• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 63 subscribers
  • Views 434 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TPS40211 LED driver application circuit questions

Emil Eriksson70652
Prodigy 130 points
Other Parts Discussed in Thread: TPS40211

Hello,

I want to use the TPS40211 boost converter as a LED driver for driving high brightness LEDs from 12V input. I am looking at the application circuit in the data sheet (page 34) and have a few questions.

1. If I undestand correctly, dimming is done by pulling down the COMP pin to ground. Since the COMP pin is essentially the output of the EA, this turns off the switching, right? What is the purpose of the diode in series with the PWM FET? It doesn't seem like it's supposed to do any rectification so I suppose it's the voltage drop that is significant here?

2. I understand that R6 is the current sense resistor for the LED string and should be set so that the voltage drop across it equals Vsense when the desired current is flowing through it. But what is the purpose of R13 in this case? Is it required? What value should it be?

  • 0
    TI__Prodigy 20 points

    Emil,

    I think you are referring to the Design Example 2 (Figure 36) on the TPS40211 datasheet.  At least, the reference designators make sense according to your comments. 

    1. Yes, dimming is achieved by pulling the COMP voltage low.  This resets the logic and turns off the MOSFET.  The COMP voltage does not have to be pulled "all the way" to GND.  So, D3 is used to pull the COMP voltage just below the internal ramp "valley" voltage, which is 1.2V typical.  In this way, the PWM dimming recovers faster, since the COMP components are not fully discharged.  In a nutshell, it improves dimming performance. 

    2. Yes, R6 is the current sense resistor.  R13 is the input resistance to the internal op-amp.  This value is typically chosen to be around 10k ohms.  It is used in conjunction with R4, C5, and C6 to tune the error amplifier's pole and zero frequencies.  R13 and C5 set the dominant pole frequency of the error amplifier. 

    Regards,

    Joel Brassfield jbrassfield@ti.com

  • 0 in reply to Joel Brassfield
    Prodigy 130 points

    Thanks for the great reply!

    I'm just wondering, you say that the internal ramp valley voltage is about 1.2V but the MMBD7000 which is listed in the data sheet has ~ 700 mV of forward voltage for each of the internal diodes which results in approximately 1.4V which is not lower than 1.2V. Why is this?

  • 0 in reply to Emil Eriksson70652
    TI__Prodigy 20 points

    Emil,

    The forward voltage of the MMBD7000 is typically 0.5V, but can be as high as 700mV (as you said).  The VF is a function of forward current and also temperature.  The Diode's Inc. datasheet gives a good graph showing this performance.  If you need to gaurantee that under all conditions (VF variations from diode lot to diode lot, and VF variations due to temperature) the COMP voltage will always be below 1.4V (to ensure turn-off), then perhaps you can get by with using one series diode, rather than two.  The dimming will still work, but the performance of the dimming may be a little slower.  For instance, the COMP voltage will be discharged more during each PWM Dimming "off-time".  So, it will take a little bit longer for the error amplifier to slew the COMP voltage from its "off" voltage to its "operating" voltage.  But, this may be ok for your application. 

    Another thought:  What happens if we use two diodes, but the COMP voltage is not fully brought below the internal ramp valley?  Well, the TPS40211 would reduce its duty cycle to a small value during the PWM Dimming "off-time".  Instead of shutting off completely, it reduces its duty cycle, which reduces the LED current.  In this case, your PWM dimming would go from regulated current, to something slightly greater than 0A.  This would still appear as "dimmed" light.  However, the color of the light may shift a little bit, since we are now PWM'ing two different average currents, rather than 100% current and zero current.  Also, the lumen value would not linearly approach "zero" when dimming from 100% to 0% duty cycle.  It would instead go from 100% lumens to something like 5% lumens, but still follow a linear trend. 

    Regards,

    Joel