EmbeddedRelated.com
Forums

MOSFET keeps burning out

Started by nathanchronister June 28, 2004
Hi. I'm using the BX24 to control a DC motor. I have the PWM output
of the BX24 hooked up to an N-channel MOSFET. (I'm not using an H-
bridge because I don't need reverse.) Even though the MOSFET is rated
25 amps and should easily handle the motor current it keeps burning
out. After this happens the motor won't shut off.

I have the BX24 wired *directly* to the Gate pin of the MOSFET. The
Gate is also wired to the ground via a 1000 ohm resistor. This is
intended to prevent static electricity from turning the MOSFET on.

The Source pin of the MOSFET is also wired to the ground.

The Drain pin is wired to the motor, and the other motor lead goes to
the + side of the 7.4 volt battery which drives the motor.

The BX24 is powered by the developers board. The motor control
circuit is grounded on the ground strip on the developers board and
the - battery lead goes to the ground strip also.

When this is all set up, it works fine for a while, and then the
motor will switch full on and I have to unplug the battery to make it
stop. (The PWM is correct however and the MOSFET failed to work
correctly when tested separate from the BX24.) Does it sound like I'm
doing something wrong which could cause the MOSFET to fail?

Thanks for your help!
Nathan



too much gate current?
.
Put a 100 ohm or so resistor in series from the bx pin
to the gate terminal. Odd are the output of the bx was
not swinging to its full hi/lo level. Mosfets
consume the most power during turn on/off transition.
Full on... the resistance should be low enough that
you shouldn't generate much heat in the mosfet.
But if it only partially turns on/off it'll cook easily.

Paul Tegler

www.teglerizer.com
----- Original Message -----
From: "nathanchronister" <>
To: <>
Sent: Monday, June 28, 2004 8:54 PM
Subject: [BasicX] MOSFET keeps burning out > Hi. I'm using the BX24 to control a DC motor. I have the PWM output
> of the BX24 hooked up to an N-channel MOSFET. (I'm not using an H-
> bridge because I don't need reverse.) Even though the MOSFET is rated
> 25 amps and should easily handle the motor current it keeps burning
> out. After this happens the motor won't shut off.
>
> I have the BX24 wired *directly* to the Gate pin of the MOSFET. The
> Gate is also wired to the ground via a 1000 ohm resistor. This is
> intended to prevent static electricity from turning the MOSFET on.
>
> The Source pin of the MOSFET is also wired to the ground.
>
> The Drain pin is wired to the motor, and the other motor lead goes to
> the + side of the 7.4 volt battery which drives the motor.
>
> The BX24 is powered by the developers board. The motor control
> circuit is grounded on the ground strip on the developers board and
> the - battery lead goes to the ground strip also.
>
> When this is all set up, it works fine for a while, and then the
> motor will switch full on and I have to unplug the battery to make it
> stop. (The PWM is correct however and the MOSFET failed to work
> correctly when tested separate from the BX24.) Does it sound like I'm
> doing something wrong which could cause the MOSFET to fail?
>
> Thanks for your help!
> Nathan > Yahoo! Groups Links






Problem 1: N_Channel MOSFETs are designed to sink current not source
as you are wiring it. Connecting the FET
in this manor will prevent the FET from ever turning "Full on" and
quickly cause it to heat up and burn out. The normal
way to connect an N-Channel fet is to connect the + of your battery
to the motor +, the motor - to the FET Drain and the
battery - to FET source.

Problem 2: A free spinning motor acts as a generator and back feeds
reverse voltage which depending on the size of your motor
can and will burn out a FET very quickly. To solve this you should
add a diode (preferably a shotkey diode) across the motors inputs.
Anode to - and Cathode (stripped side) to +. Most FETs have a built-
in diode but they may not be large enough for the motor you are
using.

You may be able to do without the diode, but it wouldn't hurt to
have it! Chris
--- In , "nathanchronister"
<nathanchronister@y...> wrote:
> Hi. I'm using the BX24 to control a DC motor. I have the PWM
output
> of the BX24 hooked up to an N-channel MOSFET. (I'm not using an H-
> bridge because I don't need reverse.) Even though the MOSFET is
rated
> 25 amps and should easily handle the motor current it keeps
burning
> out. After this happens the motor won't shut off.
>
> I have the BX24 wired *directly* to the Gate pin of the MOSFET.
The
> Gate is also wired to the ground via a 1000 ohm resistor. This is
> intended to prevent static electricity from turning the MOSFET on.
>
> The Source pin of the MOSFET is also wired to the ground.
>
> The Drain pin is wired to the motor, and the other motor lead goes
to
> the + side of the 7.4 volt battery which drives the motor.
>
> The BX24 is powered by the developers board. The motor control
> circuit is grounded on the ground strip on the developers board
and
> the - battery lead goes to the ground strip also.
>
> When this is all set up, it works fine for a while, and then the
> motor will switch full on and I have to unplug the battery to make
it
> stop. (The PWM is correct however and the MOSFET failed to work
> correctly when tested separate from the BX24.) Does it sound like
I'm
> doing something wrong which could cause the MOSFET to fail?
>
> Thanks for your help!
> Nathan




