• State
  • Locked Locked
  • Replies 8 replies
  • Subscribers 62 subscribers
  • Views 496 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.

LM3532: Impossible to dim to off without a noticeable step?

Leif Ilvedson
Prodigy 250 points
Part Number: LM3532


I am trying to dim the LEDs to OFF without any noticeable step at the end. The datasheet states an "14 bit equivalent dimming", which if true, should provide undiscernible step to OFF. 

The only approach I didn't try with the LM3532 is using the ALS. But I believe I have tried all the other combinations of approaches.

My best results were changing the brightness level w I2C synchronously with lowering either the PWM input or following a code generated exponential map.

As a reference: Driving a single LED with the PWM output of a microcontroller:

  • 16 bit has more than enough resolution to fade to off without any step to off. - Beautiful!
  • 11 or 12 bit would probably be the lower limit of resolution to fade to off without any step
  • 8 bit - not enough resolution here, a step to off is noticeable. 

Given these test results, I may be misunderstanding LM3532's stated "14 bit equivalent dimming".

If someone has been able to get a smooth transition to OFF with the LM3532, I'd love to know your approach.

  • 0
    Prodigy 250 points

    I should note that all my tests of the LM3532 were done using the LM3532EVM evaluation board.

  • 0 in reply to Leif Ilvedson
    TI__Mastermind 35550 points

    Hi Leif,

    The brightness code is 8 bits so the full-scale current adjust at the same time is recommended.

    BR

    Patrick

  • 0 in reply to Patrick Yao
    Prodigy 250 points

    My tests were done with the full-scale current at 0 (5mA).   

    Have you been able to achieve smooth dimming to off with no visible step?

  • 0 in reply to Leif Ilvedson
    TI__Mastermind 35550 points

    Hi Leif,

    Have you captured any waveform? I am not sure if I really understand what you mean as the "step".

    BR

    Patrick

  • 0 in reply to Patrick Yao
    Prodigy 250 points

    See the video and you will see there is a sharp cut off from On to Off at the end of the dimming.  

    https://vimeo.com/877990360?share=copy

    In the above setup:

    Channel A Only:    5 LEDs

    I2C control

    Tried multiple Full Scale Current settings (the one in the video is the lowest at 5mA);

    I give Zone Target a value of zero and let the LM3532 ramp down.

    Ramping was set to 32ms down (even at 64ms, it sharply cuts to off)

  • 0 in reply to Leif Ilvedson
    TI__Mastermind 35550 points

    Hi Leif,

    Understood, let me check with our EVM to see if there is a method to cut off the LEDs smoothly. And could you please send me all the registers programmed parameters?

    BTW, what was the minimum LED current you tested during your experiment? I would suggest use the multimeter to test the LED current (channel current) to see what's the minimum the IC could output for each channel. You can test with the 5mA full-scale current and exponential mapping.

    BR

    Patrick

  • 0 in reply to Patrick Yao
    Prodigy 250 points

    OUTPUT_CONFIG_VALUE = 0B110100;// each string A, B, C is controlled by its own Brightness Registers  

    // I've also tried having each string controlled only by A's brightness registers. 

    // Enable all three strings - I've also tried enabling only one string
    CONTROL_ENABLE = 0B111;

    RAMP_RATE = 50; // 2up32down

    CONTROL_A_BRIGHTNESS_CONFIG_VALUE = 29; // I2c w Exponential Mapping
    CONTROL_B_BRIGHTNESS_CONFIG_VALUE = 29; // I2c w Exponential Mapping
    CONTROL_C_BRIGHTNESS_CONFIG_VALUE = 29; // I2c w Exponential Mapping

    FULLCURRENT_A_VALUE = 0;// 0=5ma 19=20mA
    FULLCURRENT_B_VALUE = 0;
    FULLCURRENT_C_VALUE = 0;

    Working only in Zone 4, I control the brightness by changing the ZONE_TARGET_BRIGHTNESS, toggling between 255 and 0.

    I don't have a bench meter with me, so I am unable to measure below 1mA. In my former tests a couple years ago with the LM3532, I have recordings of 25uA at Brightness Code 58, Exponential Map, and Full Scale Current at 5mA.   That's close to my calculated expected calculation (30uA).

    The datasheet says the minimum string current is 9uA. So more than likely this is what I'm seeing as a sharp cutoff as it goes from 9uA to 0. 

  • 0 in reply to Leif Ilvedson
    TI__Mastermind 35550 points

    Hi Leif,

    Thanks, I will set up the bench test in my site. And for the 9uA min current, I have the same assumption with you so I will also test it on bench. The EVM shipment might take several days so I will probably update the test results next week. Thanks for your patience.

    BR

    Patrick