The common-base configuration is so little used, it doesn’t even register on many engineers’ radar. If that describes you, refresh yourself on the "basics."
The common-base configuration is so little used, it doesn’t even register on many engineers’ radar. So, as a self-professed analog guy, I’m embarrassed to say that, while it’s on my radar, I don’t recall ever designing a C-B circuit. As part of the EDN CDEL project, a C-B stage seemed the perfect choice to provide voltage gain for the MOSFET gate driver, but I required more thought than expected to design a circuit that might actually work, and I wasn’t happy till I’d actually confirmed operation on a breadboard, and worked through some very basic math.
I find I can’t really learn and internalize a circuit until I have a need for it, and have done at least some small hands-on activity with it. Unlike basic C-E & C-C arrangements that most of us can design in our sleep, it took me a few tries to draw a workable C-B stage. Roughly speaking, I needed to convert a 0-3 V swing to a 0-12 V one. One side of my brain knew that all the output current came from the input, yet the other side kept looking at the biased E-B junction and couldn’t get past the fact that it wouldn’t just turn on the transistor fully. Eventually though, the voltages and currents fell into place. True enough though, the C-B can still saturate like a C-E when driven hard enough.
Once I realized the base had to be biased for my application, not grounded, the circuit fell into place. That was after realizing I didn’t want a PNP … unless Ro were to be connected to a negative supply!
Although I prefer to understand things visually and intuitively, not just mathematically, I worked through some simple math to confirm my thinking:
Assume IC = IE (it doesn’t, but close enough with a good transistor)
IE = (VI - VE) / Ri (set VE to Vref, or Vref - 0.7V. Just don’t forget about your approximations.)
VO = 12V - IC·Ro
Since current through Ri & Ro is the same, gain = Ro/Ri
The output supply must generally be stable, as there’s no power supply rejection. A small bonus of this “coming down from 12V” behaviour for my MOSFET driver is that I can set VO to around 3V when VI is 0V, thus not wasting precious driving-DAC resolution in the MOSFET’s sub-threshold range.
I like how C-B is somewhat op-amp-like. Gain is Ro/Ri, and the emitter input behaves similarly to a summing node.
Here’s an exceedingly boring video demonstrating input vs. output, albeit on a classic 549 scope:
Of course, C-B can be used in ways beyond this basic voltage amp. It can make a digital level converter, and is regularly seen in RF and cascode circuits.
Have you designed or encountered interesting common-base circuits? If so, how, where, and why?
- Trits vs. Bits: 5-Trit ternary DAC has (almost) 8-bit resolution and high PSRR
- Off-line flyback uses HV transistor in common-base configuration
- LED-driver level-shifter includes fault detection
- Cascode MOSFET increases boost regulator's input- and output-voltage ranges
- Choosing Transistors
—Michael Dunn is Editor in Chief at EDN with several decades of electronic design experience in various areas.