• State Resolved
  • Locked Locked
  • Replies 1 reply
  • Subscribers 63 subscribers
  • Views 716 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.

TPS2373-4EVM-758: Passive PoE/DC Injection on Ethernet Pairs

Daniel Seabra de Andrade
Prodigy 10 points
Part Number: TPS2373-4EVM-758
Other Parts Discussed in Thread: TPS2373

Hi all,

We just ordered a TPS2373-4EVM-758. The board worked well when plugged into a 802.3bt-compliant PoE (the switcher down-converted to ~5.2V which was available at the output and the LED turned on).

However, the board failed after being plugged in to a passive PoE injector (that is, one that simply injects 57V DC into two Ethernet pairs and grounds the other two, with no classification/detection). We did not expect this. Is this expected behavior? What is the failure caused by - inrush current?

After some debug, it seems that the bridge rectifier FETs failed when this happened, and afterwards would short out the PoE supply (regardless of whether it was passive or 802.3bt-compliant). Perhaps more alarmingly, the TPS2373 itself failed after we pluggled in a DC supply at the adapter at 44V (the supply itself briefly shorted, and then the chip developed a sizable hole at pin 1). 

Right now we have removed the FETs and the TPS2373 and have ordered these parts to try to rework the board. It no longer shorts out the PoE supply, and the DC adapter side seems to still be working (the DC/DC converter bucks it down to 5.2V when supplied by an external 13V VC_OUT). We plan to continue testing when we have reworked the board.

However, we need to understand the failure mechanism, because we do expect both 802.3bt-compliant PSEs as well as passive PoE injectors. How can we modify the circuit to mitigate this issue? Is this an issue with the TPS2373, inrush current due to the 100uF capacitor, the active-FET bridge rectifier, or a combination of these?

Thank you,

Daniel

  • 0
    TI__Mastermind 41770 points

    Hi Daniel,

    Under the passive PoE injector and adapter plug in conditions, there was probably a large spike that exceeded the abs max of the FET bridge (100V) and PD controller (100V). I also believe this because there is damage near the VDD (pin 1) pin.

    Since the passive PoE injector is non-standard, it probably does not follow the IEEE802.3 standard for a PSE turn ON slew rate  and might be acting like a hot plug which can create a larger than 100V spike. For the adapter, there is no inrush protection since the PD controller is bypassed, and there is no TVS between VDD-RTN (I see some customer designs add a TVS at the adapter input if they expect these types of transients to be present in their adapters). It may be good to have this system level projection under your adapter and passive PoE selection.

    For protection the bridge you can try adding bidirectional TVS across the pairs of the PoE input (before the bridge).