MOSFET: N- or P-, enhance or deplete?

Bob Engelhardt · Jun 16, 2017

Here's really basic question for s.e.BASICS:
I want to use a MOSFET to control a load, via a switch. The tricky part
is that I need load current when the switch is open and no current when
the switch is closed. What I am confused about is what type of MOSFET
to use, and its orientation in the circuit (S & D connected which way?)

+
|
L
|
|
/ |-
+ --/ o--+--|
| |-
R |
| |
G G

I could put a BJT inverter before the MOSFET, but is there a way that I
don't have to?

Thanks,
Bob

rickman · Jun 16, 2017

Bob Engelhardt wrote on 6/16/2017 12:37 PM:

Here's really basic question for s.e.BASICS:
I want to use a MOSFET to control a load, via a switch. The tricky part is
that I need load current when the switch is open and no current when the
switch is closed. What I am confused about is what type of MOSFET to use,
and its orientation in the circuit (S & D connected which way?)

+
|
L
|
|
/ |-
+ --/ o--+--|
| |-
R |
| |
G G

I could put a BJT inverter before the MOSFET, but is there a way that I
don't have to?

Can the switch be wired to ground instead of V+? Tie the resistor to V+ and
an open switch lets current flow in the load and a closed switch stops
current from flowing in the load.

--

Rick C

Phil Hobbs · Jun 16, 2017

On 06/16/2017 12:37 PM, Bob Engelhardt wrote:

Here's really basic question for s.e.BASICS:
I want to use a MOSFET to control a load, via a switch. The tricky part
is that I need load current when the switch is open and no current when
the switch is closed. What I am confused about is what type of MOSFET
to use, and its orientation in the circuit (S & D connected which way?)

+
|
L
|
|
/ |-
+ --/ o--+--|
| |-
R |
| |
G G

I could put a BJT inverter before the MOSFET, but is there a way that I
don't have to?

The simple approach is to put the resistor from the gate to the positive
supply, and the switch to ground.

Cheers

Phil Hobbs

--
Dr Philip C D Hobbs
Principal Consultant
ElectroOptical Innovations LLC
Optics, Electro-optics, Photonics, Analog Electronics

160 North State Road #203
Briarcliff Manor NY 10510

hobbs at electrooptical dot net
http://electrooptical.net

Ian Field · Jun 16, 2017

"Bob Engelhardt" <BobEngelhardt@comcast.net> wrote in message
news

i11mm0etq@news6.newsguy.com...

Here's really basic question for s.e.BASICS:
I want to use a MOSFET to control a load, via a switch. The tricky part
is that I need load current when the switch is open and no current when
the switch is closed. What I am confused about is what type of MOSFET to
use, and its orientation in the circuit (S & D connected which way?)

+
|
L
|
|
/ |-
+ --/ o--+--|
| |-
R |
| |
G G

I could put a BJT inverter before the MOSFET, but is there a way that I
don't have to?

The source follower configuration is non inverting.

Assume the load is from source to GND and you pull the source up to pass
current through the load - the source will be up at Vdd and the drive will
have to be at least VGSthr higher than Vdd. You can get away with a high
side driver, but many bootstrap the gate supply, those only work if the
output is a train of pulses. If the PSU is an old iron core transformer, you
can use a charge pump voltage doubler to get the gate supply - you might get
away with it on a SMPSU secondary, but it can be a bit more tricky.

If the load must hang from Vdd, use the opposite channel polarity MOSFET -
all the same basic principles aply.

rickman · Jun 17, 2017

Ian Field wrote on 6/16/2017 4:39 PM:

"Bob Engelhardt" <BobEngelhardt@comcast.net> wrote in message
newsi11mm0etq@news6.newsguy.com...
Here's really basic question for s.e.BASICS:
I want to use a MOSFET to control a load, via a switch. The tricky part
is that I need load current when the switch is open and no current when
the switch is closed. What I am confused about is what type of MOSFET to
use, and its orientation in the circuit (S & D connected which way?)

+
|
L
|
|
/ |-
+ --/ o--+--|
| |-
R |
| |
G G

I could put a BJT inverter before the MOSFET, but is there a way that I
don't have to?

