• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Subscribers 41 subscribers
  • Views 349 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

PMP22301: PMP22301 adaptation to an ethernet SBC platform

Manu Abraham
Genius 3510 points
Part Number: PMP22301
Other Parts Discussed in Thread: TPS23754, , TPS23751, UCC28630, PMP12027, UCC28740

Greetings,

With a thought to integrate Power over Ethernet with a  SBC with Ethernet capability, came across the TPS23754 (PMP22301).

I have 2 questions at this point.

1. I need to integrate the PoE on to the same SBC. As a result, we need only a single ethernet connector on the final SBC board.

How do we integrate both together (The SBC has an Ethernet transformer, the PMP22301 has an Ethernet transformer: 7490220121) ?

I guess we could likely merge both functionality into a single transformer ?

2. This aspect is about the Flyback transformer. as per advice provided by TI, I have contacted Wurth for the transformer.

Yet to receive a reply from them. In the meantime, I have been looking for an alternate flyback transformer.

The closest (I guess it appears nearly identical ?) that I can come across is PH9006NL from Pulse Electronics.

Can this transformer (PH9006NL) be used in the PMP22301 derived design at our end, instead of the  750318587 from Wurth ?

(I am not talking about the physical layout, since we need to design our own board for integration purposes)

Thanks,

Manu

  • 0
    TI__Intellectual 2120 points

    Manu,

    1.  The RJ45 connector on PMP22301 has both data and power on it.  The power is sent to the flyback converter through the winding center-taps of the PoE transformer.  The data is available on the secondary of the PoE transformer.  Yes, you can use one RJ45.

    2.  The PMP22301 flyback converter operates with a 24VDC adapter input or PoE input (18-57VDC).  Do you need this wider range or just the PoE input (42.5-57VDC)?  The Wurth transformer is designed to operate over the wider input range in continuous conduction mode (CCM) and is built on an EFD20 core/bobbin.

    The Pulse xfmr is designed for a 35-80V input and discontinuous conduction mode (DCM), so it will meet the PoE input range, but not the 24V adapter.  25W is pushing the power limit of an EP13 core, but since the operation is DCM it might be ok.  It can probably be made to work in the PMP22301 design with some changes.  The peak/rms currents will be higher so the sense resistor (R17) value needs to change.  More output capacitors may be need to provide an acceptable voltage ripple level.  It looks like Q1 and D10 are ok as is.  The loop compensation will need some adjusting.

    If you just need the PoE input range, take a look at PMP21655 attached.  It uses the TPS23751, which is very similar to the TPS23754.  The output is 23V/1A, but can be changed to 24V by adjusting the value of R37.

    Thanks,

    David

    PMP21655A(001)_Sch.PDFPMP21655A(001)_TI-BOM.xlsPMP21655_RevA_Test Report.doc

  • 0 in reply to David Strasser
    Genius 3510 points

    Hi David,

    Thanks for the response.

    Regarding #1. That solves the issue;

    As you suggested that I can connect the ethernet phy directly to the data on the secondary of the PoE transformer. That's good news.

    Regarding #2.

    As an alternate powering method, using an AC adapter was also in the idea; I've already have at hand a UCC28630 based flyback with a 24V output, tested with another project. Based on 2 aspects the PMP22301 was a good choice to start with. But the whole issue has been with WE. Previously, had talks with the WE department handling chokes and they were quite responsive. They did send some samples too and quite fast! The transformer division appears to possess exceptional contrast. They have not responded at all. If an organization does not respond to basic business queries, for anything else it would be much more worse.

    It is for these grounds, I do really appreciate TI for setting good standards, for others to follow. Much appreciated !

    Unless there is an alternative for the flyback transformer, PMP22301 is a non-starter. Pulse seemed like all ready to go, which looked good, hence the previous query with the Pulse transformer.

    As you suggested, I had a look at PMP21655. It looks okay, till I need to alternately power with an AC adapter; A UCC28630 solution that I have at hand is no longer good enough. But a search brought me to the UCC28740 based PMP12027. Which looks a bit underrated, but might be kind of okay ?, but the AC adapter needs to be redone. Is there a cost effective AC adapter solution that comes to your mind, to go along with the PMP21655 based solution, other than the PMP12027 ?

    Thanks,

    Manu

  • 0 in reply to Manu Abraham
    TI__Intellectual 2120 points

    Manu,

    Can you OR the 24V outputs together instead of OR-ing at the input?  Then you can use your existing 24V offline solution and PMP21655.  I forgot to mention that the Pulse PA3855.003NL transformer can be used in PMP21655.  If you want adapter priority application note SLVA306A shows how to split the DEN detection resistor to disable the PoE.

    I'll look for an alternate to PMP12027 in case the above does not meet your needs.

    Thanks,

    David

  • 0 in reply to David Strasser
    Genius 3510 points

    Hi David,

    The 3855 transformer is a good choice, available off the shelf. Saved me from making email request to the transformer vendors. Thanks!! I was thinking of adapter priority, somehow you managed to read my mind. :-) DEN as with Option#3 was what was lingering and toying in my head. You saved another possible future post!

    The PMP12027 outputs 0.625A and hence slightly underrated, but is way more cost effective. It still is a viable alternative. The available solution at hand provides around 3A, though unnecessary and overrated. But since it already exists; Unless, there is the need to cut costs, it is a better choice AFAICT. Depending on how the project moves, will decide between existing solution, PMP12027 or a slightly more capable design.

    Since you mentioned output OR'ing, it allows me to use the existing solution at hand, which is good! Thanks for the pointers. Really appreciate all the pointers.

    Thanks Again,

    Manu