• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 64 subscribers
  • Views 334 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.

Original question:

LM5156-Q1: SEPIC converter - MOSFET overheating

LM51561-Q1: LM5156-Q1: SEPIC converter - MOSFET and Inductor overheating

Wagner Mourthé
Prodigy 40 points

Part Number: LM51561-Q1

Hello!

On my last post (https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1521807/lm5156-q1-sepic-converter---mosfet-overheating/5850758) I shared an issue I had with a SEPIC topology using the LM51561-Q1 on which the MOSFET was overheating beyond what was expected.

When we tried a MOSFET with smaller gate capacitance the result was better, even though the on Resistance was bigger. At the time I reached the conclusion that the bigger capacitance on the gate was making the turn-on time too slow, and the MOSFET was overheating because of that.

Now, we were able to test other MOSFETs and the results are shown below:

image.png

  • BUK7Y7R8 (Original)
  • BUK7M17-80EX (Last Post substitute)
  • PSMN041-80YL (New)
  • BUK7Y38-100E (New)

image.png

Yet again, the best performing MOSFET was the one with the smaller capacitance.

We noticed that the inductor is also overheating, even though it is rated for 5.23A rms and we are using a 2A load, so we decided to investigate further.

The picture below shows the waveforms for a 2A load:

image.png

Yellow - MOSFET Drain
Purple - MOSFET Source
Blue - MOSFET Gate
Green - Diode Anode

We believe this oscillation on the MOSFET drain is the main reason the MOSFET and inductor are overheating. We tried adding gate resistance, but if we slow it down too much, it also overheats because it stays too long on the linear region. 

Do you think this oscillation may be the reason for the overheating issue? Can a snubber across Vds solve it? or should it be from Vd to GND? Also, how does it relate to the MOSFETs we tested and the temperature results? is it really a gate capacitance issue? 

Thank you in advance.

Wagner M.

  • 0
    TI__Mastermind 22935 points

    Hi Wagner M.,

    Wagner Mourthé said:
    Yet again, the best performing MOSFET was the one with the smaller capacitance.

    Yes, losses in MOSFETs depend a lot on the switching frequency. For fast switching, the gate charge is more important than the RDSon.

    Good that you found a FET that fits. 

    Wagner Mourthé said:
    We noticed that the inductor is also overheating, even though it is rated for 5.23A rms and we are using a 2A load, so we decided to investigate further.

    Inductor L1 will see much more RMS current than the output. The lower the input voltage, the higher this current will be. Please check in the power stage designer:https://www.ti.com/tool/POWERSTAGE-DESIGNER 
    It is possible that the inductor gets hot because of overloading. 

    Wagner Mourthé said:
    Do you think this oscillation may be the reason for the overheating issue?

    I think the overtemperature is mostly related to the high switching frequency and the most likely underrated inductor. Also a higher amount of copper in the PCB will help with getting the heat away. You can still try a snubber or improve the layout to get rid of the ringing. Also you could think about switching slower.

    Best regards

    Moritz

  • 0 in reply to Moritz
    Prodigy 40 points

    Hi, Moritz

    When using the coupled inductor on the Power Stage Designer, on the inductance, should I use the series value on the inductor datasheet?

    For example, I used the SRF1280A-3R3Y on my design. Which parameters should I use, the parallel or series one? I considered the Series rating



    Using 13.2uH as the inductance value, the max current on the inductor L1 is 3.28A and on L2 is 2.19A

    Should I add both these currents and compare to the Irms series rating on the datasheet? Would that be the correct way to predict my inductor temperature?

  • 0 in reply to Wagner Mourthé
    TI__Mastermind 34240 points

    Hello Wagner,

    Moritz is currently out on sick leave, so I will take over in the meantime.

    The inductor datasheet shows schematics for the definition of series and parallel connections.

    As you use a SEPIC topology, you use N1 for the primary side and N2 for the secondary side, so I would say it is neither parallel, nor series connection.
    The inductance for a use in SEPIC topology should be something between 3.3uH and 13.2uH, maybe even in the middle at 8uH.
    It is important that the calculator peak current is still below the saturation limit of the inductor, but this should be the case for your design.

    Best regards,
    Niklas