Gamma Correction Circuit Question

[Lawnboy]

quick question on gamma correction circuits: in Peter Smith's led drivers, does the current drawn by the leds determine the gamma curve? for example, if the gamma resistors are calculated for 100ma, and the led array pulls 150ma (or 75ma), will that affect the gamma curve? could someone explain to me how this method of introducing non-linearity is accomplished?

[Klaas Robers]

In principle the function is not dependant of the drawn current. Mathematically Gamma = 2 will say: the LED-current is the video input voltage to the power of 2, where black = an input of zero.

So as long as black = 0 this is always Ok.

However, in practice we don't make a circuit that fulfills the square fuction really. It will be an approximation in a certain range of input signals, e.g. zero volt (black) to one volt (white). In a ready designed circuit the output current for white then might be 100 mA. If you use that circuit onto 200 mA you will see that above 100 mA the circuit will behave linearly, not as a square function.

The circuit with the two diodes in the emitter of the video transistor simulate the square fuction in the form of three straight lines. Compare the chapter on the video amplifier - LED/neon driver on page 22 of the NBTV handbook. For small signals the circuit is more or less linear, for larger currents than for white (40 mA in this circuit) there are no further bends in the line. This approximation of three straight lines in stead of a parabola is good enough.