Is there connections between the power rails inside the DM8168?

Paul Koenig

We are bringing up a prototype board with the DM8168 and confirming the power sequence and timing. We noticed the 1.0V AVS rail seems to come through on the 1.5V ALT rail, or one of the 1.5V rails. We also observed this voltage "mirroring" on the EVM board from Spectrum Digital, so we do not think this is isolated to this particular board and layout. Anyone else observe this behavior?

Thanks,

Paul

over 14 years ago

0 MarcPyne over 13 years ago

TI__Mastermind 25055 points

Paul,

Sorry for the slow reply. Our Design teams have been looking into this, and they do not see any significant concern with the behavior observed during power sequencing. This same sequence has been approved by the owners of the relevant device modules.

Regards,
Marc

0 Omin Chou over 13 years ago in reply to MarcPyne

Genius 3555 points

Dear MarcPyne,

The below waveform denotes D1.8V(CH1 : yellow) and D1.5V(CH2 : cyan) and that is a power on sequence on DM8168EVM Rev B. The dashed line indicates the power rail ENABLE. We see both power rails start to ramp up before ENABLE. I would like to know is this acceptable? Will this lead to any problem? Will this damage the chip? How can we eliminates this pre-ramp?

Please help me to confirm the above concerns. Thanks in advance.

B.R.

0 MarcPyne over 13 years ago in reply to Omin Chou

TI__Mastermind 25055 points

Hello,

This comes from some leakage on a couple of the domains. Our Design teams do not see any significant concern with the behavior observed during power sequencing.

Regards,
Marc

0 Garth over 13 years ago in reply to MarcPyne

Prodigy 130 points

But I also meet the same ting(power rail couple) in your EVM(DDR2 slim version EVM),the same as O.C.

0 Tomokazu Miyazaki over 13 years ago in reply to Garth

Intellectual 270 points

This has happened to us also in TI's EVM and our custom board.

I thought power rail couple is started at 1.0V_CON rising timing.

0 Omin Chou over 13 years ago in reply to MarcPyne

Genius 3555 points

Hi Marc,

Can you provide a document to prove this leak does not impact the chip to work normally? I think a lot of engineers have concern about this issue. Please give us a explanation to let everyone reassurance. Thanks a lot.

B.R.

0 MarcPyne over 13 years ago in reply to Omin Chou

TI__Mastermind 25055 points

Hello,

We will add a Note on this in the next datasheet revision in the Supply Sequencing section.

Regards,
Marc

Processors

Processors forum

Is there connections between the power rails inside the DM8168?