High voltage drop across H-bridge

[shehry]

Hi,

I am having a problem in driving a 12V dc motor using an H-bridge. The H-bridge uses two tip122 (npn) and two tip127 (npn).

I was using a PIC877 to generate pwm output which i gave as an input to the H-bridge via a voltage buffer (74LS244).

The problem is that there seems to be a very large drop across the transistors and very little voltage appears at the output. I was using a 20V supply and only 2 or 3 volts appear at the motor.

Now when i replaced the pwm output with a Vcc, the motor runs. This probably means that the buffer chip experiences a sink. I am not sure. But I had better add that the pwm output with 100% duty cycle does not run the motor while the Vcc does.

Someone told me that there might be two solutions to it.
1. to use a free-wheeling diode.
2. to use a FET to drive the BJT transistors used in the H-bridge.

I am not quite sure on how to proceed, so I would appreciate any help that I could get. I would like to know whether the solutions presented above are valid or not. And if they are, then how should I use them.

 

[Nigel Goodwin]

If you haven't got protection diodes on the H-bridge, then you've probably already killed the transistors?.

Darlington's aren't a very good choice for an h-bridge, as they have a fairly high voltage loss across then - probably 2-3V each or so, but nothing like you're getting. Try posting you circuit so we can see how you're doing it.

 

[shehry]

If you haven't got protection diodes on the H-bridge, then you've probably already killed the transistors?.

Darlington's aren't a very good choice for an h-bridge, as they have a fairly high voltage loss across then - probably 2-3V each or so, but nothing like you're getting. Try posting you circuit so we can see how you're doing it.

The circuit is fairly simple and has no protection diodes attached to it. I am attaching the diagram as well. I saw some of the H-bridges posted on the web with protection diodes. Could you tell me why I need them and which ones would be suitable for my case.

Also, could you tell me how the FET driving the BJT solution works?

 

[hyedenny]

Hi
Whenever you have a transistor driving an inductive load, like a motor or a relay, you have to use diodes to protect against "back EMF." This is essentially a reverse voltage - produced when the electromagnetic field in the coil collapses when the power is removed. For your project 1N4001 diodes should be fine.

Your power transistors, if theyre still good, are probably not turning on all the way. Try driving the power transistors with smaller ones like a 2n3904 tied high.

If youre not using PWM, you CAN use FETs, but they have to be turned on hard.

 

[shehry]

If youre not using PWM, you CAN use FETs, but they have to be turned on hard.

Why not? In fact, I had thought that FETs are actually better for switching while BJTs are better for amplification

Btw, can someone tell me how I can isolate my voltage buffer from the H-bridge. I was thinking of using a phototransistor or an IR sensor of some sort but it would put a restriction on the usable frequency

 

[bananasiong]

hi,

i'm not using h-bridge for my project, but my friend do. u can try to change the tip 127 to pnp transistor. i mean, 2 pnp on top and 2 npn at the bottom. it may work.
i will ask my friend for the detail of the circuit if u want, okay?

 

[hyedenny]

FETS, being high impedance devices, can generate a lot of heat if switched too often. At high voltages or fast switching times, the Miller effect can cause the FET to turn on when you dont want it to, effectively creating a shoot-through - a short in your circuit. Shoot-through can also occur if the turn-on times of the 2 opposing FETs overlap because of logic or switching flaws. Slow switching times and incomplete turn-on will also overheat a FET. Voltage spikes, back-EMF, too low gate voltage, too high gate voltage, spurious oscillations, static, etc. can all damage a FET very quickly, and require greater design consideration to be taken.
None of that stuff usually matters in everyday hobby stuff though! I use FETs in my H-Bridge with NO problems at all...

Consider a monolithic H-Bridge and H-bridge driver. There are some available in 3amp versions - maybe higher. I think you can piggyback them for more current also.

 

[shehry]

hi,

i'm not using h-bridge for my project, but my friend do. u can try to change the tip 127 to pnp transistor. i mean, 2 pnp on top and 2 npn at the bottom. it may work.
i will ask my friend for the detail of the circuit if u want, okay?

i wouldn't mind seeing the circuit, but the thing is that TIP127 is a pnp and the top two transistors are both tip127s while the bottom two are TIP122s. In any event I am going to try putting in some protective diodes

 

[Nigel Goodwin]

hi,

i'm not using h-bridge for my project, but my friend do. u can try to change the tip 127 to pnp transistor. i mean, 2 pnp on top and 2 npn at the bottom. it may work.
i will ask my friend for the detail of the circuit if u want, okay?

i wouldn't mind seeing the circuit, but the thing is that TIP127 is a pnp and the top two transistors are both tip127s while the bottom two are TIP122s. In any event I am going to try putting in some protective diodes

Bear in mind you've probably already destroyed the transistors!.

 

[bananasiong]

this circuit may work well.

i think it is possible if u don't want to use the opto-coupler. tell me if it works without using the opto-coupler okay?

 

[shehry]

Bear in mind you've probably already destroyed the transistors!.

i rechecked the TIP122 data sheet. it has a built in protective diode. So i dont think i really need to connect them separately.