1
\$\begingroup\$

In this configuration

schematic

simulate this circuit – Schematic created using CircuitLab

if the two pMOS are equal, they provide the same drain current to both the branches: suppose that Vbias is chosen so that the drain current will have a certain (constant) value Ibias. Moreover, if a new pMOS is added and connected at the same Vdd and with the same gate potential Vbias, it will generate the same drain current as M1 and M2.

But often the current mirror configuration is used:

schematic

simulate this circuit

where the drain current of M1 is Ibias (and it is the drain current of any other transistor equal to M1 and connected in the same way).

I think that the two configurations (with appropriate Ibias and Vbias values) are equivalent.

My questions are:

  • is the second configuration the most common in CMOS circuits? I found it more often than the former;
  • if yes or if not, why?
\$\endgroup\$
3
  • 1
    \$\begingroup\$ In your first circuit, how do you know what output current you'll get for a given input voltage? Consider that if you build this circuit on several chips, the PMOS behavior will not be identical from chip to chip. \$\endgroup\$
    – The Photon
    Commented Feb 27, 2015 at 18:12
  • \$\begingroup\$ I choose Vbias according to the transistor characteristics. Yes, with different transistor I will obtain different currents. \$\endgroup\$
    – BowPark
    Commented Feb 27, 2015 at 18:14
  • 2
    \$\begingroup\$ Even if you manufacture multiple copies of the exact same chip design, the MOSFET performance will vary from chip to chip and wafer to wafer. \$\endgroup\$
    – The Photon
    Commented Feb 27, 2015 at 18:16

2 Answers 2

3
\$\begingroup\$

In your first circuit, it's uncertain what output current you'll get for a given input voltage. First, because if you build this circuit on several chips, the PMOS behavior will not be identical from chip to chip. Also because as the temperature chagnes the PMOS device behavior will change.

In the second circuit, you know the output current will be essentially identical to the input current, regardless of variation of the MOSFET performance. The trick then is just to make a current source for the input that is insensitive to process and temperature variations.

\$\endgroup\$
1
\$\begingroup\$

What is "preferred" makes no sense since the two circuits are different.

The first, assuming well-match transistors, is merely a way to control two different current sources to the same value.

The second is a current mirror as you say. The purpose of that is to have the current source produce the same current ("mirror") as the current drawn by a load. You are calling this load Ibias in your second schematic. Only one current source is being controlled, and it is controlled implicitly by showing it the current you want it to source. This happens to be accomplished by using the first circuit with feedback around the control signal so that the first source matches the load current, then the second source follows. The control signal itself, which is the input to the first circuit, is kept internal to the block.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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