I have a PIC controlled robot design which has a small 12VDC motor. The
motor speed is controlled by PWM (PIC generated) into a logic level
power FET. I now wish to be able to reverse the motor under control of
the PIC but I don't like the idea of using a relay.
I borrowed a FUTABA FET speed controller which worked well with the
motor but I can't justify purchasing one for a small hobby robot.
I have been told that a FET H-Bridge would be a possible solution. Are
there any H-bridge designs available on the web ? Are there any
alternatives to this type of circuit ?
An H-Bridge is indeed the way to go. Be VERY CAREFUL not to turn both
low and high drivers of the same side on at the same time, as it causes a
direct short across the power supply.
As an alternative, you can try an H-Bridge on a chip that has internal
logic to eliminate the crossover current problem, like the L293D from SGS
(Mouser carries them). Fewer parts, fewer I/O's to drive it, etc.
Easier all around.
>
>I have a PIC controlled robot design which has a small 12VDC motor. The
>motor speed is controlled by PWM (PIC generated) into a logic level
>power FET. I now wish to be able to reverse the motor under control of
>the PIC but I don't like the idea of using a relay.
>
>I borrowed a FUTABA FET speed controller which worked well with the
>motor but I can't justify purchasing one for a small hobby robot.
>
>I have been told that a FET H-Bridge would be a possible solution. Are
>there any H-bridge designs available on the web ? Are there any
>alternatives to this type of circuit ?
>
>Any advice very much appreciated.
Depends on your power requirement. Efficient motor and low power? Then try
the L293D. TI makes a pin compatible chip (SN754410) that has better
parameters, costs a $1.00 more. Probably worth it for us hobbyists .. easy
to use.
Relays are not so bad. Careful application of a diode can give you a nice
back E.M.F. protection (that will also stop the motor on a dime via active
braking if you turn it off then set the relay for reverse for a second or
two). Simple, effective, efficient, low cost... I'm not aware of any real
problems other then maybe wear leading to mechanical failure over time and
noise on some relays.
I have a PIC controlled robot design which has a
small 12VDC motor. The
motor speed is controlled by PWM (PIC generated)
into a logic level
power FET. I now wish to be able to reverse the
motor under control of
the PIC but I don't like the idea of using a relay.
I borrowed a FUTABA FET speed controller which
worked well with the
motor but I can't justify purchasing one for a small
hobby robot.
I have been told that a FET H-Bridge would be a
possible solution. Are
there any H-bridge designs available on the web ?
Are there any
alternatives to this type of circuit ?
The TI chip here is the L293DNE. It can handle up to 600 mA and up to
1200 mA max. I have used it before in one of my projects and have found
it to be easier to use than trying to build the h-bridge using individual
transistors. If the motor is going to change directions often, you may want
to purchase a heat sink for the chip. The current through the motors
increases
dramatically when they have to stop and reverse direction, making the chip
heat up.
In my project (a car that avoided objects) the motors could change direction
quite often, causing the chip to get rather warm...
and they are sold by Newark for about $3.00 a piece. I think there is a
minimum
order to $25.00.
search.newark.com/part_detail.phtml?PART%5FID=250&VID=250&10005=06F95
23
This link will take you directly to the info on the chip, otherwise, their
normal web
page is http://www.newark.com
I guess I'm missing the point. Why is this better than using an L293D (500
ma) or a TI SN754410 (1000 ma) - augmenting both with inverters?? You still
have to provide PWM - which probably means an external MCU, you get about
the same response from the two chip setup, you have about the same number of
wires going to the chip(s). I looked at this design, then the one I'm
working on now and I have to use an inverter to get the effect I need (and I
suspect you do with the Wirz version), but the overall cost is 4.50 in small
quantities as opposed to 23.50. With the L293D/TI setup, you get both sign
magnitude and anti-phase , .1" PC board pinouts etc, etc.
> I guess I'm missing the point. Why is this better than using an L293D (500
> ma) or a TI SN754410 (1000 ma) - augmenting both with inverters?? You
still
> have to provide PWM - which probably means an external MCU, you get about
> the same response from the two chip setup, you have about the same number
of
> wires going to the chip(s). I looked at this design, then the one I'm
> working on now and I have to use an inverter to get the effect I need (and
I
> suspect you do with the Wirz version), but the overall cost is 4.50 in
small
> quantities as opposed to 23.50. With the L293D/TI setup, you get both sign
> magnitude and anti-phase , .1" PC board pinouts etc, etc.
>
> Am I missing something very kewl?
>
> Dan
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com
Ahh.. Drat! I was hoping we had a $20 PID/PWM motor controller in one
package. Not to be, I guess.
Dan
> Dan,
> I posted the link as a resource for you. Nothing more.
>
> John
> > I guess I'm missing the point. Why is this better than using an L293D
(500
> > ma) or a TI SN754410 (1000 ma) - augmenting both with inverters?? You
> still
> > have to provide PWM - which probably means an external MCU, you get
about
> > the same response from the two chip setup, you have about the same
number
> of
> > wires going to the chip(s). I looked at this design, then the one I'm
> > working on now and I have to use an inverter to get the effect I need
(and
> I
> > suspect you do with the Wirz version), but the overall cost is 4.50 in
> small
> > quantities as opposed to 23.50. With the L293D/TI setup, you get both
sign {Quote hidden}
> > magnitude and anti-phase , .1" PC board pinouts etc, etc.
> >
> > Am I missing something very kewl?
> >
> > Dan
>
>
>
> _________________________________________________________
> Do You Yahoo!?
> Get your free @yahoo.com address at http://mail.yahoo.com