CCS Basics

CCS is an abbreviation for Constant Current Source.

CCSs are used in two places:

1: In signal/feedback paths where a DC offset is required to bias function-critical components - LTP, output of some class A amps...

2: In/as a reference by which a signal is compared and processed - Voltage regulators, voltage references...

3: To charge a battery.

These applications in particular will be discussed later on. At the moment we will discuss the concepts normally applied in constructing a CCS.

[NOTE: there should be a note in here about the simplest transistor CCS which is a resistor in parallel with the B-E junction with the output taken from the collector. It should state the disadvantages of this and namely mention temperature coefficients and why using a diode for his purpose is an intuitive way to compensate for this.

Diodes

Diodes in particular behave in a way that makes them suitable for use in a CCS.

Once you get past breakover voltage, diodes exhibit a fairly steady voltage of 0.6V across them. That is to say, a diode acts as a CVS, or Constant Voltage Source.

So, using this principle as inspiration a very crude and mostly impracticable (but conceptually correct) example is as follows:

[Diagram: diode in parallel with a 220r resistor, with resistor Rload in series with the diode and resistor]

If the current through the diode is neglected, we find that since the diode creates a voltage drop of about 0.6V across the 220R resistor, the resistor will draw about 3mA since 0.6/220=~.003

This forms the basis of diode-based CCSs.

Common CCS Circuits

[diagram: two diodes paralleling the B-E junction of a transistor, with a 220r resistor in series with the emitter. Output is taken from the collector]

The above example is one of the most common arrangements found. The first diode cancels the 0.6V voltage drop caused by the transistor's B-E junction (remember, a transistor is virtually two diodes facing or pointing away from each other), and the second causes a 0.6V Voltage drop across the 220R resistor.

There is one interesting advantage to using this method: Since diodes have nearly identical temperature coefficients to transistor B-E junctions, the first diode cancels out the tempco of the transistor (we still have the tempco of the second diode, however).

[diagram: a 6.2V zener diode replacing the two diodes in the above diagram]

This is nearly identical to the above example with the exception of the Zener diode replacing the two standard diodes. The tempco isn't so neatly compensated, but the concept is the same. Subtract 0.6V from 6.2V, which comes out to 5.6V, and then 5.6/220=~.0255. So this will draw about 25.5mA, a bit more than usual for small circuits, but a 1.8k resistor will give ~3mA, the same as in the above example.

[More discussion of various CCSs will follow later (hopefully not only contributed by me D:)]

---