Check out TI's SN754410. It is pin compatible with the L293D,
but can handle more current. I have found that both these chips
can handle more than twice the rated current if it is intermittent
and you have good heat dissipation through the ground pins
(Although I would not recommend this if you are designing life
Both chips are based on MOSFETs, so another option is to wire them
up in parallel for extra current.
The next step up is the LMD18200, which is rated for 3 Amps. Beyond
that, I think the only option is to use discrete components.
How much current do you need? You may be able to reduce the current
by using a motor wound for a higher voltage. 1 Amp at 12 volts will
give you as much power as 2 Amps at 6 Volts.
Actually, the L293D and 754410 are built of bipolar technology and not
MOSFET. There are a bunch of other h-bridge chips which are built on
MOSFET tech, like L620x series from STMico.
- dan michaels
I've never used the 754410, but I believe that it is a drop
in replacement for L293, not the L293D. The L293D has the
clamping diodes built in and the L293 does not. The spec.
sheet for the 754410 is a little ambiguous on this point since
it shows clamping diodes in the output driver schematic, but
do not mention them at the front of the document. The "typical
application" shows externally provided clamping diodes.
> There are a bunch of other h-bridge chips which are built on
The venerable L298 is a dual H-bridge where each side can handle
2 Amps. You can double up the two H-bridges to get 3.5 amps
as specified in the application note. L298's are cheap and available
from lots of sources (just type L298 into <http://www.findchips.com/
Be sure to provide external clamping diodes and a decent heat sink.
Mouser sells the 532-566010B34, which is designed to fit on to the
Multiwatt-15 package of the L298.
After about 3.5 amps, the cost of the H-bridge chips starts to
exceed the cost of just building one out of discrete components.
Wayne, Older versions of ST's datasheets were also confusing here, as
they doubled-up the L293 with the L293D. The "typical application"
schematic shows outboard diodes as well. Only in the text did it mention
the output clamp diodes.
TI's datasheet shows the clamp diodes for the 754410 in its "typical of
all outputs" diagram, even though the "typical application" schematic
shows the diodes. In fact, it's the same diagram they use for their
Unitrode L293D part, which they also sell. They're just reusing art. The
754410 doesn't mention output diodes at all, though they are shown on
I think most folks like to use the external diodes anyway, especially
with larger motors. In which case there's also the L293E, which shares
the current capacity of the 754410, and does need the external diodes.
However, as you mention, as you go up the current ladder there are
plenty of other choices.
The thing I don't like about the L298 is the Multiwatt packaging, which
is an issue if you're just prototyping. Has someone yet come out with a
*cheap* carrier for 0.100" protoboards for these -- $1 rather than $5?
After you add carrier, heatsink, and external components, you're about
as compact as an old mercury-wetted relay.
I thought I'd throw out for the OP and others the long-standing issue of
minimum voltage for the typical high-current bridge. The latest ones use
DMOS, which typically have a 10V minimum. That's okay for 12V motors run
from 12V (or higher) battery supplies, but a lot of the non-motherboard
based robots have lower power sources.
As Gordon says, one of the downsides to the DMOS h-bridge chips is they
don't operate below 10-12v or so, so you can't use them very well with
a lot of the smaller cheapo motors. The bipolar h-bridges, like L29x,
aren't as efficient as DMOS but they will work for battery voltages
down to 6-8v or so. Given a 1-2v drop in the L29x chips, this will work
well for most small motors, I think. If you use the DMOS parts, then
you're putting about 10v on the motors, which is generally too much for
low-V motors - unless you like screaming bearings. There don't seem to
be many easy ways to drive low-volt bidirectional motors, other than
possibly building a bridge out of discrete power transistors, like you
find in all of the radio shack toys, etc.
One thing I'm playing with on my new controller board is using a
current limiter on the h-bridge - L298, not Gordon's favorite - to
limit current flow when driving smaller motors at higher voltages, help
limit stall currents, etc.
- dan michaels
I tend to put L298's onto printed circuit boards so the packaging is
not an issue for me.
Every time I look at the DMOS chips, I tend to get sticker shock.
The quantity one price for the LMD18200 is $10+ for a 3 Amp part.
The ancient technology L298 can do 3.5 amps for $3. I can not
figure out why the DMOS parts refuse to come down in price.
By the way, I think the LMD18200 is an excellent part and have
no reservations against using one. I just tend to use the L298
I seem to recall them being on .10" centers; just skew the pins on one row
.05" to fit. Allegro's SLA Multiwatt is really a SIP on (I think) 3mm
centers. That one's a bit more difficult to shove onto a protoboard.
I have used Texas Instruments L293DNE and also the SN754410 for my first
little robot. On their website
http://focus.ti.com/docs/prod/folders/print/sn754410.html it states that
for the SN754410 there are:
TTL- and CMOS-Compatible High-Impedance Diode-Clamped Inputs
I have never paid for L293DNE or SN754410, as I have gotten tham as free
samples from TI ;-)
My first robot used two Tamiya motors/twin gearbox kit ( 3-6V ) on a
Tamiya caterpillar tracks kit. I only supplied 4.8V to the motor power
ports of SN754410 motor driver ( and separate regulated power supply to
the control ports ) - it has an operating range of 4.5-36V making it good
for small motors on a little bot. I think from memory it dropped maybe
1.8V or maybe even less. This left plenty of juice for the small motors.
The beauty of the Tamiya twin gearbox is that you can configure the gears
to be either high speed or high torque, giving you good options with the
I used PWM using a PIC16F876A, and provided enough juice for the little
motors using only 4 1.2V NiMH cells.
By the way, the tracks kept coming off, so I don't think they are a good
option unless you can come up with a solution (let me know if anyone has
Polytechforum.com is a website by engineers for engineers. It is not affiliated with any of manufacturers or vendors discussed here.
All logos and trade names are the property of their respective owners.