V G wrote:
> In this case, what would be the difference here in using a MOSFET or
> a BJT?

Like everything, its a tradeoff.  There are mainly two issues to consider i=
n
this case, power dissipation when on and driving the transistor.

A FET looks largely like a low value resistor when on.  In fact, that's wha=
t
the Rdson spec is (resistance drain to source when on).  If Rdson is
100mOhms, then at 10A the FET will drop 1V and dissipate 10W.  Of course
100mOhms is quite high for a good FET rated only for low voltage, like 20V.
If it's 10mOhms then the voltage drop will be 100mV and it will dissipate
1W.

A bipolar looks more like a voltage source when saturated, although there
will be enough of a resistive component to matter at 10A.  At low currents =
a
power transistor can probably do less than 100mV.  At 10A a good one
probably a few 100mV.  At 500mV it would dissipate 5W.

Both FETs and bipolars will have better on-state specs when rated for lower
off-state voltages.  Bipolars can be made to have higher gain when the C-E
voltage doesn't need to be high, and FETs lower Rdson when the D-S voltage
doesn't need to be high.

The other major issue to consider is how to drive the transistors.  A FET i=
s
driven with voltage on the gate while a bipolar requires current thru the
base.  The problem with the voltage drive of a FET is that the voltage is
often inconvenient for a microcontroller.  To get otherwise the best specs,
a good FET will want 10-15V on the gate to be fully on.  Particularly at lo=
w
D-S voltage ratings, FETs can be made to require lower gate voltage.  Those
that work well with only 5V on the gate are sometimes called "logic level"
FETs.  You can get some pretty amazing FETs if you only require 20 or 30
volts capability.  Take a look at the IRLML2502 or IRLML0030 for example.
However impressive, they can't do 10A.  10A is a lot, and to keep the
disspation down you probably buy into the more complicated full voltage gat=
e
drive in this case.

A bipolar can be driven easily from low voltage circuits since the input
signal is essentially the current thru what looks like a silicon diode to
the driving circuit.  The base voltage even at high base currents will be
under 1V, usually around 750mV.  The problem here is that this current must
be sustained continuously while the transistor is on, and can be substantia=
l
in the context of a ordinary logic circuit.  Small signal transistors can
easily have gains of 100 or more (1A collector current requires only 10mA
base current), but transistors designed for high current like 10A are going
to have lower gain.  The fact that it only needs to be rated for 20 or 30
volts will help.  That allows tradeoffs in the design of the transistor in
favor of higher gain.  You can probably find a NPN transistor with a gain o=
f
around 100 in this range.  However, that still means you have to come up
with 100mA base drive to get 10A collector current.  That's doable, but not
straight out of a microcontroller pin.

There is no free lunch.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.