Transcranial Direct Current Stimulation (tDCS) is a method of external neural modulation that uses a small current run through the brain in order to alter cortical excitability. The details of the mechanism of action and exact enhancements possible are beyond the scope of this article, but start with the wikipedia entry, examine commercially available products, and look at safety data and ethical reviews before deciding if this is something you would like to pursue. Some google scholar searches will turn up interesting things too.
The photo on this page is from this article.
Step 1: Circuit Principle of Operation
The circuit shown is a regulated current sink. You may find it a useful building block in your future projects. It regulates the current through R[L], preventing it from exceeding a set value. This circuit doesn't have active drive capacity, though, and so V[DRIVE] must be large enough to drive the desired current through R[L].
The current through R[L] is equal to I[C]. I[C] is roughly equal to ( V[REF] - (V[BE] of T1) ) / R[LIM] .
To see where this equation originates, begin by noting that the sum of the voltages around the loop formed by V[REF], the base-emitter junction of T1, and R[LIM] must be zero (by Kirchhoff's voltage law):
V[REF] - V[BE] - V[RLIM] = 0
V[RLIM] = V[REF] - V[BE] .
The current through R[LIM] (also known as I[E]) is defined by Ohm's law, and we can substitute using the previous equation:
I[E] = V[RLIM] / R[LIM] = (V[REF] - V[BE] ) / R[LIM] .
Ignoring the base current,
I[C] = I[E] ,
so the current through the load resistor is approximately defined by
I[LOAD] = I[C] = (V[REF] - V[BE] ) / R[LIM] .
If you wish to include the effects of the base current of the transistor, you must also factor in the current gain of the transistor, h[FE].
Viewing the transistor as a node, by Kirchhoff's current law,
0 = I[C] + I[B] - I[E]
I[B] = I[E] - I[C] .
We know that h[FE] is the factor we can multiply by I[B] to find our I[C]. Thus,
I[B] * h[FE] = I[C] .
Substituting for I[B] from a previous equation,
(I[E] - I[C]) * h[FE] = I[C] .
Solving for I[C],
I[C] = I[E] - (I[E] /(1 + h[FE] ) ) ,
and since I[E] = (V[REF] - V[BE] ) / R[LIM] ,
the exact equation then becomes:
I[C] = ((V[REF] - V[BE] ) / R[LIM] ) - (((V[REF] - V[BE] ) / R[LIM] ) / (1 + h[FE] ) ) .