• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 63 subscribers
  • Views 2933 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.

PD with no data passthrough

VforVendetta
Prodigy 170 points
Other Parts Discussed in Thread: TPS23753A

I'm using a TPS23753A to design a PD with no need for data passthrough. The device will be a "dumb" dead end device that just acquires power through PoE.

Reference the schematic snippet below from PMP7298:

My question is this: If I have no need to pass data back out can I eliminate the portion in the red box? I'd like to get rid of T1 (obviously J1 can go away). However the PoE spec mandates that power can come in over either pairs 1&2 and 3&6 or over pairs 4&5 and 7&8. This means I still need to tie pairs 1&2 and 3&6 into the circuit in case power comes in over them. If I don't need the data from them can I just tie 1&2 together over to one side of D3 and tie 3&6 together to the other side of D3? Basically use the same approach that is used with pairs 4&5 and 7&8.

Thanks in advance!

V

  • 0
    TI__Mastermind 40120 points
    Hi V,
    This is currently being looked into. An answer will be posted when available. Thanks!
  • 0 in reply to Darwin Fernandez
    Prodigy 170 points

    Hi Darwin,

    Sounds great. Thanks for looking into my question. I've been doing more research. SLVU318B provides a reference design for another of TI's PoE PD chips. In this design all four ethernet pairs are run through a transformer, the center tap of which is connected to a diode bridge.

    The spare pairs (4&5 and 7&8) may or may not contain data and/or power. It appears this pair can be run through a transformer (as seen in SLVU318B) or directly shorted together (as seen in PMP7298). The primary pairs (1&2 and 3&6) may or may not contain data and/or power. This is the same possibilities as the spare pair. If the spare pair can be shorted, the primary pair should be able to be shorted as well. Unfortunately what I just presented was a logical argument, not an electrical one. If anybody can confirm/deny this based on electronics knowledge that would be great. I'll probably just wire it as proposed (short pairs 1&2 and 3&6 together then feed them into a diode bridge) and see what happens. I'll report on this after I try it.

    V

  • 0 in reply to VforVendetta
    TI__Mastermind 40120 points
    Hi V,

    The current IEEE802.3at standard allows to send data through both the data and spare pairs which is why each pair are connected to the data transformer.

    The application of not using the PD reference designs/EVMs and omitting the data transformers because no data will be sent will still work and allow power to be sent through the Ethernet cable. There should be no issues. However note that this will not be considered IEEE compliant as there is no way to implement LLDP (power negotiation through data lines).
  • 0 in reply to Darwin Fernandez
    Prodigy 170 points
    Hi Darwin,

    Thanks for getting back to me. Sorry for disappearing for so long. Do you have any app notes or reference info on LLDP?

    I thought PoE functioned solely using the classification resistor to declare PD's power requirements to the PSE. What am I missing? As I mentioned before, this application is for a dumb dead-end device. All it does is acquire power through PoE. No data is used or passed through.

    Thanks,
    V
  • 0 in reply to VforVendetta
    TI__Mastermind 40120 points
    Hi V,

    After looking at the beginning of this thread, I notice you are looking at 13W of power? If this is the case, LLDP can used for Type 2 (25.5W) applications. If Type 1 (13W), then LLDP is not needed.

    Regards,
    Darwin
  • 0 in reply to Darwin Fernandez
    Prodigy 170 points

    Cool!

    Thanks Darwin!