H-bridges - chip or discrete?


Senior Member
Stan Swan wrote:

Although I applaud making an H bridge for it's educational benefits,the L293D is THE industry standard approach for modest power motors & just HAS to be considered ! I recall it's arival was akin to the 555 (1972)whereby zillions of discretes were done away with. Crucial hence you feature it at your presentation...

Hi Stan,

As a relative newcomer to the Picaxe scene I highly appreciate the major contributions you have made to this community.

Please consider my comments then as quasi- complimentary with a hint of humor 8^).

To paraphrase Shrek: chip are like onions.

Yes, "consider" using a L293 but I respectfully disagree that it is "crucial to feature it" for many of the reasons you mentioned.

Given this is the Picaxe Forum, so single chip solutions like the L293 would seem a natural extension for reasons of economy and convenience.

But in an educational setting, a presentation using a discrete version of a functional block like a motor driver especially if it is a novel solution, may get you higher marks compared to using a single chip "black box".

Without peeling back the hidden layers, most people will never know just where all that magic smoke comes from.

Using a discrete designs not only reveals useful details, helpful to the learning process, but also let's you tweak performance for a specific application.

The L293 chip has been around for a long time and while it may be widely used by hobbyists now, imho it is not an ideal motor driver and in fact is rarely used in commercial applications such as printers, consumer electronics, toys, etc.

Yes, it is a single chip solution and yes it is robust. But note that the typical output voltage drop is 2.6V at 600mA which makes it very inefficient when used with the minimum motor supply voltage of 4.5V.

If you must use a motor driver chip for a low voltage application consider newer more efficient low saturation voltage CMOS types that are widely used in consumer electronics.

Having said that, thanks again for all your helpful writings on all things Picaxe.



NB: The original H-bridges? thread started by jadesteffen specifically requested an h-bridge design using only 2N2222 transistors and 1K resistors to drive toy motors, a solution presumably dictated by limited resources.

So I wrote:

This 6 transistor quasi-complementary full feature h-bridge can be found here.



New Member
thanks wilf, for posting the link to the scehmatic, but my membership to the beam group still hasn't been aproved yet. So, i'm still trying to use ylp88's oringinal design, with your sugestion of adding resistors. Unfortunately, i only have five 1k resistors, so my question is, Are the resistors needed?

thanks for all your help!

Edited by - jadesteffen on 1/16/2006 10:58:55 PM


Senior Member
OK on discretes &amp; bravo! To really show the &quot;H&quot; concept, &amp; as a great intro to transistor use, suggest you FIRST demo a version with switches &amp;/or relays - the action is usually SO clear that even snoozing dullards in the back row (ahem -the staff I mean- NOT the students !)will get the point &amp; be keen for the extension. See =&gt;www.rctankcombat.com/articles/speed-control/081Hbridge.gif <A href='http://www.rctankcombat.com/articles/speed-control/081Hbridge.gif' Target=_Blank>External Web Link</a>

If you're being acessed by your peers, then trivial little demos like this can earn HIGH marks. I ran something similar recently where a couple of students rustled up some hilarious role play to make a <b>technical </b> point- the afternoon had been <i>VERY DULL INDEED </i> prior to this &amp; their K.I.S.S. entertainment perhaps gained a disproportionally high mark as a result!

Edited by - Stan. swan on 1/16/2006 11:59:17 PM


Senior Member
The ultimate obvious hardware is a big DPDT knife switch cross wired to reverse the +/- .

And you can chop carrots or fingers with it as well! :) cheers...


New Member
I'm not a professional educator, but have been through ALOT of school, and even tought some classes in college...My experience is this:

Chips are great, and using them, all the better. However, that bit where you UNDERSTAND what the magic black box is doing, is your key element in troubleshooting what's going wrong.

I mean it like this: The L293(D or otherwise) in't doing what you want? Why? By understanding how a discrete H bridge works, youll be able to look at what it IS doing, compared ot what it's supposed to do, and have a much better foothold on the problem now becuase you know WHY it's doing what it's doing (however vaguely, still helps!).

On the other hand, I'll never forget my first day in college calculus. We spent the whole class doing one problem (dont remember which) on paper. And yes...it could be done, and most of us got the right (ish) answers, and then the professor had us tape that to the front of our notebooks, pull out our graphing calculators, and never do another problem on paper to begin with. His point o the whole thing? Why do all the work and spend the time and money, when you can get it done in a minute or two with the proper tool.

You could argue that the proper tool, here, would be an integrated H bridge, depending on circumstances.

so it could og both ways, with benefits and detractions for each position. A war not worth winning, becuase it really cant be. Kinda along the lines of the intelligent design mess we've been going through here in the states. It's all a matter of the right tool for the job, personal preference, and a myriad of other thigns that boil down to a stalemate, no matter how you look at it.

Just my $0.02

--Andy P