TLC59108: Delay time from register set to output applied

Hi Higa,

What is the 'resister set' mean? Does customer want to change REXT resistor during operation?

Hi Wu-san

I'm sorry, it was a typo.
The correct is "register set".

Best regards,

Higa

Hi Higa-san,

Here is a test result for your reference. From sending commands to output active, it takes around 5ms.

Hi Wu-san

Thank you for your support!

Could you share your I2C command with me?

Best Regards,

Higa 

Hi Higa-san,

I wrote group PWM register 0x0A from 00 to FFh.

Hi Wu-san

Thank you for sharing.

My customer would like to know your all register setting.

In particular, they interest whether the delay time does not change depending on the OCH setting of MODE2.

Could you provide that?

Best Regards,

Higa

Hi Higa-san,

The delay time would not change whether this bit is 0 or 1. When set as changed on ACK, the output would changed after the related register is written. When set as changed on Stop condition, if auto increment function is used, several registers are written together in the one I2C command and all outputs would changed together after stop condition sent. But the delay time is around 5ms no matter what the trigger point is. 

Hi Wu-san

Thank you for your kind support.

I'm sorry for the additional request, but could you please provide the waveform data in the I2C sequence below?

This is a request from my customer.

1. PWM control (0x0A) 00h-> FFh-> 7F

2. Set the OCH bit in the MODE2 register to 0-> 1 and execute the No. 1 sequence.

The reason they asked is because they want to understand from the data that the delay time does not change under the following two conditions.
- Continuous PWM setting
- When OCH is changed from the default value "0" to "1"

Thanks and Best  Regards,

HIga

Hi Higa-san,

Yes I can test it but I am out of office today. I will reply you in next Monday.

Hi Wu-san

Thank you for supporting.

Do you have any update?

Thanks and Best Regards,

Higa

Hi Higa-san,

Sorry that I am out of office in business trip. The results will be replied for you in tomorrow.

Hi Higa-san,

Here is the test results for your reference.

The conclusions are:

1) OCH value does not influence the delay time significantly.
2) PWM value change vary from the work state:
    >From off to on, the time is most long around 5ms;
    >Changing the PWM setting in different value need around 1ms;
    >Turning off the outputs will be valid immediately.

TLC59108 bench test.zip

Hi Wu-san

Thank you for your kind support.

I may want to ask you for confirmation.

From what I see, all your data are exactly the same timestamp.

Are these all the same data?

Thanks and Best Regards,

Higa

Hi Higa-san,

Sorry that it seems there are some issues happened on the scope that only one image was saved, which replaced others. The conclusion I observed could be used for communicate with customer firstly. For the bench waveforms, I can re-test for you later. But the covid epidemic situation is not optimistic at Shanghai and we all start work from home. The results will be sent to you maybe 2 or 3 weeks later. I think you customer could test it on TLC59116's EVM, which is 16-channel version of TLC59108 and have the same behaviors.

Hi Wu-san

I understand your situation.
If your view doesn't change, no additional testing is needed.
Thank you for your devoted support so far.

Thanks and Best Regards,

Higa