I have a doubt regarding the use of a boost converter and a current sink.

This doubt arises based on the following -very simplified- circuit.

enter image description here

Following the attached circuit, the boost will take the incoming 3.7V and provide a constant 12V output at whatever demanded current. Now, in the hypothetical example of having N number of LEDs with N resistors of different values, if I want to maintain a constant current so the LED brightness is constant (neglecting manufacture mismatches,) the most logical idea -IMO- would be to implement a current sink.

In the example, a TLC5940 is used, which will set a fixed current and, as consequence, will make the voltage change to adapt to that current. Would that make a conflict to what the boost is performing?

  • \$\begingroup\$ if your voltage stays 12v, why would the LED's current change? Seems you can just tie them to gnd, no need for extra active components. \$\endgroup\$
    – dandavis
    Mar 15, 2021 at 23:25

1 Answer 1


The TLC5940 doesn't vary the voltage.

It varies the current. It acts like a potentiometer that automatically adjusts its resistance to keep the current constant.

The TLC5940 also uses pulse width modulation to vary the brightness of the LEDs.

You set the maximum current to the LEDs with a single reference resistor, then you can vary the brightness from off (duty cycle 0 percent) to full on (100 percent.)

The only concern you have with the boost converter is that it (and the lithium cell) can provide enough current for all of your LEDs.

  • \$\begingroup\$ sorry, bad choice of words. My point is that the TLC would set a fixed current and, as a consequence, the voltage will adapt/change, but that would make a conflict to what the boost is performing. \$\endgroup\$ Mar 15, 2021 at 21:45
  • \$\begingroup\$ "as a consequence, the voltage will adapt/change," Perhaps some voltage will change by a small amount, but approximately, the voltages in the circuit will stay the same. People speak about "effective voltages", and say that PWM changes the "effective voltage". But the output of the boost converter doesn't care about that. An "effective voltage" is not an actual voltage. \$\endgroup\$ Mar 15, 2021 at 22:20
  • \$\begingroup\$ If you regulate the current to be the same with a linear device, the effect will be the same as if you just used the same sized resistor on each line, so it conflicts with the point of having different value resistors. If your device is switching and does it with mostly reactive impedance, you'll end up with a "virtual resistor" that provides the same volt drop with less lost power. What is the point of the different value resistors to begin with? \$\endgroup\$
    – K H
    Mar 16, 2021 at 4:08
  • \$\begingroup\$ Hey everyone, thanks for the answers. @KH the circuit is not something I want to implement, is just to "give shape" to my confusion. Let's say there is no physical resistor but the traces per se are extremely high ohmic. so for one LED you will have 1k ohm and for another 100ohm. that would make a difference. And that is where the TLC would set the current fixed and as a consequence, the voltage will adapt, but that would make a conflict to what the boost is performing. \$\endgroup\$ Mar 16, 2021 at 8:23

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.