• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 83 subscribers
  • Views 567 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.

TMS320F28388D: ADDITIONAL VDD (1.2V) POWER CONSUMPTION USING CM AND ETHERNET

Michael Wohl
Prodigy 240 points
Part Number: TMS320F28388D

Hello,

We are finishing a design update migrating a F28379D based product to the F28388D for the express purpose of adding an Ethernet interface.   I need to verify that our existing1.2V VDD supply has adequate margin.   Looking at the data sheet (SPRSP14D) as well as the migration document (SPRACQ1) I do not find adequate information, specifically no details on the CM power consumption.   

Generally what I see are statements like "There is not a significant difference in power consumption between the F2838x and F2837x devices, if the same number of peripherals are being utilized", or table entries dictating maximum IDD in operational mode slightly decreasing for the new device?

Can TI provide any further information about additional VDD power consumption when using the CM as a communication processor for transferring data over ethernet?

Thank you.

Mike

  • 0
    TI__Genius 15385 points

    Hi Mike,

    Operational currents in Datasheet under Power Consumption Summary section include currents from CM. We do not have currents number per CPU, we provide the worst case number when all the CPUs are powered up and running. So for your used case where you have CM running and communicating with ethernet you should use the Datasheet numbers.

    Best Regards,

    Nirav

  • 0
    Prodigy 240 points

    Hi Nirav,

    Thank you for your quick response.

    I reached out because the power consumption summary section specifically showed a decrease in power consumption for the new part (which adds a processor) and I found this odd (I'm assuming feature size did not change).   I want to ensure I am interpreting this correctly...that worst case (as you say) is still sub 500mA?

    • The 28379D shows a VDD max consumption with all CPU's running of 495ma   ... this obviously does NOT have a CM.
    • The 28388D shows a VDD max consumption with all CPU's running of 475ma   .... this includes a CM running at 125MHz?

    Also , in section 7.54 (reducing current consumption), indicates a savings of 3.7mA for the Ethernet peripheral when its clock is shutdown.   Is 3.7mA the typical operating current of the Ethernet peripheral, or it's idle mode consumption so to speak?

    Thanks again

    Mike

  • 0 in reply to Michael Wohl
    TI__Genius 15385 points

    Hi Mike,

    Your observation has a valid point, we have optimized power in design on the new design 28388D hence the numbers are slightly lower even though it has an extra processor. Your interpretation is correct.

    Let me check with our validation team to confirm 3.7mA numbers. Will reply back to you in couple of days.

    Best Regards,

    Nirav

  • +1 in reply to Nirav Ginwala
    TI__Guru 60865 points

    Hi Mike,

    Excellent question! These device families do use similar technology, the F2838xD has mostly additional modules/features vs. the F2837xD, and you can actually infer that the CM core current consumption is about 30mA from the "Typical Operating Current Versus SYSCLK" plot in the datasheet, so totally understand your concern. 

    The F2838xD device is several years newer and Nirav is correct that this design has been synthesized with newer flows that are a little more power efficient, but the main difference here is how the "operating mode" test condition is defined. 

    If you look at the device datasheet definitions for these tests, in the F2837x device, the test case exhaustively and simultaneously uses every peripheral on the device at the max capacity or data rate.  For example, from the F2837x datasheet description of the operating currents test conditions for the communication peripherals "All of the communication peripherals are exercised in loop-back mode: CAN-A to CAN-B; SPI-A to SPI-C; SCI-A to SCI-D; I2C-A to I2C-B; McBSP-A to McBSP-B; USB"  Other sets of peripherals (analog, control, etc.) are similarly exercised to the max.  

    While this does create a very much "worst-case" power consumption, it is much much above what any application could achieve due to CPU/DMA/CLA bandwidth limitations on how many peripherals can be usefully utilized at one time.  F2838x family still uses a very conservative benchmark (all peripherals enabled, powered up, and clocked; all CPUs running at max speed with 100% utilization; analog and flash peripherals operating) along with appropriately conservative characterization and statistical methods to generate the max currents numbers, but it is a little less excessive than the F2837x benchmarking method, hence the slightly reduced power numbers.       

  • 0 in reply to Devin Cottier
    Prodigy 240 points

    Thanks very much for the detailed response Devin and for your support Nirav.   I am confident of our supply's margin at this point.

    -Mike