3
\$\begingroup\$

I am planning to build a device that includes an high power RGB LED (either a lightsaber or a decorative object for the girlfriend - I still have to decide), where high power means about 3x3W. I've found this little guy from Texas Instruments that seems to perfectly suit my needs, that's a sequential driver so I'll need only one inductor and one diode for the DC-DC, it has an I\$^2\$C interface and some other fancy functions.

There are a couple of questions that arise though.

1) Refer to page 4, bottom of the table, where the current accuracy is reported, and to pages 18-19, inductor choice. As per the EC table it seems that the LED current can be tuned from the maximum set by the resistors \$R_{1,2,3}\$ (first page schematic) down to approximately zero, while in the inductor choice paragraph a \$I_{LED,MIN}\$ is calculated and is as high as 600mA. Since my end application would have some 750mA as maximum current that would be unacceptable of course, so the question is: is that \$I_{LED,MIN}\$ another parameter or is it exactly what it seems to be?

2) As I said this is a sequential driver, only one LED at a time can be turned on. Since its target application seems to be pico projectors I am hoping I can hop from one LED to another fast enough to let the human brain do the mix and end up with a beautiful RGB light. To do so I'll need to bang the G,B,RCTRL inputs appropriately, i.e. one LED will be on only for one third of a given slice of time, e.g.:

time: 0  t 2t 3t 4t 5t    
RCTRL 1  0  0  1  0  0 
GCTRL 0  1  0  0  1  0
BCTRL 0  0  1  0  0  1

where \$3t\$ is the total time needed to turn on each of the three LEDs. Two questions arise then:

a) If \$t_R, t_G \text{ and } t_B\$ represent the time a given LED is on, is it correct to assume that they should be equal, i.e. \$t_R=t_G=t_B=t^*\$? Or I might want to tune them because for example for the huma eye blue seems less bright than red or whatever?

b) My LEDs are rated up to 750mA. Should I set the maximum current for each LED to a higher value because each of them is on only for \$\frac{1}{3}\$ of the time? I think that setting it to 2.25A, apart from the fact that the chip can deliver 2A at most, would burn my LEDs for sure but what about 1A or whatever? The LED I have is this (no advertising intended).

\$\endgroup\$
2
\$\begingroup\$

a) They generally look white if mixed with same duty factor. Later you may want to change the color temperature of "white" by playing with tr, tg and tb.

b) LEDs are over-driven to compensate for low brightness due to multiplexing. Thus you can overdrive them by a factor of 3 if the duty factor is 1/3. However, they have a specification for the maximum pulsed current. In your case, look into the datasheet if you got one. Otherwise, give it a try and keep an eye on their temperature while over-driving them with maximum allowed pulsed current.

1A pulses should not be a problem.

| improve this answer | |
\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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