• This is experiment # 17 in the lab manual. A brief introduction of Wilson and Widlar current sources and the steps to carry out the lab are given here. To get a good insight of Current mirrors (current sources) read the manual.
• Current mirrors are used as current sources for biasing circuits on ICs because they require a few elements and occupy very small space thus they can be easily fabricated on ICs to supply small biasing currents to circuits.
• Circuit # 1 is Wilson current source. In the current mirror both of the transistors are matched (we can add more branches to the right of the circuit like the one which is on the right to get more stages of current mirror) so they have same value of 'B' so that their collector current are equal for same value of VBE. The resistor R1 is the circuit we want to give biasing current or consider it as a load.

• If we ignore small base currents, the current Iref would be equivalent to

• The relation between I1 and Iref is obtained by applying KVL in the lower loop (neglecting base currents we have IEA = Iref and IEB = I1)

           Now using diode current expression ID = IS*e^(VD/nVT) (actually ID = IS*(e^(VD/nVT) -1), but since in forward  region the
          exponential term is quite big so we can omit 1) to make a relation between forward base emitter junction voltage of transistors (VBE)
          and current passing through it, we can write the above equation (ii) in the following form (using n = 1 and VT = 26mV @ room temp)
 

• This last relationship defines a Wilson current source. If  Iref and I1 are not too much different then their ratio would be equivalent to 1 and above relation can be approximated as

• If we remove RA from the circuit, we are left with a Widlar Source which is used to provide small amount of bias current to the circuits. The expression for Widlar source will be

• Design circuit # 1. Choose RA = 2K, RB as a variable resistor ranging upto 10K.Choose maximum value of RB for this part. Choose R1 about 2 to 3K. Now select a value of Rref to get Iref = 1 mA.

• Now select two values of RB for which I1 is less than Iref and two values of RB for which I1 is greater than Iref. Also make sure that your transistors would not go into saturation (VCE should be quite higher than 0.3V).In each case measure I1, Iref, RB, and Vce of both transistors to make sure that they are not going into saturation. 
• Make circuit # 1 a Widlar source by removing RA. Now choose two values of RB for which Iref is greater than I1. Measure I1, Iref, RB, and Vce of both transistors for these two values.
• Compare the measured values of I1 for both of the sources with the values calculated from equation (iii) (for Wilson) and equation (v) (for Widlar). Over what range of I1 can you use the approximation in equation (iv) without appreciable error?