Accurately simulate an LED
A transistor that simulates an LED lets you test the driver circuit.
Jon Roman and Donald Schelle, National Semiconductor Corp, Santa Clara, CA; Edited by Martin Rowe and Fran Granville -- EDN, October 7, 2010
| View as PDF |
| Read More Design Ideas |

Figure 1 illustrates a typical LED-driver
circuit using a low-cost simulated-LED circuit. The simulated
LED accurately mimics a
real LED at a user-programmable
threshold voltage. A simple Darlington current sink, Q1,
provides a wide range of LED
threshold voltages. The size of
the heat sink attached to Q1 and
the power capability of Q1 are
the only limits on the amount
of power the simulated LED can
dissipate.You can easily tune the circuit for any LED voltage. Place a constant voltage across the simulated LED. Tune the circuit by adjusting resistor R1 until the circuit draws the desired current. You can adjust the shape of the voltage knee by making small changes to resistor R3, although this step is not usually necessary.
Figure 2 compares the simulated LED’s current and voltage characteristics to those of a real LED and a constant resistance. The soft turn-on of the simulated LED accurately mimics that of a real LED. Furthermore, the simulated LED quickly retunes to test minimum and maximum LED characteristics, thus giving you confidence that the circuit will work over all load conditions.
Talkback
-
one question for clarification....if the input voltage to the simulated led circuit is "constant" as stated in the article, how to you evaluate the estimated forward voltage drop of the led at various currents? By adjusting the pot, you naturally control the output current, but the voltage drop across the darlington (VCE) remains relatively constant, since Vin is constant. So, for example, how did the authors arrive at their range of voltage readings in the graph (i.e. 30-35 V approx.)? At what location in the circuit were those voltage values measured? Thank you
Rick Brower - 2012-10-1 12:17:31 PST -
The red curve in Fig.2 is rugged, as compared to yellow one, belonging to real LED string.
This might be a sign of oscillating of the Q1, due to parasitic capacitances and inductances of (longer?)leads.
It would be interesting to connect a ceramic capacitor of 1 to 10 nF between base and collector of Q1 to see, if the I-V characteristic (red) becomes smoother.
Further, the current gain of Q1 is about 700 at Ic=0.2A; so the base current is roughly IB=0.3 mA and this IB influences the voltage at the wiper of the R1.
Perhaps the values of both R1, R2 could be reduced to 1/3, for example, to suppress the virtue of the IB.
Marian Stofka - 2010-7-10 12:16:03 PDT


















