Why would I want to know what a current mirror is?

A current mirror at first glance appears to just be a circuit which copies a reference current - How is this useful practically speaking? Great question! I also pondered this when learning about them as they appeared rather theoretical in nature to me!

On inspection, the above doesn't look particularly useful, however, it turns out this little building block (we will look at the magic a little later) is a key element of many integrated circuits! The question is, where would simply copying a current be useful?

One of the main areas where copying a current is beneficial is in biasing multiple transistors which turns out to be required for almost all Integrated Circuits as they can contain billions of them! However, why couldn't we just simply use resistors to bias them? The reason is that resistors within IC's are very large and inaccurate, whereas transistors are orders of magnitude smaller and can be fabricated with high accuracy in silicon which is needed to match the transistor dimensions, a key requirement for current mirrors.

(Image credit: Intel)

The Intel Loihi 2 chip shown above contains 2.3 billion transistors!

Another area they are commonly found are in current sources, recall that an ideal current source has the following properties:

  • Has the same current flowing through it regardless of the voltage across its terminals
  • Can supply a specified current to any circuit element connected to it
  • It has the voltage necessary to provide the rated current

It turns out that current mirrors are practical implementations of current sources, they supply the copied current to an attached load and maintain the voltage necessary to provide the current with minimal dependency on the input voltage.

Let's take a look at the magic in closer detail:

The basic idea is that the resistor, R, programs a current Iref and if we assume both Q1 and Q2 have the same geometry we will have the same Vbe drop across Q1 and Q2 and it is this action which causes the collector current of Q2 to match Q1, i.e. mirror the current. Note that the above is the most basic topology which is not used as-is in practical circuits due to lack of negative feedback with emitter resistance to reduce temperature dependence on the beta values of the transistors for example, however, this working concept is a key feature of them all.

Current mirrors are also key building blocks of the analog power house that are operational amplifiers, or op-amps for short. Check out page 7 of the classic LM741 here and you will see transistor pairs Q5-Q6, Q8-Q9, Q10-Q11 and Q12-Q13 forming current mirror blocks!

See the Current Mirror section of the Analog Electronics course for the deep dive treatment of this key building block!