Simple single-cell white-LED driver uses improvised transformer
Minimal component-count circuit boosts a single cell's voltage to light an LED.
Jim Grant, Scientific Controls, Orlando, FL; Edited by Brad Thompson and Fran Granville -- EDN, February 15, 2007
A white LED delivers a wide color spectrum and better visibility than do monochromatic LEDs. However, a white LED presents a higher forward-voltage drop than do its colorful counterparts and thus poses problems for operation from a single 1.5V cell. The self-oscillating step-up converter in Figure 1 features a minimal component count and an easily assembled transformer, T1.
During the time it takes to charge T1's primary inductance, resistor R1 and T1's added secondary winding provide sufficient base current to turn on Q2. Q2's collector current increases until its base current can no longer hold the transistor in saturation. When Q2 comes out of saturation, T1's magnetic flux and secondary-voltage polarity reverse. During T1's primary-discharge interval, the combination of T1's secondary voltage in series with Q1's base-emitter voltage applies reverse bias to Q2's base and turns off the transistor. When Q2 turns off, the voltage across T1's primary inductance adds to the battery voltage and applies a forward bias to the LED, D1. The current through R1 determines the power applied to the LED and applies forward bias to Q1's base-emitter junction to provide temperature-compensated bias voltage for Q2.
-
The first transistor, Q1 can be represented by a diode. However, the first transistor can be eliminated completely and no diode is needed. Substitute a 500 ohm for R1 and the circuit works perfectly.
Jeremy Jordan - 2012-5-2 15:28:43 PST -
Error?? PDF version shows the base and collector shorted together, making Q1 just a 'be' diode.
Paul Moffat - 2007-29-3 13:06:00 PDT -
I approached the design idea from another direction and got the figure. I still wish I didn't have to go looking for it.
Mark Dresser - 2007-24-3 18:27:00 PDT -
This might be just what I need-- to bad the link to the figure seems to be broken.
I wish you would publish these in a reader friendly format so we don't have to open another link to see the figures.
Mark Dresser - 2007-24-3 18:11:00 PDT -
Neat circuit! I used a FT37-43 with 40T for the primary and 8T secondary for T1, looks like around 250KHz for the osc freq.
Doug Phillips - 2007-28-2 13:42:00 PST
