• State Resolved
  • Locked Locked
  • Replies 5 replies
  • Answers 2 answers
  • Subscribers 62 subscribers
  • Views 503 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.

Original question:

TPS23730: POE Class 3 24V flyback

TPS23730: 24V 40W( pd class 5) or 51W(pd class 6) design

d_zero
Genius 4430 points
Part Number: TPS23730
Other Parts Discussed in Thread: PMP20878, , TPS2373, UCC24612

Dear all,

1)

would you recommend one of following transformers for the 40W and/or 51W , 24V flyback design?

a) Wurth 750343576 

b)  POE40Q-24ED , POE51Q-24ED 

2)  Would you recommend a Flayback design or a Forward for 24V ( 40W and/or 51W) ?

In previous replay you recommended Diode flyback for 24V is this also true when currents are greater then 1.5A? Because PMP20878.

3) What is the recommended controller mode if flyback is used CCM or DCM?

Best Regards,

David.

  • 0
    TI__Mastermind 21320 points

    Hey David, 

    I hope you are doing well, good to hear from you again.

    I would suggest picking the topology first, and then the transformer. 

    Either a flyback or an active clamp forward (ACF) would be fine with this output voltage/current. I think the typical design trade-offs are present: a diode flyback is going to be simpler, smaller, lower cost but lower performance. The ACF will have better performance but higher cost and more physical space. 

    I think looking at the reference designs already done will give you an idea of what performance will look like. 

    With the TPS23730, you can do a primary side regulation (PSR) flyback which will enable even lower cost and space savings. The TPS23730 can do either PSR or traditional optocoupler-feedback. 

    For the transformers, the Wurth one looks to be rated for 72W. Flyback and ACF topologies typically peak in efficiency at full load -- so since this transformer will only ever be at 50%-70%, your efficiency will take a hit. You can look at the efficiency curve in PMP20878 and find the output load at 51W to see the difference. 

    So I would recommend the Coil-craft parts here. 

    Note, if you want to consider PSR, you would need a PSR transformer. 

    For CCM vs DCM, this is a hotly contested debate among vendors and designers. The PoE Team at TI believes CCM is the best option for the typical PoE application because CCM provides a better load transient response and higher efficiencies. Our controllers are intended to run in CCM, but you can design the DCDC in a way where it will run in DCM. Other IC vendors intend their ICs to be run in DCM. Additionally, TI offers many DCDC controllers, so you could get a standalone PoE PD (like the TPS2373) and pair it with a DCDC controller that best suits your needs.

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

     

    Regards, 

     

    Michael P.

    Applications Engineer

    Texas Instruments 

  • 0 in reply to Michael Pahl
    Genius 4430 points

    Hi Michael,

    yes we are doing well, thank you for the fast replay :).

    1) Topology

    "I would suggest picking the topology first, and then the transformer. "

    I see that the active clamp forward has better efficiency, but it also has much more components that PSR flyback.

    I know in the past you suggested not to use active clap flyback because of component count , but i think it is still less components than active clam Forward ( eg. SLUUC91A) .

    Because we are space critical i would like to go with PSR Flyback, also here are some points that i think can improve efficiency and not increase component count so much, can you make your comments?

    a) Add active clamp to PSR flyback ( one additional PMOS connected to GAT2) 

    b) Add self rectifying mosfet on secondary side with UCC24612

    2)Transformer

    I' looking at POE40Q-24ED and POE51Q-24ED, they are pin compatible and i can test them both on same PCB.

    I think they are good for PSR flyback because they have AUX winding ?

    3) CCM

    As you said you think that CCM is better, and your controller TPS23730 is intended to run in CCM, i would like to take this approach.

    But how can i check if the 22uH or 27uH inductance of the coilcraft transformer will bi ok that the controller works in CCM ?

    Also if i do compensation for CCM at full-load, when the converter goes to light load condition is it possible that the converter then goes into DCM and become unstable?

    Sorry to maybe ask this again, are you aware of some CCM flyback design note from TI?

    I tried to do a calculation and i think with this trafo we can make max duty cycle for CCM 40%, also if we fall below 15W we go to DCM, and i would like to have enter DCM if we fall below 5W.

    Trafo.xlsx

    Best Regards,

    David.

  • 0 in reply to d_zero
    TI__Mastermind 21320 points

    Hello David,

    1. a.We don't recommend active clamp flybacks

    b. making it a synchronous flyback will be difficult at 24V. The VDS rating of the FET will need to be 200V at least. You will also need a snubber that will cost efficiency. You will need a gate driver like you mentioned. Even with all of this, we have never seen a sync flyback at 24V, so I can't confirm it works. 

    2. These transformers cannot be used in PSR. PSR is dependent on the transformer construction method, so if it is not specifically made for PSR then it will have poor regulation in a PSR design.

    3. The following link provides training on flybacks. "understand the basics of flybacks" talks about DCM and CCM. Many of our diode flybacks slip into DCM at light loads and that is normal and acceptable. The regulation is not as tight in DCM, so yes the output can rise a few volts at light loads. The TPS23730 does support traditional optocoupler feedback if tight regulation is required at light loads (typically <10%).  This cannot happen with sync flybacks. 


     

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

     

    Regards, 

     

    Michael P.

    Applications Engineer

    Texas Instruments 

  • 0 in reply to Michael Pahl
    Genius 4430 points

    Hi Michael,

    thanks for answers.

    One thing, how can i now or check if a transformer is made for PSR Flyback?

    How did you know that these Coilcraft transformers are not for PSR flyback?

    Best Regards,

    David.

  • 0 in reply to d_zero
    TI__Mastermind 21320 points

    Hello David, 

    We at TI have defined all PSR transformers used in our designs, so I know which ones are. 

    The vendor will have to specify that it is made for PSR designs. 

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

     

    Regards, 

     

    Michael P.

    Applications Engineer

    Texas Instruments