100BASE-T1 PoDL Coupled inductor

There is a Wurth group on the forum if you want the manufacturer to chime in...

 Würth Elektronik 

It is a bit odd that there doesn't seem to be a TI group.

Hi Taip ,

thank you for your question. I will forward your question to a colleague and get back to you with an answer.

Best regards,

Nicola

Würth Elektronik

Hi Taip ,

we made serveral tests and simulation regarding 100Base-T with PoDL. To fulfill the limits of the 802.3bw (mode conversion & return loss)  we recommend this frontend setup as below.

But we also investigated, that the inductance value of L2 can also be lower than 100uH. But we cannot recommend lower than 47uH.

Suitable double chokes for 500mA are:

300nH à 744894300033 Datasheet WE-TDC SMT Shielded Coupled Inductor (we-online.com) or 74479762133 Datasheet WE-PMI Power Multilayer Inductor (we-online.com)

47uH à 744878470 Datasheet WE-DD SMT Shielded Coupled Inductor (we-online.com)

100uH à 744877101 Datasheet WE-DD SMT Shielded Coupled Inductor (we-online.com)

Best Regards

Nicola,

Thank you very much for the details. Please help me understand the notations of the circuit. Are the TX+/- on the left from the PHY, or are they from the far end T1 line with PoDL?  L1 (10uH) in the Evals of both Microchip's and Ti's are 200uH, do I need to change htem to 10uH for it to work with L2 (100uH)? Can I use the same component values on both PSE and PD?

Thank you very much for your help.

Best regards,

Tai

Hi Taip ,

the left side is directed tot he PHY. We have different inductance values of the CMC because we don’t use the 100nF but the transformator T1 for to achieve the 1,5kV insulation which is required for industrial applications. The T1 itself provide extra common mode rejection, therefore the CMC can be smaller. This solution is more elegant than using 100nF/2kV MLCCs .

If you only need the 50V 100nF MLCC like in automotive applications it´s a different story.

Usually you can use the same circuit for PSE and PD

Nicola,

Thank you very much for the clarification. I will order the parts and wire up the prototypes. Will let you know once I have it done.

Regards,

Tai

Nicola,

Actually, since this is a closed, engineered system, we don't need to stay with the 1.5kV isolation. Do you have suggestions on components if we want to stay with the simple 50V 100nF MLCC?

Thanks,

Tai

Hi Taip ,

you can use the 885012206095 in the size 0603 or if you want to have a robust component the 885012207098 as 0805.

Nicola, 

I am sorry I did not ask my question clearly. What I want to know is if I want to stay with the 100nF capacitive coupling - No T1. Can I still use the same L1 and L2 suggested in this table? 

Hi Taip,

if you want to use the 100nF/50V you should stay at the recommendations in the TI Ref Design. Sure with your lower current, we must adjust the size of the components:

CMC(no PoDL current through it): 744235251 Datasheet WE-CNSW SMT Common Mode Line Filter (we-online.com)

6,8µH: 7448841068 Datasheet WE-DPC HV SMT Dual Powerchoke (we-online.com) or  7448844068 Datasheet WE-DPC SMT Dual Powerchoke (we-online.com)

330nH: 744894300033 Datasheet WE-TDC SMT Shielded Coupled Inductor (we-online.com) or 2x 74479762133 Datasheet WE-PMI Power Multilayer Inductor (we-online.com)

TVS(super low capacitance): 824012823 Datasheet WE-TVS TVS Diode – Super Speed Series (we-online.com)