NPN and PNP transistors

Why do we have two transistor types while a single can do everything

Not true.

Perhaps read what you actually question actually asks, clearly a single NPN cannot do the same job as a single PNP.

Find a decent book on basic semiconductors and give it a read.

"everything" is a bit generous in this context.
Do the following searches "transistor high side switch" and "transistor low side switch" to find examples of where these two types are not interchangeable.

Well, there was no equivalent of pnp/npn in vacuum tubes, so I can see what the OP is getting at. If we only had npn transistors, we would find a way of delivering the functionality we want.

The reason we have both is that it basically "fell out" of the technology - we have both because semiconductor technology makes both types possible. The reason we use both is because it greatly simplifies circuit design, reducing costs by providing functionality that would be much more difficult to deliver if we only had one or the other.

1 Like

Have you tried designing a discrete operational amplifier using only NPN? Yes it can be done but its harder to do, generally performs worse that if you can use symmetrical circuit topologies in judicious places.

You could ask why do we have power transistors when you can put small signal transistors in parallel for higher currents - because its less convenient, someone else is going to figure out that selling power transistors is profitable in that hypothetical situation.

Why have you been asked the question ?

BTW PNP transistors can have lower noise than NPN, and NPN are faster than PNP, so if you want ultimate performance you pick the polarity as appropriate. People pay big money for ultimate performance.

1 Like

It is interesting to note that the original "7400" logic family ICs used only NPN transistors.

CMOS is vastly better in many respects, but the 7400 series was the basis of serious computers, along with NMOS LSI.

1 Like

Because we can, and using both makes some circuit designs simpler.

I can see that's an NPN but with two emitters? I've seen that before but didn't know what it was beyond a transistor, sometimes multiple bases and/or collectors too. What is that kind of transistor is that, so I can research it a bit more?

I fail to see the purpose of the question.
Do you have a question regarding how to do something or are you just questioning
electronic circuit design practices while not knowing the difference between an NPN
and a PNP. Getting a little ahead of yourself aren't you ?

non sequitur....
Isn't that like saying if we only had horses we would still be able to travel. We wouldn't need cellphones if we had landlines, or pony express.

Transistor technology is able to make NPN and PNP versions. Circuit designers find the PNP versions very useful. There are many applications where a PNP transistor will work where an NPN will not without a lot of support circuitry.

The OP's question was "Why do we have two transistor types while a single can do everything".

I wanted to begin by confirming the OP's assertion that "a single [type] can do everything".

I then pointed out that using both types greatly simplifies designs and reduces costs. A point which, for some reason, you felt compelled to repeat. :roll_eyes:

Its something you do on an integrated surface to save space - its two transistors in effect, but its made in the same well. Integrated circuit design allows lots of tricks like this. The emitters can be different sizes, allowing clever effects like bandgap references.

For TTL the multi-emitter device gives AND gate functionality since only if both inputs are HIGH can the multi-emitter transistor turn off.

1 Like


Tom.. :smiley: :+1: :coffee: :australia:

1 Like

Actually, in the above diagram, Q1 appears to be just a diode array. The only time it might function as a transistor is to pull accumulated base charge out of Q2 to turn it off faster.

What does Mark think?

Q1 can pull Base of Q2 to min(A,B). When you replace it by two diodes they will pull to min (A,B) + 0.6 V, decreasing noise margin.

Three diodes, not two. :roll_eyes:

Re reply #17: I don't see a voltage source for Q1 collector. Should there be a resistor between Vcc and Q1 collector?