• State TI Thinks Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 776 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.

TLC59581: ES-PWM and LOD functions in the LED driver

Kasia Nowak
Prodigy 30 points
Part Number: TLC59581

Hi,

I am using TI TLC59581 LED driver but I am confused about the operation of the ES-PWM in this device. This device has a 16-bit Grayscale control so 65536 GCLK pulses  are needed. In the 8+8 ES-PWM mode, the total number of GCLK pulses is divided into 256 segments (in the 7+9 mode, these are 512 segments). The TLC59581 application note mentions that each segment contains 256 x 32 GCLK pulses in the static driver case and 257 x 32 X 256 GCLK pulses when 32 multiplexing is used (one sub-period with 32 segments). The timing diagram of this case is shown in Fig. 15. For the 7+9 ES-PWM,  the application note describes that each segment is divided into 512 segments and each segment has 129 GCLK pulses in the static driver case and 129 x 32 x 512 GCLK pulses when 32 multiplexing is used (one sub-period with 32 segments). The timing diagram of this case is shown in Fig. 17.

Questions:

1. Why do you use 257 x 32 GCLK pulses in each segment (129 x 32 GCLK pulses for the 7+9 mode) and not 256 x 32 GCLK pulses? Please, note that 257 x 32 x 256 is higher than the total number of GS pulses (=65536). This is also the case for the 7+9 mode: 129 x 32 x 512>65536. 

2. Why there is a difference for the static driver case for the 8+8 ES-PWM mode (256 GCLK pulses per segment) and the 7+9 ES-PWM mode (129 GCLK pulses per segment)?

Another question is about the LED open detection (LOD) function. When exactly the LOD is monitored? Does this happen in each segment?

Thank you for your feedback.

Kind regards,

Kasia

  • 0
    TI__Genius 14295 points
    Hi, Kasia,

    1. The one more GCLK is to make sure there's enough time for like 256 GCLK time, that is one more is redundant and only 256 of it is used.
    2. 7+9 can achieve double high refresh rate since it has double segments.
    3. Yes. The OUT voltage is monitored each segment.

    Please click Resolve if this answers your questions.

    Thanks.
    Regards,
    Kenneth
  • 0 in reply to Kenneth Du
    Prodigy 30 points

    Hi Kenneth,

    Thank you very much for your answers. Just three more questions to be sure I get it right:

    1. What do you mean by "one more GCLK is to make sure there is enough time for 256 GCLK time"? Why 256 GCLK pulses would not be enough? If this to make sure that in case of a Grayscale data Fh (all 1's --> 65536 GCLK pulses must be counted) there is enough time for the transition to the next segment?

    2. Is this extra GCLK added automatically by the TLC58591 or the incoming Grayscale data must be so that they include this extra clock? I guess the former (= an extra clock is added by GCLK) as the input data are always 16-bit. Is this correct?

    3. I understand that the 7+9 mode is meant to increase the refreshment rate. However, why in the static driver case (=no multiplexing), an extra GCLK  at the end of the segment is added only in the 7+9 mode and not in the 8+8 mode?

    Thank you once again for your explanation.

    Kind regards,

    Kasia.

    PS. Is Fig. 17, upper picture for the Static Driver, the waveform showing GCLK pulses correct? If each segments has 129 GCLK pulses, than after 512 segments the last pulse should have a number 66,048 (=129 x 512) but in the figure the last pulse is 65,536. Also for earlier segments the shown numbers are incorrect (e.g. 63 segments should finish at 8,256 = 64 x 129 but the marking is 16,384, like in the 8+8 mode). Is this wrong or I am missing something?

  • 0 in reply to Kenneth Du
    Prodigy 30 points
    Hi Kenneth,

    Thank you very much for your answers. Just three more questions to be sure I get it right:

    1. What do you mean by "one more GCLK is to make sure there is enough time for 256 GCLK time"? Why 256 GCLK pulses would not be enough? If this to make sure that in case of a Grayscale data Fh (all 1's --> 65536 GCLK pulses must be counted) there is enough time for the transition to the next segment?

    2. Is this extra GCLK added automatically by the TLC58591 or the incoming Grayscale data must be so that they include this extra clock? I guess the former (= an extra clock is added by GCLK) as the input data are always 16-bit. Is this correct?

    3. I understand that the 7+9 mode is meant to increase the refreshment rate. However, why in the static driver case (=no multiplexing), an extra GCLK at the end of the segment is added only in the 7+9 mode and not in the 8+8 mode?

    Thank you once again for your explanation.

    Kind regards,

    Kasia.

    PS. Is Fig. 17, upper picture for the Static Driver, the waveform showing GCLK pulses correct? If each segments has 129 GCLK pulses, than after 512 segments the last pulse should have a number 66,048 (=129 x 512) but in the figure the last pulse is 65,536. Also for earlier segments the shown numbers are incorrect (e.g. 63 segments should finish at 8,256 = 64 x 129 but the marking is 16,384, like in the 8+8 mode). Is this wrong or I am missing something?
  • 0 in reply to Kasia Nowak
    TI__Genius 14295 points
    Hi, Kasia,

    1. One more is to make sure there's no mistake for example only 255 GCLK is detected.
    2. Correct.
    3. No matter 129 or 257 GCLKs is in put, only 128 or 256 will be counted, so it's both 65536 in total.

    Kenneth