1. Put a 1-10k resistor is series with the gate ( to protect the bx24.
*Always* avoid connecting any BX24 I/O pin "directly "to anything).
2. Put a diode such as a 1n4003 across the motor with the bar to the
battery +ve
3. Check that the voltage rating of the MOSFET is at least 3x the
battery voltage
4. When the motor is running at full speed ( ie DC +5 on the gate
resistor, not pwm) check the voltage between D and S. It should be lower
than 0.5v DC
5. Does the mosfet get hot when it is running? It shouldn't. If it
does, its not turning on hard enough due to gate voltage too low or
gate rise time too slow.

neil wrote:

> too much gate current?
> .
> Put a 100 ohm or so resistor in series from the bx pin
> to the gate terminal. Odd are the output of the bx was
> not swinging to its full hi/lo level. Mosfets
> consume the most power during turn on/off transition.
> Full on... the resistance should be low enough that
> you shouldn't generate much heat in the mosfet.
> But if it only partially turns on/off it'll cook easily.
>
> Paul Tegler
>
> www.teglerizer.com
> ----- Original Message -----
> From: "nathanchronister" <>
> To: <>
> Sent: Monday, June 28, 2004 8:54 PM
> Subject: [BasicX] MOSFET keeps burning out > > Hi. I'm using the BX24 to control a DC motor. I have the PWM output
> > of the BX24 hooked up to an N-channel MOSFET. (I'm not using an H-
> > bridge because I don't need reverse.) Even though the MOSFET is rated
> > 25 amps and should easily handle the motor current it keeps burning
> > out. After this happens the motor won't shut off.
> >
> > I have the BX24 wired *directly* to the Gate pin of the MOSFET. The
> > Gate is also wired to the ground via a 1000 ohm resistor. This is
> > intended to prevent static electricity from turning the MOSFET on.
> >
> > The Source pin of the MOSFET is also wired to the ground.
> >
> > The Drain pin is wired to the motor, and the other motor lead goes to
> > the + side of the 7.4 volt battery which drives the motor.
> >
> > The BX24 is powered by the developers board. The motor control
> > circuit is grounded on the ground strip on the developers board and
> > the - battery lead goes to the ground strip also.
> >
> > When this is all set up, it works fine for a while, and then the
> > motor will switch full on and I have to unplug the battery to make it
> > stop. (The PWM is correct however and the MOSFET failed to work
> > correctly when tested separate from the BX24.) Does it sound like I'm
> > doing something wrong which could cause the MOSFET to fail?
> >
> > Thanks for your help!
> > Nathan
> >
> >
> >
> >
> >
> >
> >
> >
> > >.




another point... the MOSFET should be a logic level version that turns on at a lower voltage than the normal rated MOSFETS. Paul and Neil covered the other critical items.

best regards, Steve Thatcher

-----Original Message-----
From: Neil Jepsen <>
Sent: Jun 29, 2004 1:56 AM
To:
Subject: Re: [BasicX] MOSFET keeps burning out

1. Put a 1-10k resistor is series with the gate ( to protect the bx24.
*Always* avoid connecting any BX24 I/O pin "directly "to anything).
2. Put a diode such as a 1n4003 across the motor with the bar to the
battery +ve
3. Check that the voltage rating of the MOSFET is at least 3x the
battery voltage
4. When the motor is running at full speed ( ie DC +5 on the gate
resistor, not pwm) check the voltage between D and S. It should be lower
than 0.5v DC
5. Does the mosfet get hot when it is running? It shouldn't. If it
does, its not turning on hard enough due to gate voltage too low or
gate rise time too slow.

neil wrote:

> too much gate current?
> .
> Put a 100 ohm or so resistor in series from the bx pin
> to the gate terminal. Odd are the output of the bx was
> not swinging to its full hi/lo level. Mosfets
> consume the most power during turn on/off transition.
> Full on... the resistance should be low enough that
> you shouldn't generate much heat in the mosfet.
> But if it only partially turns on/off it'll cook easily.
>





Is the Mosfet on a heatsink?
They are not designed to run without it with PWM.
If I understand Mosfet's even if they are rated
at a much greater current than you require they still
need to dissipate a great deal of heat when in transition.

jmtcw,
Don Lewis

--- In , "nathanchronister"
<nathanchronister@y...> wrote:
> Hi. I'm using the BX24 to control a DC motor. I have the PWM output
> of the BX24 hooked up to an N-channel MOSFET. (I'm not using an H-
> bridge because I don't need reverse.) Even though the MOSFET is rated
> 25 amps and should easily handle the motor current it keeps burning
> out. After this happens the motor won't shut off.
>
> I have the BX24 wired *directly* to the Gate pin of the MOSFET. The
> Gate is also wired to the ground via a 1000 ohm resistor. This is
> intended to prevent static electricity from turning the MOSFET on.
>
> The Source pin of the MOSFET is also wired to the ground.
>
> The Drain pin is wired to the motor, and the other motor lead goes to
> the + side of the 7.4 volt battery which drives the motor.
>
> The BX24 is powered by the developers board. The motor control
> circuit is grounded on the ground strip on the developers board and
> the - battery lead goes to the ground strip also.
>
> When this is all set up, it works fine for a while, and then the
> motor will switch full on and I have to unplug the battery to make it
> stop. (The PWM is correct however and the MOSFET failed to work
> correctly when tested separate from the BX24.) Does it sound like I'm
> doing something wrong which could cause the MOSFET to fail?
>
> Thanks for your help!
> Nathan





> ... Is the Mosfet on a heatsink?

