Driving Dual MOSFETs From A NPN Transistor

[Jamie]

Thanks Chris,

Actually I was playing with the circuit today and hit something that's
puzzling me (I wired it as a h-bridge)...

Could you explain why when one the inputs (the little squares at the base of
Q1 and Q17) is high, the motor turns as expected but only 5.51V reaches it
with the 30.49V being dropped accross the top mosfets?

I'm guessing it has something to do with the load connected to the source as
opposed to the drain.....

Schematic:
http://www.mhims.co.uk/MOSFET Schematic5.png

Cheers,

Michael

if you want my opinion? you need to use P and N Channel fets.
unless i have missed something, only one side of the bridge is
actually going to fully conduct.
The last time i knew, the voltage between the SS and Gate must
be what spec calls for to fully turn them on. Seeing that you have the
motor on side of the Set SS side, I'm guessing that you will not get a
full bias to the Gate unless you're using boot strap type FETS?, in
which case, you will see a difference in readings.
I could be wrong, just and observation.

 

[Chris]

Thanks Chris,

Actually I was playing with the circuit today and hit something that's
puzzling me (I wired it as a h-bridge)...

Could you explain why when one the inputs (the little squares at the base of
Q1 and Q17) is high, the motor turns as expected but only 5.51V reaches it
with the 30.49V being dropped accross the top mosfets?

I'm guessing it has something to do with the load connected to the source as
opposed to the drain.....

Schematic:
http://www.mhims.co.uk/MOSFET Schematic5.png

Cheers,

Michael

Geez, Michael. This is a SPICE project? And now there's a heretofore
undisclosed H-bridge?

Alright. A few more comments:

1) Your top (sourcing) MOSFETs are choking themselves off. As more
current is conducted, the source voltage relative to the gate rises.
That means it will stabilize with a gate voltage about 4.5V higher than
the source voltage. If you work it out (Vg = 10V), you can see it.
The remainder has to appear across the MOSFET -- there's noplace else
for it to be.

2) Typically, H-bridge drivers are done with N-type and P-type MOSFETs
-- N-type on the bottom and P-type on the top. I'm guessing this
confusion is part of SPICE syndrome. Do some research (meaning go to
st.com and Google and read a couple of appnotes and articles on
H-bridge drivers), and then come back enlightened.

3) You're showing that you've got eight MOSFETs in parallel for each
part of the driver. That's definitely a plus. However, the total
capacitance you're going to be driving (as well as the critical need to
turn the MOSFETs on and off quickly) mean that you should rebias the
NPN logic level conversion switch with lower ohm resistors. Also,
you'll want to set up a different driver for the P-type MOSFETs. And
use one resistor (22 ohms is probably better than 100) for each MOSFET.

4) If you're doing H-bridge drivers, you have to be very careful that
the top and bottom transistors aren't both on at the same time. That's
called shoot through, and it's fatal. 36V / 2.2 milliohms = flames.
Even for nanoseconds. It's important to establish a delay so that
doesn't happen. Usually the top one is turned off before the bottom
one is turned on, but it can be the other way just as easily. The
thing is, it makes more sense to use an H-bridge driver IC for this,
and hook it to your 4 sets of MOSFETs.

5) If you ever plan on hooking this up for real, don't forget to
include the diodes and watch your current paths and voltage drops.
But, since this is apparently SPICE world, I guess it's OK.

Good luck with your homework
Chris

 

[Michael]

Geez, Michael. This is a SPICE project? And now there's a heretofore

undisclosed H-bridge?

Nope, private project. Some motors need to run one way only, some need to
run both....

Alright. A few more comments:

1) Your top (sourcing) MOSFETs are choking themselves off. As more
current is conducted, the source voltage relative to the gate rises.
That means it will stabilize with a gate voltage about 4.5V higher than
the source voltage. If you work it out (Vg = 10V), you can see it.
The remainder has to appear across the MOSFET -- there's noplace else
for it to be.

I understand.....but how do you reduce the drop? By increasing the gate
voltage?

2) Typically, H-bridge drivers are done with N-type and P-type MOSFETs
-- N-type on the bottom and P-type on the top. I'm guessing this
confusion is part of SPICE syndrome. Do some research (meaning go to
st.com and Google and read a couple of appnotes and articles on
H-bridge drivers), and then come back enlightened.

I did alot of research before coming here (though obviously not enough) and
tried this circuit:
http://www.cadvision.com/blanchas/hexfet/np-s.pdf
But again, I got a huge voltage drop so I thought it might be my software
(Livewire) at fault...

