Walter! how timely.  I just finished a DC motor speed controller for a client
 and
they rejected it.  I met all of the design goals, and my downfall was a very
slight ripple on the voltage on the motor.  So slight that you could not see the
variation in shaft motor.  The motors were Japanese, though, and not rated
horsepower.  I designed for two motors, a 45 volt and a 21 volt (same
controller).  I did a bang up job of controlling torque by monitoring back EMF
 on
the motor and comparing to command signal.  I used PI (skipped D) to perform
feedback.  I actually used a MOSFET, and when a very heavy load was put on the
motor, the controller shot current so high that I burnt paint off of the PCB.  I
had to use a lower RDSon FET and implement current limitting.  The answer to
 your
question is International Rectifier part IRF3710.  And 100 mil traces through
 the
current path.

There are other details, but I'll let you in on the biggest revelation.  You
 must
sample the motor voltage just before you begin your PWM on phase.  This way the
voltage has settled to as close to free wheel EMF as possible.  And you must PWM
much slower than the built in PWM so that this settling takes place. I found 60
to 120 Hz to be ideal.

Best of luck, need any more advice I'll throw in what I know.

Chris Eddy
Pioneer Microsystems, Inc.
Pittsburgh, PA

Walter Markiw wrote:

> Hi,
>    I need to drive a DC 1/8 HP DC motor via PWM using a 16c74.My problem is
> which transistors I should use to do this and how to trigger them.I have read
> something on IGBT being used in power inverters,but I have also been told
> they are somewhat more complicated to trigger than a regular MOSFET.
>    Could someone tell me what's my best choice?
>    Thanks.