PWM again

[Brian Danielson <BDANIELS@stb.com>]

Forgive me for not being more clear in my original message. Yes, I am
doing PWM. The "on/off" idea was really PWM (i.e. on/off real fast! :-).
Sorry for misleading you. If this changes the technical aspects of your
answer, and if you have a moment, please let me know. Thanks.

	-----Original Message-----
	From:	Subscuba [SMTP:Subscuba@aol.com]
	Sent:	Wednesday, May 06, 1998 5:34 AM
	To:	BDANIELS@STB.com
	Cc:	personal_submersibles@psubs.org
	Subject:	Re: PWM

	Brian,

	If I understand your question correctly, all you want is ON/OFF
switches to
	interface between your 5 volt logic and the motor including
reversing.

	I would suggest a  full bridge also called a H bridge of four
MOSFETs. You
	could use two P channel FETs for the upper devices and two N
channel FETs for
	the lower devices. The P channels could be International
Rectifier IRF4905-ND.
	The lower FETs  IRFZ46N-ND. The power dissipation would be about
3 watts per
	FET when the FET is on and carrying 10 amperes. Heat sinks would
be required.
	The devices can be purchased from Digi-Key at 1-800-4539. This
assumes that
	you are not using PWM techniques and that you would only switch
the motor on
	or off. This also permits you to use the body diodes in the FETs
for the motor
	inductance flyback (a sometimes risky thing to do due to body
diode reverse
	recovery). Take a look at the Unitrode 1997 applications manual
note U-102
	Figure 8 page 3-73 for an example of the H bridge topology. The
UC1637 could
	be used to do what you want with some interface circuits. This
application
	note would prove useful to you and shows a PWM technique with
current limit.
	Unitrode has a website you can download from and you should be
able to find
	this application note (http://www.unitrode.com). As you are
aware don't take
	application notes too seriously, they are primarily sales
oriented and have
	disclaimers.

	Be aware that the di/dt of modern MOSFETs can exceed the di/dt
of Emitter
	Coupled Logic (ECL) by 100 times so layout is critical. Lead
lengths to the
	motor is not very critical but to the power source is. Local
bypassing with
	low Q capacitors is usually required.

	I am also a EE. Shayne, you might also be interested in the
application note.

	Brian, copy to Psubs when you send messages to me. Thanks!

	Good Luck,

	Ken Martindale