I understand how an H Bridge works, but the common ones use pnp transistors on top and npn on bottom. In L298N's datasheet it is attached this circuit and I just can't figure out how the Vs is going to supply the motor if there is a npn on top. I know that the emitter voltage of the transistor on top is depending on base voltage as follows: Ve = Vb - Vbe, so where is the Vs? Also, shouldn't the base voltage (MCU output voltage) be greater than the collector voltage (Vs) in such a circuit?
2 Answers
The high side NPN is used as an emitter follower.
Yes, it cannot provide up to VCC, but about a Vbe drop less.
The MCU interface does not affect the output, because there is an internal voltage translation to VCC, so that is why the motor output is VCC minus the Vbe drop.
-
\$\begingroup\$ I didn't known about this internal voltage translation because there is no wiring in the circuit diagram to suggest it. Thanks! \$\endgroup\$– AnnaJul 2, 2021 at 5:47
Have a look in the datasheet of the L298, at the bottom of page 3. There the source saturation Voltage is mentioned. This is the voltage that is dropped inside the L298 when an output is high and a current is flowing:
Ideally we don't want any voltage to be dropped, we want 0 V or more realisticly: as little as possible.
You're correct that because it is using an NPN there, you would need a higher voltage than +Vs to be able to make those NPNs lift the output voltage up to +Vs. But the L298 isn't designed that way so it simply cannot pull its output all the way up to +Vs. In the table we can see that typically at 1 A flowing, the L298's outputs drop 1.35 V.
Also note that things aren't much better on the "low" side either, the Sink saturation voltage is 1.2 V at 1 A.
If you want lower voltage drops than this, I suggest using a more modern IC. This L298 is ancient, I am surprised that ST still sells them!
-
\$\begingroup\$ I bet ST has at one point made a couple dozen million of these and now is just selling off stock. Probably, the mask is used to produce low-quality wafers when they need to run a machine "clean" without wasting time on a high-precision process... \$\endgroup\$ Jul 1, 2021 at 20:20
-
\$\begingroup\$ (note that I'm just as surprised. This thing is really ancient, and from the top of my head I can't think of a single professional use of it, unlike for ancient-and-bad-but-cheap opamps; you always get a better product if you replace it with something more modern, and the modern alternatives aren't inherently more expensive to produce) \$\endgroup\$ Jul 1, 2021 at 20:32