Newsgroups: comp.robotics
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!udel!gatech!swrinde!cs.utexas.edu!news.sprintlink.net!news.indirect.com!seeker
From: seeker@indirect.com (Stan Eker)
Subject: Re: Efficient low-V H-bridges
Message-ID: <D7F8Fq.15B@indirect.com>
Sender: usenet@indirect.com (Internet Direct Admin)
Organization: Internet Direct, indirect.com
Date: Sat, 22 Apr 1995 05:26:13 GMT
References: <3n5su1$p7d$1@mhafc.production.compuserve.com>
X-Newsreader: TIN [version 1.2 PL2]
Lines: 41

Scott Edwards (72037.2612@CompuServe.COM) wrote:
: I have surveyed packaged H-bridge drivers available from National 
: Semiconductor, Allegro and others, and I've come to the following 
: (somewhat obvious) conclusions: 

: (1) Bipolar devices like the LM18293 (Nat'l version of L293D) will 
: work at low supply voltages, but their Darlington-transistor outputs
: have terrible Vce specs--something like 1.5 volts max. I believe 
: that's for both legs of the H bridge, so at 1 amp you're losing 3 
: watts of power. 

: (2) Mosfet devices like the LM18200 have low RDSon figures in the 
: neighborhood 0.5 ohms per leg, meaning that they'd waste 1 watt at 1
: amp, but the stated minimum operating voltage is 12 volts. 

: I realize that these problems are inherent characteristics of the 
: two semiconductor technologies used. What I'm interested in knowing 
: is whether there are  efficient, inexpensive H-bridge drivers for 
: small motors in the 3- to 7.5-volt range. If not, is there a market 
: for a highly efficent H-bridge module for these low voltages? My 
: thinking is that a four-cell battery pack (C or D, NiCd or alkaline)
: could be a really hot ticket for small bots if only it were used 
: efficiently. 

It's a common observation.  There's not really anything good unless you can
live within the constraints of the existing drivers.  I've been thinking of
producing a small board that's pin-compatible with the L293D, but haven't
done the research yet to find if I-R or Siliconix have a FET with 2-3A
capability in a TO-92 or TO-237 package, which'd be needed if it were to fit
within the same footprint.  Add a small CMOS charge-pump and some drive FETs
in SOT-23 and you'd have a nice little beastie that'd do many of us better.
It'd need to have Rds(on) of about .05 ohms to handle 12V motors, maybe .07
in the TO-237 package.  The IRFZ40 comes close, but the TO-220 package is
way too big to fit 8 of 'em on a 16-dip footprint board.

It's probably not impossible, but I just haven't checked into the current
crop of low-Rds(on) FETs in a few years, and I'm out of date...
If you're not limited to compatibility with the L293 and can afford the
room, you can do it *right now* for some hellaciously powerful hobby motors
instead of the 1-2A units most folks are forced into using.

