LM393 driving small MOSFET

I

ian field

Guest
My problem is "how fast can an LM393 drive the gate of a small MOSFET"? At
the moment the exact type of MOSFET isn't finalised, and even if it was I've
yet to see a datasheet give a straight answer on the actual value of gate
capacitance, if I knew that I could calculate the charge/discharge time with
the minimum allowable 393 O/P pull up resistor.

One thing that would ease the problem, is if it is possible to wire the pair
of comparators in parallel to increase the O/P sink capability, the
alternative is to drive an external transistor to get more current - but the
whole point of using the 393 was to eliminate a quantity of discrete
transistors.

TIA.
 
ian field wrote:

My problem is "how fast can an LM393 drive the gate of a small MOSFET"? At
the moment the exact type of MOSFET isn't finalised, and even if it was I've
yet to see a datasheet give a straight answer on the actual value of gate
capacitance,
Really ?


if I knew that I could calculate the charge/discharge time with
the minimum allowable 393 O/P pull up resistor.
Indeed.


One thing that would ease the problem, is if it is possible to wire the pair
of comparators in parallel to increase the O/P sink capability, the
alternative is to drive an external transistor to get more current - but the
whole point of using the 393 was to eliminate a quantity of discrete
transistors.
You might be better off with an op-amp then.

Graham
 
"Eeyore" <rabbitsfriendsandrelations@hotmail.com> wrote in message
news:45DA2E9E.B3D12FCC@hotmail.com...
ian field wrote:

My problem is "how fast can an LM393 drive the gate of a small MOSFET"?
At
the moment the exact type of MOSFET isn't finalised, and even if it was
I've
yet to see a datasheet give a straight answer on the actual value of gate
capacitance,

Really ?


if I knew that I could calculate the charge/discharge time with
the minimum allowable 393 O/P pull up resistor.

Indeed.


One thing that would ease the problem, is if it is possible to wire the
pair
of comparators in parallel to increase the O/P sink capability, the
alternative is to drive an external transistor to get more current - but
the
whole point of using the 393 was to eliminate a quantity of discrete
transistors.

You might be better off with an op-amp then.

Graham
Already looked at OP-AMPs - with the low Vcc headroom none of the devices I
looked at could guarantee close enough to true "rail to rail" O/P.
 
ian field wrote:

"Eeyore" <rabbitsfriendsandrelations@hotmail.com> wrote in message

You might be better off with an op-amp then.

Graham

Already looked at OP-AMPs - with the low Vcc headroom none of the devices I
looked at could guarantee close enough to true "rail to rail" O/P.
What supply voltage is available ?

Graham
 
"Eeyore" <rabbitsfriendsandrelations@hotmail.com> wrote in message
news:45DB2F23.4D647D98@hotmail.com...
ian field wrote:

"Eeyore" <rabbitsfriendsandrelations@hotmail.com> wrote in message

You might be better off with an op-amp then.

Graham

Already looked at OP-AMPs - with the low Vcc headroom none of the devices
I
looked at could guarantee close enough to true "rail to rail" O/P.

What supply voltage is available ?

Graham
5.4 - 7.2V
 
"ian field" <dai.ode@ntlworld.com> wrote in message
news:UnoCh.19039$s47.4068@newsfe4-gui.ntli.net...
My problem is "how fast can an LM393 drive the gate of a small MOSFET"? At
the moment the exact type of MOSFET isn't finalised, and even if it was
I've yet to see a datasheet give a straight answer on the actual value of
gate capacitance, if I knew that I could calculate the charge/discharge
time with the minimum allowable 393 O/P pull up resistor.
A 2N7000 or 2N7002 is a "small" MOSFET. The data sheet shows that about 1 nC
of gate charge will get it through the Miller region. The 393 output current
is specified at 6 mA minimum so you will be using about 1200R pull-up. That
should get you about 270 ns if I've done my arithmetic correctly.

Good luck.

John
 
ian field wrote:

"Eeyore" <rabbitsfriendsandrelations@hotmail.com> wrote in message
ian field wrote:
"Eeyore" <rabbitsfriendsandrelations@hotmail.com> wrote in message

You might be better off with an op-amp then.

