Maker Pro
Maker Pro

A PNP-NPN output with common-base level shifting

L

Lauri Alanko

Jan 1, 1970
0
Hello.

Here is a circuit I came up with. It is meant to be a general-purpose
digital output stage: it takes in an inverted digital control signal
(typically low voltage and low current), and uses that to switch between
sourcing from a higher voltage and sinking to ground.

Out
Q1 | Q2
V+---v_/-------------+-----------\_^---Gnd
|Q3 D1 R1 R2 D2 |
\_^--\<|--\/\--+--\/\--\<|--+
| |
Vhi ~In

Q1 and Q2 do the switching between sourcing and sinking. I'm not sure
what the proper name for this configuration is. "Push-pull" seems to be
when NPN is on high side and PNP on low side, and "totem pole" seems to
refer to either that or to two NPNs on top of each other. But I haven't
seen this PNP-NPN configuration very much.

To me it seems it's much better to have PNP high and NPN low, since the
voltage drop is then much lower. Is this PNP-NPN switch in common use? If
so, what is it called? If not, what is the problem with it? The fact
that the transistors get saturated?

Q3 is an NPN in common-base configuration used to drive Q1 with a lower
voltage. Vhi is the high level of the input signal, so when input is
high, Q3 is at cutoff and so is Q1. When input is low, Q3 is active and
drives Q1 to saturation.

Do I need a pull-up resistor from the base of Q1 to V+? If so, why?
Is there some danger in leaving the base of Q1 floating when Q3 is at
cutoff?

The zener diodes D1 and D2 are there to prevent strike-through when
transitioning between high and low. They should drop about Vhi/2 to
ensure that one transistor won't activate before the other one has gone
into cutoff.

To me this seems like the simplest way of, say, driving a mosfet with a
microcontroller (using only cheap discrete parts). But I haven't seen
this particular circuit anywhere. Is it okay or is there a problem I have
overlooked?

Thanks in advance.


Lauri
 
Top