM or ABSOLUTE MAXIMUM ratings table. first thing I notice is that abs. max current through N type mosfet is 6.5A. you are using it to drive 10A motor which is obviously a mistake. this thing should not be used for loads requiring more than 5A and even then, there are additional constraints (voltage, power, temperature etc.) so you may have to further derate it. absolute maximum is to not be exceeded even in the ideal (best case) scenario (infinite heatsink etc.).

next (an even bigger) discrepancy is diodes. you are using 1A diodes (1N4004) in place where diode rated 10A should be used. when Q1 is on and you get 10A through the motor (ten amp!). but when Q1 turns off, this 10A is like runaway train, it has to go somewhere (it does NOT turn off instantly like through resistor because motor is inductive type of load, capacitors and inductors are energy storage elements). using diode with lower rating is possible but needs careful analysis (duty cycle, temeprature, etc.). if you want to be safe, just use proper size components.

need an example?

you would not go to an Ikea, find some dining table rated for ABSOLUTE MAXIMUM load of 500kg, bring it home, and immediately place 500kg hunk of load on it and then sit at that table with your legs under that same table, would you? you can put a book on it, or a vase, or even have kid or yourself climb on it, but you would never want to place max rated load on it (or close to it).

why do you think that 1A diode or 6.5A mosfet can or should be used in a 10A circuit? moreover, what is the motivation to use an undersized component? savings? maaybe 40 years ago, but semiconductors are so common, cheap and compact, it costs nothing to supersize them and have some elbow room too (and you will have piece of mind too).

i'd probably use BUZ11 to run that motor. for bigger current i have STH260N6F6-2 (getting the idea now?).

 

In testing the damaged MOSFETs, it appears that the P-channel (top on schematic) half is the one taking the hit.

yes that is the one with lower rating but you can look at Q1 and Q2 as two separate things because:
a) the ratings are exceeded (by a lot!)
b) they are in the same package and on same piece of silicon. perhaps the only reason Q1 survives is because Q2 is constantly taking the hit.

since you managed to blow the Q2 more than once, why don't you use separate IC as Q2 and see what happens?

 

Hi Scott,

Not sure I understand. Under normal use of the motor, no diodes are conducting. They are there to protect Q1, the motor's MOSFET.

The parts that I added as shown below.

Just don't worry about the Back EMF(Electromotive force), because it didn't affect the Motor of MOSFET, you can see the arrow that I drew, it will release the reverse current through the diode as arrow.

 

can you show a photo of actual circuit?

 

The switched circuit only switches 50mA and should be decoupled from the high power 12v circuit.

I would use a 10 ohm series resistor between the battery and the circuit, and take the power from the + battery terminal, keeping the high power section also powered directly from the + battery terminal.

Then after the series resistor add a large cap to ground, a 470uF 25v should be fine. That cap should be close to all the other parts of the circuit. (ie on the same small PCB).

A photo of your construction will tell a lot.

Hello RB,

Good ideas! I didn't think about the need for a decoupling capacitor. I will try it with the cap but may not use the resistor yet as sometimes it draws up to 2 amps thru the low power side and don't want the voltage to drop.

Marty

 

hi Marty,


did you even read the datasheets? they always start with the MAXIMUM or ABSOLUTE MAXIMUM ratings table. first thing I notice is that abs. max current through N type mosfet is 6.5A. you are using it to drive 10A motor which is obviously a mistake. this thing should not be used for loads requiring more than 5A and even then, there are additional constraints (voltage, power, temperature etc.) so you may have to further derate it. absolute maximum is to not be exceeded even in the ideal (best case) scenario (infinite heatsink etc.).

next (an even bigger) discrepancy is diodes. you are using 1A diodes (1N4004) in place where diode rated 10A should be used. when Q1 is on and you get 10A through the motor (ten amp!). but when Q1 turns off, this 10A is like runaway train, it has to go somewhere (it does NOT turn off instantly like through resistor because motor is inductive type of load, capacitors and inductors are energy storage elements). using diode with lower rating is possible but needs careful analysis (duty cycle, temeprature, etc.). if you want to be safe, just use proper size components.