The source follower configuration is non inverting.

Assume the load is from source to GND and you pull the source up to pass
current through the load - the source will be up at Vdd and the drive will
have to be at least VGSthr higher than Vdd. You can get away with a high
side driver, but many bootstrap the gate supply, those only work if the
output is a train of pulses. If the PSU is an old iron core transformer, you
can use a charge pump voltage doubler to get the gate supply - you might get
away with it on a SMPSU secondary, but it can be a bit more tricky.

If the load must hang from Vdd, use the opposite channel polarity MOSFET -
all the same basic principles aply.

You aren't looking at it the right way. The difference between putting the
load on the source or the drain is one of voltage polarity. In both cases
the current flows when the gate is more positive and turns off when the gate
is negative.

One thing he could do is to use a P-channel device. That would turn on when
the gate is more negative and off when the gate is more positive. I didn't
think of that before... lol

--

Rick C

Bob Engelhardt · Jun 17, 2017

On 6/16/2017 2:28 PM, rickman wrote:

> Can the switch be wired to ground instead of V+? ...

No - it does other stuff too and a double pole isn't available in the style.

Bob Engelhardt · Jun 17, 2017

On 6/16/2017 7:25 PM, rickman wrote:
....

One thing he could do is to use a P-channel device. That would turn on
when the gate is more negative and off when the gate is more positive.
I didn't think of that before... lol

That's the kind of answer that I was looking for - thanks. So it would
be like this (P-channel, enhancement mode)? :

+
|
/ |- S
+ --/ o--+--|
| |- D
R |
| |
G L
|
G

Switch open: Vgs = -V+, MOSFET on
Switch closed: Vgs = 0, MOSFET off

Rheilly Phoull · Jun 17, 2017

On 6/17/17 12:44 PM, Bob Engelhardt wrote:

On 6/16/2017 7:25 PM, rickman wrote:
...
One thing he could do is to use a P-channel device. That would turn on
when the gate is more negative and off when the gate is more positive.
I didn't think of that before... lol

That's the kind of answer that I was looking for - thanks. So it would
be like this (P-channel, enhancement mode)? :

+
|
/ |- S
+ --/ o--+--|
| |- D
R |
| |
G L
|
G

Switch open: Vgs = -V+, MOSFET on
Switch closed: Vgs = 0, MOSFET off

YUP!!

Jun 17, 2017

"So it would
be like this (P-channel, enhancement mode)? : "

It has to be enhancement mode or you will need a voltage source higher than the FET source to turn it off.

Bob Engelhardt · Jun 17, 2017

On 6/17/2017 3:25 AM, jurb6006@gmail.com wrote:

"So it would
be like this (P-channel, enhancement mode)? : "

It has to be enhancement mode ...

Thanks.

Furturlec has a bunch of P-channel MOSFETs & I picked IRFL9110 more or
less randomly. Looking at the IR spec sheet for it, it doesn't say
"enhancement":
http://www.futurlec.com/Transistors/IRFL9110pr.shtml
The Mouser listing for it does say it's enhancement, but why isn't it on
the spec sheet? Is it because enhancement is the default mode & it
"goes without saying"?

Bob

rickman · Jun 17, 2017

Bob Engelhardt wrote on 6/17/2017 10:10 AM:

On 6/17/2017 3:25 AM, jurb6006@gmail.com wrote:
"So it would
be like this (P-channel, enhancement mode)? : "

It has to be enhancement mode ...

Thanks.

Furturlec has a bunch of P-channel MOSFETs & I picked IRFL9110 more or less
randomly. Looking at the IR spec sheet for it, it doesn't say "enhancement":
http://www.futurlec.com/Transistors/IRFL9110pr.shtml
The Mouser listing for it does say it's enhancement, but why isn't it on the
spec sheet? Is it because enhancement is the default mode & it "goes
without saying"?

