3
\$\begingroup\$

Sorry if the question seems stupid, im not much into electornics just trying to get into it... I'm trying to figure out the purpose of current mirror, I read everywhere that it copies reference current to another node but why we're doing it? why can't we just use that current source without connecting them to N-channel transistors? also why 2 transistors are used?enter image description here

\$\endgroup\$
  • \$\begingroup\$ Your title asks "how does it work", but your text asks "why do we use it instead of something else". Can you edit one or the other to make it clear what kind of answer you want? \$\endgroup\$ – The Photon Sep 11 '16 at 15:12
2
\$\begingroup\$

Current mirrors take a positive current (source) and generate a negative current (sink).

Current mirrors have low input impedance, so the 'reference' current can be from a poor current source (output impedance 1k ohms) and give rise to a mirrored current that is a better current source (output impedance 100k ohms). They isolate the input from variations in the output, so you can wire together a dozen current mirror outputs to make a summing junction, without causing any signal flowing back into the input nodes.

The illustrated current mirror makes an excellent pull-down load for an NPN emitter follower, sinking equal current at a wide range of output voltages (unlike a resistor pulldown).

The use of two transistors is not required, of course; you could use dozens, and the single reference input can drive two parallel output transistors to make a gain-of-2 mirror, or four to make a gain-of-four mirror (there are D/A converters built this way). Integrated circuit operational amplifiers often use a single current reference and current mirror to generate many bias currents. Matched transistors are very easy to integrate, and are more accurate than resistors in most IC processes.

Current mirrors with sense-resistor inputs are used as current monitors high side current sensor as well, with highly attenuated output current. The low input impedance, like an ammeter, is a major benefit.

Current mirrors have a wide range of uses.

| improve this answer | |
\$\endgroup\$
1
\$\begingroup\$

There are many answers to your questions. Regarding why don't we use the current source directly, one of the reasons is that the 'original' current source may not be strong enough to drive the load.

Another reason to use current mirrors is to stabilize the current against changes on temperature.

| improve this answer | |
\$\endgroup\$
1
\$\begingroup\$

Other possible reasons:

  1. You only have one source but your circuit needs many copies of the same amount of current
  2. You need several copies of the same current to match. MOSFETs from the same die built with proper layout techniques can match very well. In many cases, better than your ability to find multiple sources that match each other.
  3. You want the currents to "track" each other. If the master current moves up or down, your circuit will be more robust if all the current "copies" follow the same trend
  4. Your current source has a compliance limit that is not useful in your target circuit. You may have to look this term up for more details
| improve this answer | |
\$\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.