• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 1 answer
  • Subscribers 64 subscribers
  • Views 675 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.

TPS61500: LED Driver for Large number of LEDs from Li-Ion Battery

J.Schweiger
Expert 1745 points
Part Number: TPS61500
Other Parts Discussed in Thread: LM5155, TPS92200, TPS92390, LM3481, LM3478, LM5156, LM5157, LM3488, TPS92692

Hello,

I have an application where I need to drive around 100 white LEDs. Each LED will be driven at around 3V, 40mA.  I am currently planning to use 1 or 2 18650 batteries to power the application (2.9-4.2V).  These specifications are flexible at the moment, the key being how to drive large number of LED's efficiently.  The application will not include an MCU, but I do plan to put in a potentiometer to control power.  We want to optimize efficiency across all ranges of power (I plan to add circuitry to put boost in standby when pot is set to off).

From my research, it seems the most efficient approach is to boost the voltage, regulate total current, then use current mirrors on each series connection to split current evenly.  My need for 12W of output power seems to be a concern in this direction though.

1. TPS61500 - I am leaning towards this as it seems like the most integrated solution, however I am worried that it will not provide enough power for my desired 12W.  I am not sure if I can use two of these as we want the control to be based off a single potentiometer.  Do you have any similar solutions that allow higher power output?

2. LM5155 - It appears this can provide more power, however I'm not sure if it allows for current based feedback meaning lower overall efficiency as I need to adjust current externally.  I am considering using something like this with the TPS92200 current-based buck regulator, or perhaps simply plan for less efficiency by dumping power in resistors.

3. If I put 2x 18650's in series, I may use something like the TPS92200 to efficiently buck the current down to 100 parallel connections, however this would require 100 transistors for current mirrors which I'd really like to avoid.

I appreciate any insights you have of the best method and/or suggested parts for driving 100x LEDs from a battery with high efficiency.
Thank you,
-Jon

  • 0
    TI__Genius 15776 points

    Hi Hardy. could you help in this case.

  • 0
    TI__Genius 12915 points

    Hi Jon,

    TPS61500 is our highest current limit device with integrated power switch. No higher power device could be recommended. What the application you want to design with?

  • 0 in reply to Hardy Wu
    Expert 1745 points

    Hello Hardy.  The application is driving a large number of white LEDs (e.g. 100x 3V/40mA LEDs) from a Li-Ion cell (3-4.2V).  Can I use two of the TPS61500 with a single potentiometer for current setpoint?  What do you think of the other options I mentioned in my original post?

  • 0 in reply to J.Schweiger
    TI__Genius 12915 points

    Hi Jon,

    Do you mean using one potentiometer on FB to control 2 the current of 2 device? I did not try this solution before but it seems workable.

    You could take a look of TPS92390/1 with a external switch FET solution.

  • 0 in reply to Hardy Wu
    Expert 1745 points

    Thank you Hardy. 

    Yes, we want a single potentiometer to be able to set the current output.  Can you clarify how you would suggest this solution?  It appears both TPS61500 and TPS92390 do not have analog voltage settings, just PWM.  As I do not have an MCU in this design, do I need to put in a circuit to create a PWM from the analog potentiometer?

    Can you clarify what maximum output power for the TPS61500 and TPS92390 is?  Note that I can arrange series/parallel configuration to achieve maximum efficiency, but overall I'm wanting to be able to go up to about 12W from a 4V or 8V source.  I do value the simplicity of the TPS61500 over the TPS92390, however think if multiple TPS61500 are required for my desired power output then the TPS92390 may end up being a simpler solution.

    I appreciate your support.
    -Jon

  • 0 in reply to J.Schweiger
    TI__Genius 15776 points

    Hi Hardy, please help to keep following this case.

  • 0 in reply to J.Schweiger
    Expert 1745 points

    Hardy, I haven't heard back on my last message. 

    Based on Figure 1 in the datasheet, the TPS61500 is going to be limited around 12W with a 5V input which doesn't allow any overhead.  I have been looking at other parts including LM5155, LM5156, LM5157, LM3481, LM3488, LM3478, and LM3481.  There seem to be way too many options for me to feel confident I'm selecting the optimal part for my application, but I'm leaning towards the LM5155.  It shows an example in section 10.2 with 6V input, 24V 2A output operating at 94% efficiency.  It also shows in Figure 10-9 an LED driver which seems to mimic the same functionality as the purpose-built TPS61500.  It seems like the LM5155 is a much better part than the TPS61500 for power capability and efficiency (with understanding the need for external FET), am I missing anything here?  Do you have any comments on the slew of other components I've found that could fit my application?

    I appreciate any insight you have to help me make a decision of which part to proceed with.

    Thank you,
    -Jon

  • 0 in reply to J.Schweiger
    Expert 1745 points

    I should also add TPS92692 into the conversation as it is advertised as a "LED Controller" (I didn't mention earlier as it has stocking issue at the moment).  This has a 5A switch and from Figure 56 it can be 90% efficient at 9 LEDs x 500mA (I would split this across 10 parallel connections to get 90 LEDs).  The LM5155 seems like it's more efficient, but I'm interested in understanding the difference.

  • 0 in reply to J.Schweiger
    TI__Genius 15776 points

    Hardy, please keep following this case.