• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 62 subscribers
  • Views 141 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

Original question:

TPS2372-4EVM-006: AutoClass Function

TPS2372-4EVM-006: Pclass < Ppeak_PD compatibilty problem?

Martin Sukup
Prodigy 10 points
Part Number: TPS2372-4EVM-006

Can you explain how IEEE 802.3at-2009 Class 1 and 2 power levels between PSE (Pclass) and PD (Ppeak_PD) can be compatible given Table 33-7 Pclass supply limits (4W and7W) are less than Table 33-18 Ppeak_PD consumption limits (5W and 8.36W)?

From a previous thread on this part number it seemed like there is a bit of a spec gap and that the PSE with limits at the minimum would remove power if  Ppeak_PD is exercised, and systems need to consider this possibility. Is that true? Seems odd for an otherwise pretty thorough spec...

  • 0
    TI__Mastermind 20961 points

    Hi Martin,

    Thanks for reaching out! I will take a look and get back to you later.

    Best regards,

    Diang

  • 0 in reply to Diang Xing
    TI__Mastermind 20961 points

    Hi Martin,

    Thanks for your waiting.

    Since PD is the load side which demand how much power the PSE needs to send, there could be an instantaneous demanding power (< P_peak_PD) from PD which is higher than the PSE's steady operating power meanwhile not trigger PSE's power cut protection. 

    The P_peak_PD values at Class1 and Class2 are in below:

     802.3at:

    802.3bt:

    In 802.3bt, more details about the P_peak_PD values calculation are discussed:

    Best regards,

    Diang

  • 0 in reply to Diang Xing
    Prodigy 10 points

    The PSE may not trigger it's power cut protection, but under the standard it would be allowed to, correct? See Table 33-7 and section 33.2.7.10 from 802.3at. So this mean that PSE and PD could both meet the standard, but that the PSE could still cut power to the PD during a peak demand event. Is this true? It seems like the standard doesn't handle this well. And it is odd because it is only true for class 1 and 2, not 0 or 3.

  • 0 in reply to Martin Sukup
    Prodigy 10 points

    I guess at this point the best explanation is that Pclass is average power and to avoid resets the PSE would also have to be able to supply the peak loads of the PD + path loss. It is tricky/confusing because the peak power the PSE must supply isn't defined in 802.3at-2009 so I guess you just derive it yourself. Adding to that it is interesting that Pclass is greater than Ppeak_PD for class 0 and 3, but now 1 and 2.

    Does this assessment seem correct?

  • +1 in reply to Martin Sukup
    TI__Mastermind 20961 points

    Hi Martin,

    Thanks for your reply. 

    I agreed that "A PSE may remove power from a PD that causes the PSE to source more than P_Class." means there is no hard restriction for PSE to remove power or not for an instantaneous event P_Class < P_PD. The safest way is always below P_Class

    In the T_cut description, it also uses "may remove power" too. And I_cut here is optional limit. I guess for PD the load power is the key request, while for PSE & Ethernet cable the copper current is the key request, and the standard needs to cover both of PD and PSE along with a 0-100m cable.

    Limiting the power too low it may not utilize the Ethernet cable very well and makes PoE not an attractive power source solution (and some instantaneous high power is not truly harmful). Limiting the power too high needs to consider the risk of cable overheating in continuous power accumulation. There may be a middle zone for customer to do design optimization for both performance and reliability, meanwhile some of PSE is programmable in power cutting with host.  

    Best regards,

    Diang