• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 2 answers
  • Subscribers 63 subscribers
  • Views 1815 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.

TPS2375: Power output is cycling

AQ
Expert 1580 points
Part Number: TPS2375
Other Parts Discussed in Thread: TPS2376HEVM, , , TPS2372

I designed a PoE PD board using the TPS2375PWR-1 device.  When providing a PoE source the output power cycles, as shown in the attached scope trace.

I have the TPS2376HEVM eval board and when I connect the same PoE source it works fine, so I don't think there is a problem with the source (even though the eval boards uses a slightly different version of the PD chip).  The design is based on the example in the datasheet and I tried modifying some values but I always get basically the same result.  The power comes on for about 0.5s, and then goes off again, repeating every ~3s.

  • 0
    TI__Genius 9230 points
    Hi AQ,

    Can you described what the signals are on each of your channels? If CH1 is Vout, are you designing for 57Vout?

    Did you say your design is based on TPS2376HEVM or TPS2375EVM? Could you post a schematic to show the changes you made?

    Thanks,
    Tom
  • 0 in reply to Thomas Amlee
    Expert 1580 points

    I have attached the board schematic below.  I had designed it based on the TPS2375-1 datasheet.  The scope trace had Ch1 between +Vin and -Vin on the DC/DC converter. CH2 was not connected.

    POE_VP and POE_VN is supplied directly from an RJ-45 jack with built in magnetics and bridge rectifiers.  https://www.digikey.com/products/en?keywords=732-4974-ND.

    I noticed that the design was missing the input capacitor, so I added a 1uF cap in parallel with D10 but this didn't change the result.  I tried other values also but it didn't stabilize.  I confirmed the PTC fuse F3 wasn't cycling (I shorted it).  I tried different class setting resistors and ILIM resistors.  I increased the value of C38 to 22uF.  This caused the decay to last longer, which increased the cycle period, but it still would shut down after about 500 ms.

    I also tested with and without the DC/DC converter populated.

  • 0 in reply to AQ
    TI__Genius 9230 points
    Hi AQ,

    Try removing C37 & seeing if it works properly. That isolation cap should be connected from +Vin (U12) to -Vout (U12).

    - Tom
  • 0 in reply to Thomas Amlee
    Expert 1580 points

    Thanks Tom for the suggestion.  I tried that but unfortunately it didn't change the operation.

  • 0 in reply to AQ
    TI__Genius 9230 points
    Hi AQ,

    I saw the datasheet for the Meanwell part, so you can keep C37. I had not seen that capacitor before in our flyback designs.

    It appears that you are in inrush, it is taking too long (500ms) & get hot, the part goes into thermal shutdown, & the PD controller attempts inrush again.

    I suggest that you make sure the converter is turned off until after inrush is completed (PG goes open drain) so that the bulk cap can fully charge.

    Thanks,
    Tom
  • 0 in reply to Thomas Amlee
    Expert 1580 points

    Thanks Tom.  I suspected the inrush might be a concern, but I removed the DC/DC converter and I still saw the same behavior.  The 2.2uF caps and inductor were still there, but I didn't think that would cause significant inrush.

    I'm wondering if it's related to the MPS (maintain power signature).  The datasheet shows I need a minimum load of 10mA and specifies a maximum impedance.  I tried adding some load on the 5V output and it didn't help, but it was probably far too low.  I assume the 10mA spec refers to the high voltage side, so I would need to draw something like 150mA at 5V.  This circuit feeds a battery charger, so not a constant load, and eventually it would drop well below 150mA when the battery is near or at full charge.  It seems strange that I would need to impose such a high fixed load just to keep this chip running.  I'm misinterpreting something?

  • 0 in reply to AQ
    Expert 1580 points

    It turns out the issue was the MPS and not having sufficient load.  I added a 10k 3W resistor to the input to the DC-DC converter and it stays on now.  Because the system will have to run at low power at times (the PoE charges a battery) I will probably change to TPS2372 or something similar with auto-MPS pulsing.