Look at the threshold voltage. That is what you care about. Make sure your
gate drive is significantly more than that wrt the source. This part has a
threshold of -4 volts max, so the polarity is right, but what is your gate
drive? Since you are using it as a high side driver you need to bring the
gate to less than 2 volts of the positive rail and pull down so the gate is
at least 4 volts more negative than the source. Should I assume the control
voltage through the switch is the same as the V+ on the source?

It may also be important to consider the load. What are you driving?

--

Rick C

John Larkin · Jun 17, 2017

On Sat, 17 Jun 2017 10:10:24 -0400, Bob Engelhardt
<BobEngelhardt@comcast.net> wrote:

On 6/17/2017 3:25 AM, jurb6006@gmail.com wrote:
"So it would
be like this (P-channel, enhancement mode)? : "

It has to be enhancement mode ...

Thanks.

Furturlec has a bunch of P-channel MOSFETs & I picked IRFL9110 more or
less randomly. Looking at the IR spec sheet for it, it doesn't say
"enhancement":
http://www.futurlec.com/Transistors/IRFL9110pr.shtml
The Mouser listing for it does say it's enhancement, but why isn't it on
the spec sheet? Is it because enhancement is the default mode & it
"goes without saying"?

Bob

Yes; depletion mosfets are rare and they make a point about it.

Look at the transfer curves and gate threshold specs to see what sorts
of gate voltages make what kind of drain currents.

I don't think I've ever seen a p-channel depletion mosfet. Well, not
in 30 years or so.

--

John Larkin Highland Technology, Inc

lunatic fringe electronics

Jim Thompson · Jun 17, 2017

On Sat, 17 Jun 2017 11:14:17 -0700, John Larkin
<jjlarkin@highlandtechnology.com> wrote:

On Sat, 17 Jun 2017 10:10:24 -0400, Bob Engelhardt
BobEngelhardt@comcast.net> wrote:

On 6/17/2017 3:25 AM, jurb6006@gmail.com wrote:
"So it would
be like this (P-channel, enhancement mode)? : "

It has to be enhancement mode ...

Thanks.

Furturlec has a bunch of P-channel MOSFETs & I picked IRFL9110 more or
less randomly. Looking at the IR spec sheet for it, it doesn't say
"enhancement":
http://www.futurlec.com/Transistors/IRFL9110pr.shtml
The Mouser listing for it does say it's enhancement, but why isn't it on
the spec sheet? Is it because enhancement is the default mode & it
"goes without saying"?

Bob

Yes; depletion mosfets are rare and they make a point about it.

Look at the transfer curves and gate threshold specs to see what sorts
of gate voltages make what kind of drain currents.

I don't think I've ever seen a p-channel depletion mosfet. Well, not
in 30 years or so.

480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |

Thinking outside the box... producing elegant solutions.

"It is not in doing what you like, but in liking what you do that
is the secret of happiness." -James Barrie

Bob Engelhardt · Jun 18, 2017

On 6/17/2017 12:34 PM, rickman wrote:

Look at the threshold voltage. That is what you care about. Make sure
your gate drive is significantly more than that wrt the source. ...

Oh, wait ... "more than the source"? Vgs needs to be negative, doesn't
it, for P-channel?

Every V+ is a 12v battery. The load is 2-90ma relay coils. Bandwidth is
in the 1 Hz range <G>.

Thanks,
Bob

Bob Engelhardt · Jun 18, 2017

On 6/17/2017 2:14 PM, John Larkin wrote:

Yes; depletion mosfets are rare and they make a point about it.

Look at the transfer curves and gate threshold specs to see what sorts
of gate voltages make what kind of drain currents.

I don't think I've ever seen a p-channel depletion mosfet. Well, not
in 30 years or so.

Thanks

whit3rd · Jun 18, 2017

On Friday, June 16, 2017 at 9:45:34 PM UTC-7, Bob Engelhardt wrote:

On 6/16/2017 7:25 PM, rickman wrote:
...
One thing he could do is to use a P-channel device. ...

