Four transistor versus six transistor H bridge
Martin_H
Posts: 4,051
I've been reading the BEAM wiki and a few things confused me, so I figured that the forum could correct my understanding.
Here's the four transistor H bridge:
http://www.solarbotics.net/library/circuits/driver_4varHbridge.html
The bias resistors seem reasonable, and the basic structure is what I would expect. But it also includes a 74HC14 with a caveat "if you don't include inverters (or, at least buffers) on the control inputs, you now have to take great care to avoid having the bridge influence the circuit that's driving it." I guess I'm wondering what they mean by influence? If it is a microcontroller raising pins high and low, what could it do?
They also have a six transistor H bridge:
http://www.solarbotics.net/library/circuits/driver_tilden.html
The first thing I noticed is that the bias resistors are huge! The second thing I noticed is that I don't understand why there are six transistors?
Here's the four transistor H bridge:
http://www.solarbotics.net/library/circuits/driver_4varHbridge.html
The bias resistors seem reasonable, and the basic structure is what I would expect. But it also includes a 74HC14 with a caveat "if you don't include inverters (or, at least buffers) on the control inputs, you now have to take great care to avoid having the bridge influence the circuit that's driving it." I guess I'm wondering what they mean by influence? If it is a microcontroller raising pins high and low, what could it do?
They also have a six transistor H bridge:
http://www.solarbotics.net/library/circuits/driver_tilden.html
The first thing I noticed is that the bias resistors are huge! The second thing I noticed is that I don't understand why there are six transistors?
Comments
The 4-transistor H-bridge will work fine driven from a microprocesor pin as long as you don't let that pin float. (rather hard to do actually... so an inverter is a good idea.) You'll also need to insure that the microprocessor and H-bridge have similar voltage supplies. The 2K base resistors will keep the load on the microprocessor pins low. The comment on loading is most likely due to the low power finicky nature of BEAM circuits, it just doesn't take much current to disturb them.
Lawson
Fast-blow fuses are always a good idea. Stuff happens.
I also think diodes are important. I don't mind destroying transistors. I keep a good supply. The diodes protect my Propeller.
My kinda guy, I love a good scorched earth policy! Transistors can stunt double as fast-blow fuses. Use sockets for EZ replacement.
Yes, this is a common problem with higher power h-bridges. And in some cases, even using Darlingtons might require an additional pair of transistors to drive a powerful motor.
Consider that 742 watts equal one horse power and that W = Amps x Volts.
When you do start addiing Darlingtons the voltage drop in the H-bridge doubles, so a much higher supply voltage is required to get the desired voltage to the motor.
These internal voltage drops for a TIP120/TIP125 pair can add up to 4 Volts if driven to saturation. And the devices run quite hot. So MOSfets with very low Rds have been preferred as a cooler h-bridge with allowing one to use a lower voltage and not to waste so much power in heat.
I found that the Beam H-bridges were not that wonderful, though they are easy to build and do work. They can run quite hot and many of the reasons that they work well is that the motors they power are so small that ignoring good engineering can be gotten away with. They are mainly optimized to run tiny motors for short burst periods from solar power.
The TIPxx series are matching power transistors and the TIP1xx are matching Darlington power transistors with internal diodes.