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.

TPS23754: Use of PoE rail after classification

Part Number: TPS23754

Hi,

we need to use the 48 to 57V PoE rail voltage after classification for a local load, which does not need galvanic isolation. See our schematic attached.

We have designed a low-side switch, which disconnects the load at power-up, in order not to interfere with classification. After the classification is done and PoE voltage rises to nominal values, the 39V Zener D17 breaks through and puts a positive bias on the gate, effectively connecting the load.

When we do it with a commercially available PoE converter, this works fine. However, our tps23754 schematic will not work: LED3 will not come on, the PoE voltage stays at around 36 V and the tps23754 gets really hot. I suspect that the hotswap FET is not fully engaged for some reason.

Do you see anything wrong in our design?

The 12v output of the flyback only supplies app. 2-3W of power.

BR

Michael

  • Hello Michael,

    There is only 18.8uF of bulk capacitance - and they are all ceramic. This is too low, especially if the VDD rail is going to be powering another load.

    I would suggest 47uF of electrolytic cap. 

    I have tried to use all ceramics before, and the DCDC will not startup without ESR. If you must use ceramics, insert a 1ohm resistor to mimic ESR. But you must increase it. 

    Does the TPS23754 design work if you remove D17? That will help us understand if the DCDC is working properly.

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

     

    Regards, 

     

    Michael P.

    Applications Engineer

    Texas Instruments 

  • Hello,

    thanks for the afst answer. I added an 100uF electrolytic cap to the 4x 4.7uF ceramics. However, this did not solve the problem.

    When I disconnect D17, the tps23754 is working fine. Also, when I connect D17 manually after startup, it works fine.

    I am thinking of using the TPS26625DRC for feeding the PoE rail.

    BR

    Michael

  • Hello Michael,

    A 100uF would be too much. There is a bulk capacitance range for every PoE class, as defined in the IEEE802.3 standard. 

    I've been thinking about it and I wonder if this setup is messing with the in-rush phase? 

    If there is a successful PoE handshake with a Type 2 PSE, but T2P stays low, I wonder if the device is stuck in the in-rush phase. Potentially the diode is taking enough current to make the PSE and PD stuck here. 

    So some suggestions I have are: 

    - maybe trying a higher voltage diode? Try something above 42.5V. 

    - Could you use T2P signal to drive the MOSFET? The T2P signal is active low when a Type 2 PSE is connected, so this signal could work here. 

    - I am also thinking you could try to use the VC voltage instead of T2P? You would probably need a bigger cap on that pin, but it would be able to provide more current than T2P since it is coming from the transformer aux winding instead of the T2P pin. I guess it depends on how much current you need to drive the FET. 

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

     

    Regards, 

     

    Michael P.

    Applications Engineer

    Texas Instruments 

  • Hello Michael,

    using the VC voltage is a brillant idea - thank you so much.

    The MOSFET does not need any current, just voltage on the gate. So this is an easy solution which doesn't even need a zener. 10 stars and cudos to you!

    Best regards,

    Michael