TMS320C6678: Power rail bleed-back during power up sequence (from 1.0V CVDD1 to 1.5V DVDD15)

Kyle Steiner

Prodigy 25 points

Part Number: TMS320C6678
Other Parts Discussed in Thread: TPS40322

I have a question regarding power sequencing of the TMS320C6678 DSP.

I noticed the following behavior upon powering up the device:

For the sequencing shown below,

Yellow - CVDD

Cyan - CVDD1/VDDT1

Magenta - DVDD18/AVDDA1/AVDDA2/AVDDA3

Green - DVDD15/VDDR1/VDDR2/VDDR3/VDDR4

There appears to be bleed-back from the CVDD1/VDDT1 (cyan) input to the DVDD15/VDDR1/VDDR2/VDDR3/VDDR4 (green) rail. I was not expecting this behavior based on the sequence given in 7.3.1 Power-Supply Sequencing from document SPRS691E.

I'm looking for input on what may be the cause of this and what should be done to remedy the issue. What are the implications for reliability of the device?

In the design, the TMS320C6678 is the only component powered by the CVDD (yellow), CVDD1/VDDT1 (Cyan), or DVDD15/VDDR1/VDDR2/VDDR3/VDDR4 (green) regulators. Each of those three regulators supply power only to the DSP (there is some ancillary filtering using passive components that are unique to a given voltage rail).

The regulator powering DVDD18/AVDDA1/AVDDA2/AVDDA3 (magenta) does power a few other components, but there does not appear to be an bleed-back issue with that power form.

From our testing, I'm not aware of any issues with the device at this time; but we are still in the early stages.

Thank you for your assistance.

over 7 years ago

0 Yordan Kovachev over 7 years ago

TI__Guru**** 161600 points

We're looking into this.

Best Regards,
Yordan

0 Tom Johnson 16214 over 7 years ago in reply to Yordan Kovachev

TI__Mastermind 46460 points

Kyle,

This is known and documented. Please refer to the document TMS320C6678 Multicore Fixed and Floating-Point Digital Signal Processor Silicon Revision 1.0, 2.0 Silicon Errata (SPRZ334H) at: in "Usage Note 18 Current Flow Between VDDT and VDDR Rails During Power Sequencing of DSP".

Tom

0 Kyle Steiner over 7 years ago in reply to Tom Johnson 16214

Prodigy 25 points

Thank you for your response.

Kyle

0 RandyP over 7 years ago in reply to Kyle Steiner

TI__Guru* 84110 points

Tom,

The switching power supply providing CVDD to the DSP shows a blip during the initial rise (yellow trace above). Is this a potential issue?

The switcher consists of the TPS40322 controller. The rise time is set to 2ms. From the design side, it appears that this behavior is part of the switcher (simulations show a similar transition between 2 rising slopes).

Regards,
RandyP

0 Tom Johnson 16214 over 7 years ago in reply to RandyP

TI__Mastermind 46460 points

Randy,

Ideally, the power supply ramp-up should be monotonic. You say this is coming from the switcher power stage. Can it be eliminated by changing ramp time or other similar change? Can you see in the simulations what causes this? You may need to discuss this with an FAE.

Tom

0 Tom Johnson 16214 over 7 years ago in reply to Tom Johnson 16214

TI__Mastermind 46460 points

Randy,

This thread had fallen out of view since it was previously closed out. Is there any more open on this or can it be closed again?

Tom

Processors

Processors forum

TMS320C6678: Power rail bleed-back during power up sequence (from 1.0V CVDD1 to 1.5V DVDD15)