3) <snipped>

4) <snipped>

5) If you ever plan on hooking this up for real, don't forget to
include the diodes and watch your current paths and voltage drops.
But, since this is apparently SPICE world, I guess it's OK.

Once I'm happy with the simulations I'll be building it....so it's not just
a little homework exercise so it were.....

Last question, I've spent the last hour googling to try and find an answer
and testing theories.....but is there anyway to turn off a P channel mosfet
by applying a voltage to the gate that isn't 4.5V (or there abouts) over the
source voltage? Nothing else seems to work.....

Sorry for all the questions but this thing is really starting to bug me and
with such large voltages and currents involved I'd like to have some idea of
what I'm doing.....

Cheers,

Michael

 

[Chris]

Nope, private project. Some motors need to run one way only, some need to
run both....


I understand.....but how do you reduce the drop? By increasing the gate
voltage?

I did alot of research before coming here (though obviously not enough) and
tried this circuit:
http://www.cadvision.com/blanchas/hexfet/np-s.pdf
But again, I got a huge voltage drop so I thought it might be my software
(Livewire) at fault...


Once I'm happy with the simulations I'll be building it....so it's not just
a little homework exercise so it were.....

Last question, I've spent the last hour googling to try and find an answer
and testing theories.....but is there anyway to turn off a P channel mosfet
by applying a voltage to the gate that isn't 4.5V (or there abouts) over the
source voltage? Nothing else seems to work.....

Sorry for all the questions but this thing is really starting to bug me and
with such large voltages and currents involved I'd like to have some idea of
what I'm doing.....

Cheers,

Michael

Hi, Michael. You got a large voltage drop on your sample circuit in
SPICE because you punched in N-channel MOSFETs for Q2 and Q6. Try it
again with Q2 and Q6 as P-channel MOSFETs, and see the difference.

To turn ON a P-channel MOSFET, the gate voltage has to be *lower* than
the source, just as a PNP transistor requires current to be drawn
*from* the base rather than injected into the base. Simple, really.

In the diagram, when logic input A goes high, it turns on transistor
Q1. That presents about +1V at the gate of the MOSFET. The 11V
difference between the gate (1V) and the source (12V) is enough to turn
on MOSFET Q2.

Look -- this circuit should work well for you if you can separately
control A/B/C/D, you use voltage translation to ensure that A/B/C/D are
36V-level signals, and you can ensure that there will be a delay of a
microsecond or ten between the time Q8 really turns off and Q2 turns
on. Same for the other way, and for the other pair. I'm not so
worried about slow turn on, but slow turnoff will kill you. Use 2N4401
/ 2N4403 transistors instead of 2N3904 / 2N3906 -- they have higher
V(ceo)(max).

I still think using a contactor/two contactors, or just using four
surplus honker automotive relays is a much better way for a newbie to
control this very serious motor. You might want to try All Electronics
24 VDC, SPDT 80 AMP RELAY CAT# RLY-472 for $2.95USD ea. (be sure to use
a series resistor that will drop the extra 12V):

http://www.allelectronics.com/cgi-bin/category/500700/Automotive_Power.html

Then just use four NPN darlington transistors to drive the relay coils.
It would be wise. Control of large motors with MOSFETs isn't for the
faint of heart or those without access to tubs or at least tubes of
N-channel and P-channel MOSFETs. And I'm not sure you can pull this
together with the background and experience level you're at.

Good luck
Chris

 

[Michael]

Hi, Michael. You got a large voltage drop on your sample circuit in
SPICE because you punched in N-channel MOSFETs for Q2 and Q6. Try it
again with Q2 and Q6 as P-channel MOSFETs, and see the difference.

To turn ON a P-channel MOSFET, the gate voltage has to be *lower* than
the source, just as a PNP transistor requires current to be drawn
*from* the base rather than injected into the base. Simple, really.

In the diagram, when logic input A goes high, it turns on transistor
Q1. That presents about +1V at the gate of the MOSFET. The 11V
difference between the gate (1V) and the source (12V) is enough to turn
on MOSFET Q2.

Look -- this circuit should work well for you if you can separately
control A/B/C/D, you use voltage translation to ensure that A/B/C/D are
36V-level signals, and you can ensure that there will be a delay of a
microsecond or ten between the time Q8 really turns off and Q2 turns
on. Same for the other way, and for the other pair. I'm not so
worried about slow turn on, but slow turnoff will kill you. Use 2N4401
/ 2N4403 transistors instead of 2N3904 / 2N3906 -- they have higher
V(ceo)(max).

