• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Subscribers 63 subscribers
  • Views 1034 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.

TPS2372: Over-current Protection

Nicholas Smith57318
Prodigy 195 points
Part Number: TPS2372
Other Parts Discussed in Thread: , TPS2373

Hello,

I was performing over-current tests on the TPS2372, using both the TPS2372-4EVM-006 Eval board and a board I designed myself with the TPS2372 was the PD PoE controller.   

When TPS2372 is configured for Type 4 PoE and I pull more than 70W with a 300+ foot Ethernet cable (or 90W with a 1-foot cable, equivalent to 70W with a 300ft cable), instead of the TPS2372 protecting itself and opening up the internal MOSFET to stop current consumptions, it seems the TPS2372 is trying to regulate the PoE output voltage/current even beyond these power consumptions (see oscilloscope screenshots below).  During a PoE over-current event, the TPS2372 gets very hot (~130C) very quickly.  Instead of the TPS2372 completely shutting down the current flow, the PoE output voltage and current oscillates at 50Hz~100Hz causing the rapid 130C self-heating.  I did not leave it in this state long enough to damage the TPS3272 part, but I assume it would eventual be damaged if I did leave.

By the way, the same thing happens if the TPS2372 is configured for Type 1 PoE, and I try to draw 30W.  Same over-current event happens, and the voltage and current oscillations occur resulting in the same self-heating up to 130C.  Please help!

My setup includes a PoE injector connected to the TPS2372-4EVM-006 Eval board (also tried with my own board with same results), and a resistor load box connected to the VDD/RTN output power lines.  I tried to pull up to and beyond 90W.  I was anticipating that the TPS2372 would protect itself and shut off the power if tried to pull to much power, but it instead the voltage and current 

     

Thanks,

Nick Smith

  • 0
    TI__Mastermind 21320 points

    Hey nick, 

    Thank you for the question. In the TPS2372-4EVM-006, Figure 5 shows what we want the over current protection feature to look like. Can you get these waveforms during your test and show them here?

    I will try to replicate this issue in the lab tomorrow. 

    If this post answers your question, please indicate so by marking this thread as resolved. Thank you.

     

    Regards, 

     

    Michael P.

    Applications Engineer

    Texas Instruments 

  • 0 in reply to Michael Pahl
    TI__Mastermind 21320 points

    Hello Nick, 

    I was able to reproduce this in the lab. What is happening is during the overcurrent, the IC reverts into the in-rush phase after the overcurrent phase is done. The internal hotswap FET opens, so there is a voltage across it. This makes the IC get hot. If it gets out of inrush, there is now a full load current through the internal hotswap, which also contributes to the heat. 

    Therefore it is expected that the IC will get hot during an overcurrent event. However, the IC has Over Thermal protection as well, so it should be protected in this situation too. 

    Are you observing any damage?

    If this post answers your question, please indicate so by marking this thread as resolved. Thank you.

     

    Regards, 

     

    Michael P.

    Applications Engineer

    Texas Instruments 

  • 0
    Prodigy 195 points

    Thanks Michael. 

    No, I'm not observing any damage to the TPS2372, but I'm also not leaving it in this over-current state very long.  I have not tested to see if the part will in-fact protect itself if it reaches the thermal shutdown state.  Has this been tested in the factory?

    In any case, I would have expected the part to have over-current protection without getting so hot (>130C) and not have to rely on the thermal shutdown to protect against an over-current event.  Also, I could add an external over-current supervisor chip but was wanting to avoid this for board space.  There are other PoE PD controller chips that to do protect against over-current events. Is there a way to configure the TPS2372 to protect against an over-current event without getting so hot?  We still want hot-swap protection however.

    Thanks,

    Nick  

  • +1 in reply to Nicholas Smith57318
    TI__Mastermind 21320 points

    Hey Nick, 

    Yes we let the IC sit for 10 minutes to reach thermal stability to ensure the IC does not damage from a short. 

    Part of this OCP strategy is to keep the output maintained. This type of short is on VDD_RTN, which shorts input power, which can sometimes be brief. Our other devices that have a PWM controller integrated will use Hiccup mode to address a short on the output. But this is how we address this type of short. 

    One strategy you could use is to disable the PSE. Using an optocoupler like below, we can disrupt detection and thus the PSE will shut off power. 

    Other IC's like the TPS2373 has an adapter pin, APD, which could be used to turn off the internal hotswap FET altogether. That would be a more cost effective solution but require an IC change. 

    If this post answers your question, please indicate so by marking this thread as resolved. Thank you.

     

    Regards, 

     

    Michael P.

    Applications Engineer

    Texas Instruments