• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 64 subscribers
  • Views 340 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.

LM25116: Inductor ringing/heating up

Neehar Shah
Prodigy 30 points
Part Number: LM25116

Hello,

I currently use the LM25116MH in our designs extensively as our DC/DC buck converters.
We use 2x 6S LiPo batteries (nominal 24V, max is 26V fully charged each with 4500mAh capacity) as our power source and we use this controller to step-down to 5V nominal.
For this particular design for which I am having issues with, we actually had designed this buck converter to output 5V with a max of 7A on a previous revision and it worked perfectly fine (i.e. heat dissipation was nominal and handled well, and good line regulation)
Due to issues with an on-board USB hub controller, we found that filtering the buck converter with an RC circuit and some value changes reduced the ringing that we suspected was causing our USB issues.
We created a new revision a few days ago trying to incorporate the changes and make PCB layout changes that were more closely following the evaluation board layout.
Another notable change is the SW copper area which is much less than before with no real thermal vias for dissipation other than through the MOSFET pad themselves. I read an article for how to reduce power supply noise and one thing that was mentioned was to reduce the SW node copper area.

When we first turned it on with about a 2A load on the 5V line, the inductor heats up significantly almost immediately. About a minute of operation, the inductor starts audibly ringing. In less than 2 minute, the inductor gets to about 125degC and eventually shuts the 5V line down.

Since that issue, I have updated the error compensation network to match the WEBENCH results and used another inductor which has a higher current rating but the same inductance value, but still see the same results.

I was wondering if anyone can help guide me into rectifying this issue? Measurements that might be useful to know or anything else.

I have attached a screenshot of the schematic and a PDF of the PCB layout. The only values in the schematic that have changed are the error compensation network components, all other component values match the schematic below.

For the PCB:
Top Layer - GND Plane
Inner Layer 1 - GND Plane
Inner Layer 2 - Power Plane
Bottom Layer - GND Plane

RADIO_CARRIER_UNIT_RevC1.pdf

  • 0
    TI__Guru 51971 points

    Hello Neehar 

    Would you please forward your Quick Start excel file with your component selections https://www.ti.com/lit/zip/snvu051 , and measure( inductor current, SW , VCC, SS, UVLO, EN) for more investigation ? 

    Meanwhile, please try these items. 

    • The CSG pin shouldn't be directly connected to the PGND pin. It should be directly connected to the ground connection of R5
    • Ground the VCCX pin if not used
    • Retest after changing C6 to 0.47uF
    • Check if any component touching the inductor. 

    - Eric Lee

  • 0 in reply to EricLee
    Prodigy 30 points

    Hi Eric,

    I appreciate you taking the time to help me out with this issue.

    Please see attached quick start excel document that you attached to this post completed. The bill of materials page also has the components that we had selected for this application.

    As for the measurements, please give me by the end of the week to give you some results. We don't have a current probe so getting the inductor current would be difficult.

    - Neehar Shah

    LM5116_quickstart (version 1) Updated.xls

  • 0 in reply to Neehar Shah
    TI__Guru 51971 points

    Hello Neehar 

    Thank you for sharing the file. When you have a chance, please try Rcomp = 4.3 kOhm, Cccomp=33 nF, Chf = 270 pF and check if any improvement . 

    - Eric Lee 

  • 0 in reply to EricLee
    Prodigy 30 points

    Hi Eric,

    I did a couple things:
    - I now have CSG connected directly to the sense resistor
    - VCCX is now grounded
    - I'm still waiting on parts to arrive so I could not use a 0.47uF capacitor (I should be able to do this today)
    - Compensation network that you suggest I should also be able to do today once the parts show up

    The ringing and heating up are still an issue.

    These were the results:
    VCC - 7.3V
    EN - Starts at 12.5V and slowly creeps up in voltage
    SS PinThis is the SS pin

    This is the SW node on the controller

    This is the UVLO pin

    - Neehar

  • 0 in reply to Neehar Shah
    TI__Guru 51971 points

    Hello Neehar

    By decreasing Rcomp and increasing Ccomp together, you can reduce the crossover frequency a decade (Try 430 Ohm and 330 nF).

    If the system is still unstable after reducing the crossover frequency, I think you have to measure inductor current for more investigation.

    -  Eric Lee

  • 0 in reply to EricLee
    Prodigy 30 points

    Hi Eric,

    Rcomp = 4.3 kOhm, Cccomp=33 nF, Chf = 270 pF

    This seems to have worked, the inductor no longer heats up and there is no more ringing.
    I did a load test and all works ok at a 7A draw.

    I followed the WEBENCH results for compensation network, I'm wondering how you came to these values?
    If WEBENCH isn't the best tool to use for compensation network, what is a good tool to use?
    We use the LM25116 in many parts of our system and if there is a better way to calculate these values, some guidance would be appreciated!

    Regards,
    Neehar

  • 0 in reply to Neehar Shah
    TI__Guru 51971 points

    Hello Neehar

    I used the Quick Start calculator which you sent, but reduced the crossover frequency from 25kHz to 12kHz (which is 1/20 of Fsw) and put a de-rated COUT value 150uF ( which is a half of 300uF). 

    - Eric Lee