Already looked at OP-AMPs - with the low Vcc headroom none of the devices
I
looked at could guarantee close enough to true "rail to rail" O/P.

What supply voltage is available ?

Graham


5.4 - 7.2V
I see. Is that even enough to turn on the device well ?

Graham
 
John wrote:

"ian field" <dai.ode@ntlworld.com> wrote in message

My problem is "how fast can an LM393 drive the gate of a small MOSFET"? At
the moment the exact type of MOSFET isn't finalised, and even if it was
I've yet to see a datasheet give a straight answer on the actual value of
gate capacitance, if I knew that I could calculate the charge/discharge
time with the minimum allowable 393 O/P pull up resistor.

A 2N7000 or 2N7002 is a "small" MOSFET. The data sheet shows that about 1 nC
of gate charge will get it through the Miller region. The 393 output current
is specified at 6 mA minimum so you will be using about 1200R pull-up. That
should get you about 270 ns if I've done my arithmetic correctly.
It'll pull up faster with a constant current load but then you're back to using
discretes.

Graham
 
"Eeyore" <rabbitsfriendsandrelations@hotmail.com> wrote in message
news:45DBF0E0.B781ED01@hotmail.com...
ian field wrote:

"Eeyore" <rabbitsfriendsandrelations@hotmail.com> wrote in message
ian field wrote:
"Eeyore" <rabbitsfriendsandrelations@hotmail.com> wrote in message

You might be better off with an op-amp then.

Already looked at OP-AMPs - with the low Vcc headroom none of the
devices
I
looked at could guarantee close enough to true "rail to rail" O/P.

What supply voltage is available ?

Graham


5.4 - 7.2V

I see. Is that even enough to turn on the device well ?

Graham
Which device are you referring to? The comparator should be perfectly fine
on TTL supply and the MOSFETs I have to hand have a Vgs-thr upper tolerance
of about 3 or 4V.

The Vgs-thr lower tolerance is about 1V, looking at the graph for the O/P
transistor Vce-sat V's sink current shows it reaches 1V at just over 10mA.
Don't forget the transistor has to discharge the gate capacitance as well as
sinking the load resistor current, since the MOSFET is driving a flyback
inductor the cut off needs to be as fast as possible.
 
"John" <groups5not@verizon.net> wrote in message
news:erg53c$jtf$1@aioe.org...

.. The 393 output current
is specified at 6 mA minimum so you will be using about 1200R pull-up.
That should get you about 270 ns if I've done my arithmetic correctly.

Good luck.

John

Thanks for that, the data sheet I have neglected to mention that figure.

By the look of your figures it should be plenty fast enough, I'll have to
look if the data sheet for the MOSFET gives that parameter.

Thanks again.
 
"Eeyore" <rabbitsfriendsandrelations@hotmail.com> wrote in message
news:45DBF128.F01AB01@hotmail.com...
John wrote:

"ian field" <dai.ode@ntlworld.com> wrote in message

My problem is "how fast can an LM393 drive the gate of a small MOSFET"?
At
the moment the exact type of MOSFET isn't finalised, and even if it was
I've yet to see a datasheet give a straight answer on the actual value
of
gate capacitance, if I knew that I could calculate the charge/discharge
time with the minimum allowable 393 O/P pull up resistor.

A 2N7000 or 2N7002 is a "small" MOSFET. The data sheet shows that about 1
nC
of gate charge will get it through the Miller region. The 393 output
current
is specified at 6 mA minimum so you will be using about 1200R pull-up.
That
should get you about 270 ns if I've done my arithmetic correctly.

It'll pull up faster with a constant current load but then you're back to
using
discretes.

Graham
The main motivation for using a chip was I'd run out of E-line BC182s and
the TO92 equivalents use more strip board because they're bulkier, however
someone left a Sanyo TV in the bin room and it had loads of 2SC1740s which
are even smaller than E-line and almost identical spec to the 182.

OTOH the figures John provided suggest I shouldn't be having difficulty with
this, if the figures translate well into the MOSFET I'm using the comparator
circuit should switch more efficiently than the discretes with little
difference in board size.

Thanks all.
 

Welcome to EDABoard.com

Sponsor

Back
Top