I still think using a contactor/two contactors, or just using four
surplus honker automotive relays is a much better way for a newbie to
control this very serious motor. You might want to try All Electronics
24 VDC, SPDT 80 AMP RELAY CAT# RLY-472 for $2.95USD ea. (be sure to use
a series resistor that will drop the extra 12V):

http://www.allelectronics.com/cgi-bin/category/500700/Automotive_Power.html

Then just use four NPN darlington transistors to drive the relay coils.
It would be wise. Control of large motors with MOSFETs isn't for the
faint of heart or those without access to tubs or at least tubes of
N-channel and P-channel MOSFETs. And I'm not sure you can pull this
together with the background and experience level you're at.

Good luck
Chris

Thanks Chris,

I think it's best if I go with your method and used one 'set' of MOSFETs for
speed control and use a contactor or two for direction control. I take it I
just stick the contactor in above the motor on the schematic you designed
for me?

Cheers and apologies for not quite 'getting it',

Michael

 

[Michael]

To turn ON a P-channel MOSFET, the gate voltage has to be *lower* than
the source, just as a PNP transistor requires current to be drawn
*from* the base rather than injected into the base. Simple, really.

It turns out that changing the MOSFET from 'Ideal' to an actual model
changes it's properties quite considerably..

Michael

 

[Chris]

Thanks Chris,

I think it's best if I go with your method and used one 'set' of MOSFETs for
speed control and use a contactor or two for direction control. I take it I
just stick the contactor in above the motor on the schematic you designed
for me?

Cheers and apologies for not quite 'getting it',

Michael

Wow. And now speed control, too. This is starting to sound like nail
soup here.

If you want bidirectional motor control, you might go with four of the
automotive relays mentioned, andone set of the N-channel MOSFETs
between the bottom side of the relays and GND, like this (view in fixed
font or M$ Notepad):

| +36V
| |
| .-------o-----o-----o--------.
| | | | |
| | - - |
| | ^ ^ |
| | CRY1 | | CRY3 |
| | || | | || |
| '---||--o _ o--||----'
| || | / \ | ||
| o--o( M )o--o
| || | \_/ | ||
| .---||--' '--||----.
| | || || |
| | CRY2 CRY4 |
| | |
| '-------------o--------------'
| |
| .--------o-o-------o--------.
| | | | |
| ||-+ ||-+ ||-+ ||-+
| ||<- ||<- ||<- ||<-
| .-||-+ .-||-+ .-||-+ .-||-+
| | | | | | | | |
| .-. | .-. | .-. | .-. |
| 4 X 33| | === | | === | | === | | ===
| Control | | GND | | GND | | GND | | GND
| (10V = ON,'-' '-' '-' '-'
| 0V - OFF) | | | |
| >--o--------o---------o--------'
(created by AACircuit v1.28.6 beta 04/19/05 www.tech-chat.de)

This should work. Try it. And if you make sure all MOSFETs are off
and wait a second or so before switching anything off, your relays will
love you.

Good luck
Chris

 

[Michael]

Wow. And now speed control, too. This is starting to sound like nail
soup here.

If you want bidirectional motor control, you might go with four of the
automotive relays mentioned, andone set of the N-channel MOSFETs
between the bottom side of the relays and GND, like this (view in fixed
font or M$ Notepad):

| +36V
| |
| .-------o-----o-----o--------.
| | | | |
| | - - |
| | ^ ^ |
| | CRY1 | | CRY3 |
| | || | | || |
| '---||--o _ o--||----'
| || | / \ | ||
| o--o( M )o--o
| || | \_/ | ||
| .---||--' '--||----.
| | || || |
| | CRY2 CRY4 |
| | |
| '-------------o--------------'
| |
| .--------o-o-------o--------.
| | | | |
| ||-+ ||-+ ||-+ ||-+
| ||<- ||<- ||<- ||<-
| .-||-+ .-||-+ .-||-+ .-||-+
| | | | | | | | |
| .-. | .-. | .-. | .-. |
| 4 X 33| | === | | === | | === | | ===
| Control | | GND | | GND | | GND | | GND
| (10V = ON,'-' '-' '-' '-'
| 0V - OFF) | | | |
| >--o--------o---------o--------'
(created by AACircuit v1.28.6 beta 04/19/05 www.tech-chat.de)

This should work. Try it. And if you make sure all MOSFETs are off
and wait a second or so before switching anything off, your relays will
love you.

Good luck
Chris

Ok thanks Chris,

I'll give it a go in a couple of weeks and let you know,

Cheers,

Michael