TMS320C6748: Impacts of Errata Advisory 2.3.18 for TMS320C6748 on other devices

Hi Nomo,

I am following up with experts who worked on these parts. I expect to get back to you within a day or two.

Regards,

Colin

Hi Colin-san,

Thank you for your reply.
It would really be appreciated, if you would follow up the issue.

Best Regards,
Nomo

Hi Nomo-san,

The experts confirmed that this particular C6748-Silicon Errata issue has no history or reason to be present in the DM368 units. 

Has this issue been observed on multiple units or boards? Has the customer tried an A-B-A swap to see if the issue follows a particular unit or board?

Regards,

Colin

Hi Colin-san,

Thank you for your reply.
This issue has been observed on multiple boards.
And the customer tried an A-B-A swap and they found that the issue depends on the power-up sequence (The raising time of 3.3V).

Could you confirm the following figure?
This phenomen seems to exact matches to Errata Advisory 2.3.18 for TMS320C6748.

We found that this issue seems to be resolved by dealing with one of the following three.
- Shorten the raising time of 3.3V from 4.3ms to 610us
- Disconnect 3.3V power from dual voltage IOs
- Add large load capacitor on the 1.8V power line

Could you show us your perspective for the results?

Best Regards,
Nomo

Hi Nomo-san,

let me discuss this issue internally with more experts. Your waveforms and observations are very interesting. I would like to discuss with additional engineers.

Regards,
Colin

Hi Colin-san,

Thank you for your cooperation.
The issue is critical for us, so it would really be appreciated if you would keep working on this.

Best Regards,
Nomo

Hi Colin-san,

Do you have any update on this query?
It would really be appreciated if you would tell us the status.

Best Regards,
Nomo

Nomo-san,

The belief right now is that the IO and interconnects are similar but not the exact same. They may be similar enough though that the issue on both devices could be the same.

Can you implement the Work-Around from the C67xx Silicon Errata on your newer DM368 application? If so would that solution suffice?

Regards,

Colin

Hello Colin-san,

Thank you for your reply.

The work around from the C67xx Silicon Errata seems to be effective on the DM368 application.
However, they need an evidence to apply the work around on their DM368 application.

Because the issue occurred after their mass production (i.e. in the market), they cannot make major design modifications.
So, they're considering add the large capacitive load on 1.8V power line.
However, they cannot determine the size of the capacitive load, because the detail value is described only for C67xx not DM368.

They're wondering if the same capacitive load described in the C67xx Silicon Errata can be applied on their DM368 application.

Best Regards,
Nomo