According to my experience, if the switch failed due to overdissipation
the cause is obvious. The component gets _very_ hot, far too hot to not
notice; in my case, the un-sunk package literally smoked and nearly
unsoldered itself. I suppose transitional overdissipation could make an
instantaneous fuse and allow the package to stay cool, but my failure
was not fast - it was HOT. Tom
Tom Becker
--... ...--
www.RighTime.com
The RighTime Clock Company, Inc., Cape Coral, Florida USA
+1239 540 5700


Thank you all for the very helpful suggestions. I think the MOSFET
may not be turning on all the way, as it *was* getting hot. I didn't
know it would burn out if it wasn't on all the way but that makes
sense to me now.

Paul, were you suggesting that a 100 ohm resistor between the BX and
the Gate would help it turn on all the way? Could you explain how
that works?

The MOSFET is Fairchild Semi's #NDP603AL. I think it has all the
right specs but I'm new to this so I could be wrong. Here is the
documentation page: http://www.fairchildsemi.com/ds/ND/NDP603AL.pdf

The internal diode is rated 25 amps but only 1.5 volts. Could that be
the problem?

Nathan



As someone else mentioned as well.... if the gate draws too
much current from the bx pin, the voltage transition of the bx
pin may not be swinging to its full hi/lo states.

A mosfet should only require a milliamp or two at the max
to turn on fully. Without a limiting resistor, the gate 'could' pull
a lot more current based on your ground references and etc.

Paul Tegler

www.teglerizer.com
----- Original Message -----
From: "nathanchronister" <>
To: <>
Sent: Tuesday, June 29, 2004 1:52 PM
Subject: [BasicX] Re: MOSFET keeps burning out > Thank you all for the very helpful suggestions. I think the MOSFET
> may not be turning on all the way, as it *was* getting hot. I didn't
> know it would burn out if it wasn't on all the way but that makes
> sense to me now.
>
> Paul, were you suggesting that a 100 ohm resistor between the BX and
> the Gate would help it turn on all the way? Could you explain how
> that works?
>
> The MOSFET is Fairchild Semi's #NDP603AL. I think it has all the
> right specs but I'm new to this so I could be wrong. Here is the
> documentation page: http://www.fairchildsemi.com/ds/ND/NDP603AL.pdf
>
> The internal diode is rated 25 amps but only 1.5 volts. Could that be
> the problem?
>
> Nathan >
> Yahoo! Groups Links



Nathan
I've checked the specs of the fet and it is OK for the
job. It is also connected to "sink" correctly as you have it connected
now.( in spite of another post to the contrary). Your problem will be
either no heatsink ( too hot) and /or no back emf diode. The missing
gate resistor has nothing to do with it as the gate current is
micro-amps. Howver put one in to protect the bx24
neil Steve Thatcher wrote:

> another point... the MOSFET should be a logic level version that turns
> on at a lower voltage than the normal rated MOSFETS. Paul and Neil
> covered the other critical items.
>
> best regards, Steve Thatcher
>
> -----Original Message-----
> From: Neil Jepsen <>
> Sent: Jun 29, 2004 1:56 AM
> To:
> Subject: Re: [BasicX] MOSFET keeps burning out
>
> 1. Put a 1-10k resistor is series with the gate ( to protect the bx24.
> *Always* avoid connecting any BX24 I/O pin "directly "to anything).
> 2. Put a diode such as a 1n4003 across the motor with the bar to the
> battery +ve
> 3. Check that the voltage rating of the MOSFET is at least 3x the
> battery voltage
> 4. When the motor is running at full speed ( ie DC +5 on the gate
> resistor, not pwm) check the voltage between D and S. It should be lower
> than 0.5v DC
> 5. Does the mosfet get hot when it is running? It shouldn't. If it
> does, its not turning on hard enough due to gate voltage too low or
> gate rise time too slow.
>
> neil > wrote:
>
> > too much gate current?
> > .
> > Put a 100 ohm or so resistor in series from the bx pin
> > to the gate terminal. Odd are the output of the bx was
> > not swinging to its full hi/lo level. Mosfets
> > consume the most power during turn on/off transition.
> > Full on... the resistance should be low enough that
> > you shouldn't generate much heat in the mosfet.
> > But if it only partially turns on/off it'll cook easily.
> > *
> <http://companion.yahoo.com" target="_blank" rel="nofollow">http://us.ard.yahoo.com/SIG9uppnug/M)5196.4901138.6071305.3001176/D=groups/S06554205:HM/EXP88590928/A!28215/R=0/SIGse96mf6/*http://companion.yahoo.com >
>
> *>.