need an example?

you would not go to an Ikea, find some dining table rated for ABSOLUTE MAXIMUM load of 500kg, bring it home, and immediately place 500kg hunk of load on it and then sit at that table with your legs under that same table, would you? you can put a book on it, or a vase, or even have kid or yourself climb on it, but you would never want to place max rated load on it (or close to it).

why do you think that 1A diode or 6.5A mosfet can or should be used in a 10A circuit? moreover, what is the motivation to use an undersized component? savings? maaybe 40 years ago, but semiconductors are so common, cheap and compact, it costs nothing to supersize them and have some elbow room too (and you will have piece of mind too).

i'd probably use BUZ11 to run that motor. for bigger current i have STH260N6F6-2 (getting the idea now?).

Hello Panic Mode,

Q1, the motor driver, is rated for 75 amps (Id), 55 Vdss. It seems to work fine for the 10 amp motor.

Q2 and Q3 are packaged together the P-channel side is limited to 4 amps, I'm drawing no more than 2 amps (only 50ma during this test that causes failures).

You're right, the 1N4004 diodes I'm using are only rated for 1 amp. They don't seem to be failing though. Can you suggest an inexpensive diode rated for over 10 amps?

I like the idea of inserting a decoupling cap at the front end of the unswitched 12v.

What do you think about installing a Transient-Voltage_Supression diode across the same place?

Thanks,

Marty

 

The parts that I added as shown below.

Just don't worry about the Back EMF(Electromotive force), because it didn't affect the Motor of MOSFET, you can see the arrow that I drew, it will release the reverse current through the diode as arrow.

Hi Scott,

I understand now. Thanks for modifying the drawing. I like your ideas. My concerns are the voltage drop caused by each added diode will effect the motor voltage and the unswitched 12v. Each side needs to be as high a voltage as available (one side goes to the pump motor and the other side goes to an audio amplifier).

Will it still work without the 1N4004 you inserted?

Thanks,

Marty

 

Q1 is http://www.nxp.com/documents/data_sheet/BUK9508-55B.pdf

Lower Rds (7 m ohm) campared to BUZ11 (40 m ohm)

 

Hi Scott,

I understand now. Thanks for modifying the drawing. I like your ideas. My concerns are the voltage drop caused by each added diode will effect the motor voltage and the unswitched 12v. Each side needs to be as high a voltage as available (one side goes to the pump motor and the other side goes to an audio amplifier).

Will it still work without the 1N4004 you inserted?

Thanks,

Marty

The DC Motor can be working for a big range, you can check yours.
The unswitched 12v just need 50mA, so in series 1N4004 as a limit resistor, you can also using a 10 resistor to replace the 1N4004, Vdrop = 50mA * 10Ω = 0.5V

 

The DC Motor can be working for a big range, you can check yours.
The unswitched 12v just need 50mA, so in series 1N4004 as a limit resistor, you can also using a 10 resistor to replace the 1N4004, Vdrop = 50mA * 10Ω = 0.5V

Thanks Scott,

What are we trying to limit on the 50ma, low power, side? Current or Voltage? Have you ever used a Transient-Voltage-Supression diode? I've heard of others using them to protect MOSFETs.

Marty

 

Thanks Scott,

What are we trying to limit on the 50ma, low power, side? Current or Voltage? Have you ever used a Transient-Voltage-Supression diode? I've heard of others using them to protect MOSFETs.

Marty

I'm using Schottky diodes and rectifier diodes to protect the Motor, relay or some other Inductive products, I'm not using TVS to protect MSOFET, maybe someday I will do that.

About three weeks ago, one of my friend asked me to help him to debug for a interference of transformerless circuit, I saw that he using TVS to protect the DC circuit for surge voltage.

Using TVS to protect MOSFET.

 

Thanks Scott,

Nice article on protecting MOSFETs with TVS diodes.

Marty