[chbot] H-Bridge (was Re: Sensor Information)

Charles Manning manningc2 at actrix.gen.nz
Wed Mar 25 23:11:59 GMT 2009


Yes, that is true. I misspoke by saying there is *no* voltage loss.

What I should have said is that there is substantially less voltage loss than 
when compared to a simple common collector Hbridge which will commonly have 
voltage losses of around 1V per transistor, due mainly to the problem that 
the Vbe is limited by the voltage rails.

With BD139/BD140s etc we're typically seeing a Vce loss of 0.2V per transistor 
(ie a total of approx 0.4V in an H-bridge) which is far better than losing 2V 
or more per H-Bridge. That makes a very noticeable difference in heat, speed 
and torque in the 6V systems we work with most of the time. 



On Thursday 26 March 2009 11:15:39 Michael Pearce wrote:
> > This circuit is common emitter so there is no voltage loss across the
> > main driving transistors.
>
> Not a true statement they are still BJT's so there is still a
> significant voltage drop (0.2 to  0.7V per transistor).
>
> Modern MOSFET's with low RDS-ON  (<0.02ohm) will be much less lossy.
> Can get MOSFETS with 20 Amp peak and 4 Amp continuous in a SOT-23 or
> smaller package.
> But Bridges are a little more complex to make using FET's (May Need
> dead time gaps etc)
>
>
> Mike
>
> 2009/3/25 Charles Manning <manningc2 at actrix.gen.nz>:
> > Herewith a reasonably good circuit for a simple H-bridge, roughly done
> > but the idea is there.
> >
> >
> > It was designed by Tilden (the BEAM guy) and we've used them a bit.
> >
> > How it works:
> >
> > The inputs go to the resistors at the left and right of the schematic.
> >
> > Let's consider an input on R1. When it does high this causes T6 to
> > conduct which, in turn turns on T4 and T2. causing the motor to be
> > powered in one direction.
> >
> > When R3 is taken high, then T7, T5 and T1 conduct, turning the motor the
> > other way.
> >
> > If all the above were turned on simultaneously there would be a short
> > circuit, so T3 is added as an "anti-smoke" transistor. If R1 input goes
> > high then so does R4, causing T3 to turn on which forces T7 off, thus
> > saving the day.
> >
> > This circuit is common emitter so there is no voltage loss across the
> > main driving transistors.
> >
> > You need to experiment a bit with resistors. I would suggest starting
> > with R1, R4, R3 all being 10k and R2 being 100 ohms.
> >
> > R1 and (R3 + R2) control the maximum current so changing these to 1k
> > would increase the current.
> >
> > T1,2,4,5 are the main power transistors.
> >  T3,T6 and T7 can be low power devices.
> >
> >
> > Have fun.
> >
> > -- CHarles
> >
> > On Wednesday 25 March 2009 17:55:12 Hanno Sander wrote:
> >> Agree with Charles- from personal experience!
> >> Don't use the L293 for low voltage (<12V) systems!  You'll lose too much
> >> voltage to heat the device.  Best bet is probably to make your own
> >> h-bridge - it's not that hard...
> >> Hanno
> >>
> >> 2009/3/25 Charles Manning <cdhmanning at gmail.com>
> >>
> >> > I would not put the L298 high on the list because it is a common
> >> > collector device, just like the L293.
> >> >
> >> > The internal configuration of these devices means that the output
> >> > voltage does not swing all the way to the rail and you lose approx 1.5
> >> > to 2 volts in the driver chip. This turns into heat, but also means
> >> > that your motor does not get driven as hard as it should meaning that
> >> > it does not get enough Weetbix to win that Sumo competition. That is
> >> > perhaps not too much of an issue when you're using 12V or more input,
> >> > but obviously throwing away 1.5V or so is bad news for low voltage
> >> > systems.
> >> >
> >> > Many modern HBridge designs are "rail to rail" and do not suffer this
> >> > problem.
> >> >
> >> >
> >> >
> >> >
> >> > On Wed, Mar 25, 2009 at 3:54 PM, Andrew Errington <
> >> >
> >> > a.errington at lancaster.ac.uk> wrote:
> >> >> I actually like this design:
> >> >>
> >> >> http://www.mcmanis.com/chuck/Robotics/tutorial/h-bridge/bjt-circuit.h
> >> >>tml
> >> >>
> >> >> Because the opto circuit is very elegant (it combines optical
> >> >> isolation with 'smoke-free' logic).  I've never built one as I was
> >> >> not able to find a fast enough opto-isolator to do high-frequency
> >> >> PWM.   It may be that that is a non-issue however.
> >> >>
> >> >> Good luck with whatever design you choose.
> >> >>
> >> >> Andrew
> >> >>
> >> >>
> >> >> _______________________________________________
> >> >> Chchrobotics mailing list Chchrobotics at lists.linuxnut.co.nz
> >> >> http://lists.ourshack.com/mailman/listinfo/chchrobotics
> >> >> web site: http://kiwibots.org
> >> >> When replying, please edit your Subject line to reflect new content.
> >> >
> >> > _______________________________________________
> >> > Chchrobotics mailing list Chchrobotics at lists.linuxnut.co.nz
> >> > http://lists.ourshack.com/mailman/listinfo/chchrobotics
> >> > web site: http://kiwibots.org
> >> > When replying, please edit your Subject line to reflect new content.
> >
> > _______________________________________________
> > Chchrobotics mailing list Chchrobotics at lists.linuxnut.co.nz
> > http://lists.ourshack.com/mailman/listinfo/chchrobotics
> > web site: http://kiwibots.org
> > When replying, please edit your Subject line to reflect new content.





More information about the Chchrobotics mailing list