H-Bridge Circuit

[catman232]

Hey guys, I recently made this H-Bridge circuit in a simulator, and want to make sure there aren't any blatant issues before I invest in the parts. I had never seen an H-Bridge circuit until after I made this, so it deviates from most of the typical designs I have seen.



Basically, setting "In" to high will set Out1 high and Out2 low, and the opposite when it is set to low. Setting Enable to high will set both Out1 and Out2 to high and disable the motor. This design is quite different from the others I have seen, mostly by the fact it only has two inputs.

Again, it'd be awesome if you guys could point out any newbie mistakes I made, as this is one of my first circuits!

[Zero999]

What's it for?

That design is only suitable for very low currents. You couldn't run a motor off it.

[Kremmen]

Hmm. Different is right, that's for sure.
Technically it resembles a H-bridge but your point in this particular circuit escapes at least me. Since you have a relay with mechanical movement, the bridge is not suitable for pwm control. That leaves simple direction + enable. But for that much more efficient circuits exist. One such would consist of a DPDT relay where the common pins are connected to the motor and the NO/NC contacts to V+ and gnd in antiparallel. That way the motor changes direction as the relay switches. If you need enable, a simple power fet between the relay and gnd takes care of that. At the same time it makes pwm speed control possible. How about that?

[PA0PBZ]

Also, the only function of R1 and R2 seems to be to drain the battery?

[madires]

Nice! A H bridge which drains the battery in about 2h without running any load. 

[Zero999]

It looks like he's been playing around with a simulator without fully understanding it.

Why not just use an H-bridge IC such as the L293NE or SN754410?

[vpjaseem]

This circuit is 100% perfect and working.
I simulated it with Proteus 8

Here I'm attaching the simulation file (as .gif animation)

[PA0PBZ]

This circuit is 100% perfect and working.

Working.... yes. Perfect... far from that.

[Kremmen]

This circuit is 100% perfect and working.
I simulated it with Proteus 8

Here I'm attaching the simulation file (as .gif animation)

You'll live & learn. Now would perhaps be a good time to learn that there is one small but vital difference in the world. That is, that in theory, theory and practice are the same thing, but in practice that is not the case.
Simulators are just simulators with all kinds of shortcomings. They never reflect the real world with perfect accuracy, so please disabuse yourself quickly from using the words 'simulation' and 'perfect' in the same sentence.

The original circuit works - kinda. But it is not one a skilled designer would come up with. For the reasons stated and others, not mentioned yet.

[vpjaseem]

At least we can expect the circuit will work in practice,

[Zero999]

This circuit is 100% perfect and working.
I simulated it with Proteus 8

Here I'm attaching the simulation file (as .gif animation)

I rest my case.

[ignator]

At least we can expect the circuit will work in practice,

This is not a H bridge. You can delete R3, R4, D1, D2, and short out D3, D4 (from your second drawing, nice animation). And your wasting power through R1, R2.  These 2 resistors require being replaced with switches for it to be a H Bridge, as Kremmen has indicated. H bridges have 4 switches. I've seen RF switches like you show, but they work by applying DC bias to the diodes.

[Dave]

This circuit is 100% perfect and working.

Self-confidence is good, the excess of it is not. The sooner you realize this, the better.

[RoadRunner]

this circuit is no good. why don't you use  L293 or if you wanted to build it with jelly bean parts make simple transistor based h-bridge .

[KTP]

Actually a single switch or relay works but it has to be a double pole double throw.  Contacts won't last long though unless you add some components.

[catman232]

Haha,  thanks for the feedback guys. By perfect I meant the simulation gave me the numbers I wanted,  but now I see the faults in this.  Perfect wasn't really the right word to use,  sorry. On another note,  I don't want to use an ic because I want to learn how to design it myself.  Again,  thanks for the help! 

[KTP]

It is fine to learn and I think educational to use discretes but you should probably start with transistors instead of diodes :-)

4 bjt or mosfets and a sprinkling of other parts and you can have a functional h-bridge, then start worrying about things like shoot-through and turn on/off times due to gate capacitance.

[catman232]

It is fine to learn and I think educational to use discretes but you should probably start with transistors instead of diodes :-)

4 bjt or mosfets and a sprinkling of other parts and you can have a functional h-bridge, then start worrying about things like shoot-through and turn on/off times due to gate capacitance.

Sounds great. I am going to take another crack at designing this, and I'll post here when I finish. I also might add that I am 14 and not exactly swimming in cash, so that's why I was relying on a simulator. Hopefully soon I'll be able to afford some more simple parts so I can start prototyping things in real life.

If you can believe it, I started learning how to program when I was 8, and only recently came across Electronics Engineering. Still unsure as to which one I should pursue in my career, but I've got plenty of time to decide.

[ben_r_]

Oh, man... Do yourself a favor and just use an IC. Soo much easier to deal with. Unless youre just doing this to learn and have fun, in which case have at it! And google some H bridge circuits.

[Stonent]

I think we're missing the point. He had a function that he needed that he came up with his own circuit to complete the function which is pretty cool.

[KTP]

I think we're missing the point. He had a function that he needed that he came up with his own circuit to complete the function which is pretty cool.

Well, sort of.  I mean it was an obfuscated and inefficient way of getting a desired effect.  It is ok to reinvent the wheel and can be a great learning experience, but there are reasons you would want to start with a round shape instead of square.  In other words, go for building an hbridge out of discrete components to learn how they work, but use transistors.

[ben_r_]

I think we're missing the point. He had a function that he needed that he came up with his own circuit to complete the function which is pretty cool.

Well, sort of.  I mean it was an obfuscated and inefficient way of getting a desired effect.  It is ok to reinvent the wheel and can be a great learning experience, but there are reasons you would want to start with a round shape instead of square.  In other words, go for building an hbridge out of discrete components to learn how they work, but use transistors.

Exactly. If the OP is doing it just to play and/or play, fine. But if he's doing it as part of a larger project why waste the time and effort and added cost? Just find a cheaper, easier and quicker to implement IC and move on to the more important things.

[jeffro]

I am a newbie ,compared to some of the more experienced guys here , I would like to say good effort , even more so that yr only 14.
My suggestion doesnt relate to the design ,thats all been covered , more about suppression of back EMF from any relay .Without a freewheel diode , a relay can induce some nasty spikes . Anyway once again good effort ,only way to learn.