That's the kind of answer that I was looking for - thanks. So it would
be like this (P-channel, enhancement mode)? :

Probably. There's more to it, like maybe a zener (8V plus or minus two) to
limit the gate-to-source voltage. A PNP transistor, with only a limit
resistor to the base, would work, too. The load could stand a snubber
diode, if it's a relay coil.

rickman · Jun 18, 2017

Bob Engelhardt wrote on 6/17/2017 7:47 PM:

On 6/17/2017 12:34 PM, rickman wrote:

Look at the threshold voltage. That is what you care about. Make sure
your gate drive is significantly more than that wrt the source. ...

Oh, wait ... "more than the source"? Vgs needs to be negative, doesn't it,
for P-channel?

Every V+ is a 12v battery. The load is 2-90ma relay coils. Bandwidth is in
the 1 Hz range <G>.

"More" in the sense of magnitude. With a P-channel FET it works just like
an N-channel FET but the polarities are opposite. The gate threshold for
this part is 4 volts max, so you need to provide more than 4 volts drive.
Just pay attention to the polarities required and it will be fine.

--

Rick C

Jun 18, 2017

It (IRFL9110)) looks like a good choice, but of course if you got with P channel there is inversion. It is not an insurountable problem. The cheapest transistor in the world can do it, and in SMD still no holes.

If you want to just drive it in polarity (most people call it phase) you need a special device unless the current involved is very low. But unless you go to great (I think because I am lazy LOL) lengths to do this the P channels with NPN bipolars on them and be done with it.

I heard nothing about Âź, Âž or whatever being an acceptable output.. I have been assuming you want the FET as about saturated as needs to be under reason. What is reasonable ?

Ian Field · Jun 19, 2017

"rickman" <gnuarm@gmail.com> wrote in message
news

i1p5l$kmo$1@dont-email.me...

Ian Field wrote on 6/16/2017 4:39 PM:

"Bob Engelhardt" <BobEngelhardt@comcast.net> wrote in message
newsi11mm0etq@news6.newsguy.com...
Here's really basic question for s.e.BASICS:
I want to use a MOSFET to control a load, via a switch. The tricky part
is that I need load current when the switch is open and no current when
the switch is closed. What I am confused about is what type of MOSFET
to
use, and its orientation in the circuit (S & D connected which way?)

+
|
L
|
|
/ |-
+ --/ o--+--|
| |-
R |
| |
G G

I could put a BJT inverter before the MOSFET, but is there a way that I
don't have to?

The source follower configuration is non inverting.

Assume the load is from source to GND and you pull the source up to pass
current through the load - the source will be up at Vdd and the drive
will
have to be at least VGSthr higher than Vdd. You can get away with a high
side driver, but many bootstrap the gate supply, those only work if the
output is a train of pulses. If the PSU is an old iron core transformer,
you
can use a charge pump voltage doubler to get the gate supply - you might
get
away with it on a SMPSU secondary, but it can be a bit more tricky.

If the load must hang from Vdd, use the opposite channel polarity
MOSFET -
all the same basic principles aply.

You aren't looking at it the right way. The difference between putting
the load on the source or the drain is one of voltage polarity. In both
cases the current flows when the gate is more positive and turns off when
the gate is negative.

Wrong - the other option uses a P-channel MOSFET.

Ian Field · Jun 19, 2017

<jurb6006@gmail.com> wrote in message
news:4c25d1d0-9141-48b3-8ade-6e36801e882f@googlegroups.com...

"So it would
be like this (P-channel, enhancement mode)? : "

It has to be enhancement mode or you will need a voltage source higher
than the FET source to turn it off.

Eh?!!! - depletion mode needs the gate *LOWER* than the source to pinch it
off.

MOSFET: N- or P-, enhance or deplete?

Bob Engelhardt

Guest

rickman

Guest

Phil Hobbs

Guest

Ian Field

Guest

rickman

Guest

Bob Engelhardt

Guest

Bob Engelhardt

Guest

Rheilly Phoull

Guest

Guest

Bob Engelhardt

Guest

rickman

Guest

John Larkin

Guest

Jim Thompson

Guest

Bob Engelhardt

Guest

Bob Engelhardt

Guest

whit3rd

Guest

rickman

Guest

Guest

Ian Field

Guest

Ian Field

Guest

Log in

Welcome to EDABoard.com

Sponsor