4
\$\begingroup\$

Last night I was experimenting with and motors and transistors. I was testing out NPN and PNP transistors to learn more about how they work and how I can use microcontrollers to control the speed with PWN. I was having some issues, including the transistors getting very hot, but the most confusing part of the night was when I turned the transistor around and the circuit still seemed to function the same. I thought transistors allowed current flow in only one direction. Maybe I am wrong about this?

\$\endgroup\$
  • 2
    \$\begingroup\$ You need to post a "before" and "after" schematic so we can see how you connected them. \$\endgroup\$ – pipe Apr 21 '16 at 11:57
  • \$\begingroup\$ yeah.. what do you mean, "turned them around"? What is your circuit? \$\endgroup\$ – Roker Pivic Apr 21 '16 at 11:59
  • \$\begingroup\$ You mean you exchanged collector and emitter? This is not recommendable. Because of their respective functions, they have different doping levels. Emitters are not made to be used as collectors and v.v. \$\endgroup\$ – Bart Apr 21 '16 at 12:10
  • \$\begingroup\$ Yes, that's right. I reversed the collector and emitter. \$\endgroup\$ – Listen Apr 21 '16 at 13:06
12
\$\begingroup\$

A NPN transistor basically is a stack of three differently doped areas of a semiconductor. The first is N-doped, the middle P-doped and the last N-doped.

So yes, on the first sight, you can swap collector and emitter, and the transistor might still work. But there's more magic: A real transistor is optimized to fulfill its specs when connected correctly, i.e. the interior is not symmetric. If you swap C and E, you get a new transistor with completely different specs, which you don't know. It's also easy to overload and destroy the transistor by swapping C and E.

An anecdote: When I was about 10, I built my first circuits, and also started to etch PCBs. My first try was an astable multivibrator with two LEDs, and the PCB layout was absolutely mirror symmetric. The circuit worked, but the one LED was ten times longer on than the other. It turned out I had mirrored the second transistor with respect to the first, i.e. accidentally I swapped C and E. This caused the strange behavior, and after removing and soldering the second transistor, the circuit did what it should.

\$\endgroup\$
  • \$\begingroup\$ This was something that kept bothering me way back in the day at school. And yes, you get "different" specs, usually really bad ones by swapping the pins. The physical geometry of the pins is quite different, see here: fourier.eng.hmc.edu/e84/lectures/ch4/node3.html \$\endgroup\$ – Barleyman Apr 21 '16 at 12:22
3
\$\begingroup\$

As Bart suggested, I'm assuming by backwards, you mean the emitter and collector are switched. Yes current can flow in both directions.

An NPN transistor backwards is also an NPN. There will still be a reverse beta, however, the backwards NPN transistor won't work as well as a correctly oriented one will. It's not recommended.

You're thinking of diodes when talking about allowing current to go in one direction; however, at a certain reverse bias (negative breakdown voltage), the diode will let current through even though it is placed backwards in the circuit.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.