• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 519 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.

TPS2378: TPS2378 damaged when work with PoE injector and PSE in the meantime

chentsu chung
Mastermind 6180 points
Part Number: TPS2378

Hi,

The sever in the picture is PSE, connect to the midspan(UPoE injector) then connect to PTZ.

The detection mechanism of different vendor's PSE are also different. Some of them are keeping dectect and class..., but some of them detect of only if RJ-45 plug-in.

When using PSE(detect only when RJ-45 plug-in)+ UPoE injector, the power sourcing work well.

When using PSE(keeping detect type)+ UPoE injector, the power sourcing doesn't worked sometimes(result form the PSE's sequence or voltage can't meet injector's, then PoE start-up sequence error)

And this is what our end-customer found, when use sever with PSE + UPoE injector, the PTZ can't get power after power plug-in. After analysis, we found the hot-swap MOS is shorted to GND.

So we suspect that the inrush current of PSE and injector makes internal MOS(hotswap mos) damaged. The fail rate is 0.01%.

 

Here are the question:

1.May I know is there any issue happened like this before?

2.Or any way to prevent this? 

  • 0
    TI__Mastermind 41770 points

    Hi Chentsu,

    Can you help clarify the block diagram by answering below questions:

    1. TPS2378 is in injector block or PTZ block?
    2. If in PTZ block, is there PD in injector block? Or is injector only getting data (not power) from sever block?

    Note that some PSEs will cycle between detection/classification (if there is compliant PD connected) in certain operating modes (for example semi auto mode) but it should still be okay.

    I suspect that there may be a large spike between VDD-VSS of the PD. Do you have schematic for review? Also waveform of the failure? Does the injector damage any other PD? or is it just the TPZ end equipment? 

    Thanks!

    Regards,

    Darwin 

  • 0 in reply to Darwin Fernandez
    Mastermind 6180 points
    Hi Darwin,

    1. TPS2378 is at PTZ side
    2. Yes, injector block combine the injector power and the data from server

    For other question:
    TPS2378 damaged in PTZ side. And I will ask customer for schematic providing, but could you let us know these 2 things first?
    1.May I know is there any issue happened like this before?
    (you have mentioned that some PSEs will cycle between detection/classification in certain operating modes, so means nothing happened like this before? )
    2.Or any way to prevent this?

    Thank you.
  • 0 in reply to chentsu chung
    TI__Mastermind 41770 points

    Hi Chentsu,

    Thanks for clarifying. No I have not seen this type of issue before. Since the injector is not a valid PD, the PSE(server) should not be outputting power but rather only data. If the injector is sensitive to the server such that the injector is misbehaving, then the limitation could be in the injector and not the PD.

    Nonetheless, a waveform of how the injector is damaging the PD will be helpful. Probably a voltage and current measurement on the PD (between VDD and VSS) will help.

    1. I have not seen an issue like this in the past.

    2. We must first review the PD design first to see limitations. Also we must see the injector's behavior when it damages the PD with the waveform requested above. Then we can try to draw some conclusions/solutions.

    Thanks!

    Regards,

    Darwin

  • 0 in reply to Darwin Fernandez
    TI__Mastermind 41770 points
    Hi Chentsu,

    Were you able to resolve the issue? Please let us know. Thanks!
  • 0 in reply to Darwin Fernandez
    Mastermind 6180 points
    Hi Darwin,

    Yes, thanks for your hlep!