NpN transistor in hbridge problems

[Jerran]

Im having troubles getting the transistors in my H-bridge from turning into miniture heating element To test my values I tried wiring a single transistor to the motor.

How should I wire the transistor properly and what value of resistance should I use on the base?

I've tried wiring the transistor on the high and low side of the motor, both result in a really hot transistor. Ive also used values from 10K to 10 on the base and cant find a combo that will provide current to the motor and keep the transistor from becoming a mini-sun on my proto board.

The motor is rated at 5V 1 A, voltage source is 6V.

 

[Jerran]

With a HFE of 220 I would need 4 mA into the base for a emitter to collector current of 1 amp correct?

 

[John Sorensen]

4mA * 220 = 880 mA < 1A. What kind of transistors are you using in the high side, NPN also? NPNs don't work in the high side of a bridge without special considerations (which I won't get into, because you probably are using PNPs). I would guess your problem is probably "shoot through," the high side and low side transistors in the same leg on at the same time. You need some dead time between turning one off and the other on to allow the transistor to turn off. This is particularly important when driving an inductive load. Do you have "snubbers" and or reverse diodes across the transistors as well, otherwise they may be failing from overvoltage. Do you have a scope? If you do, you can monitor the current through the transistors (on the low side, if you don't have isolated channels) by putting a small (1/4 ohmish) resistor in series with the emitter and monitoring the voltage across it. If you have shoot through, you will see very short spikes of extremely high current, which won't last long because the transistor will probably fail.

j.

 

[John Sorensen]

BTW, hFE is a poor design quantity.

j.

 

[Jerran]

I got the hFe of the transistor from a component tester, and yes the transistors have diodes to stear voltages spikes to +v and ground. The high side is PnP transistors. Sadly my scope isnt here yet, should have it next weekend I hope.

Im using a relay controlled by a pic to switch direction with a .1 second delay between switching.

I took the original schematic from this site **broken link removed**

 

[zevon8]

Hi

To add to the previous posts, I would place a snubber across the motor at least, something like a 2 watt 100 Ohm resistor in series with a 0.1 uF capacitor. This will help tame the back EMF from the motor. Also, you really need fairly fast diodes on the transistors, general rectifiers don't always cut it. You may want to consider FET's too, lower Rds ( on resistance ) built in reverse diodes, easier drive, you can parallel up more for bigger loads, etc.

 

[John Sorensen]

Make sure if you go with FETs with a back diode that the diode is intended for the purpose. Sometimes the back diode is a parasitic side-effect and not suitable for this kind of free-wheeling. The free wheeling diodes should point opposite the transitors (from emitter to collector on the low side and from collector to emitter on the high side), unless I misunderstand what you said, "stear voltages spikes to +v and ground."

Relay?! PIC?! 0.1S? All those parameters sound incredible! Especially the PIC! Just kidding (Atmel fan here). What are you doing with 1) Is that an electromechanical relay? 2) 0.1S is slooooow.

You shouldn't design using hFE (even if you've measured it). A good design should assume worst case of hFE, and be largely independent of it (that goes for a linear design, and a switching design like this). You should use enough base current make sure your transistors are solidly on.

Now, I'm not an analog guy; I'm all digital all day. Nigel can correct me if I'm in left field on this :wink:

j.

 

[Jerran]

Thanks for the replies.

Yes the relay is a electromechanical relay, I purposely made the switching delay exagurated to be sure that both legs weren't on at the same time.

I have considered using FETs, one thing Ive noticed on a few schematics I have in my circuit libraries is that when FETs are used to drive motors the are running 2 diodes parrellel to the FET, one switching and one zener . Wired with the cathodes together, the anode of the switching on the V+ side and anode of the zener on the ground side of the mosfet. What exactly is the purpose of the two diodes? It seems to me that the zener would be suffecient to prevent voltage spikes.

 

[Jerran]

Found my problem, Ive been looking at the wrong data sheet. My NpNs can only handle 800 mA. Seems to work perfectly with the proper transistors.