MP4541 Shutdown Issue at 48V Input, 5V Output, 0.5A Load – IC Damage

DC-DC Power Converters
1

Hi Technical Team,

I’m working on a project using the MP4541 step-down converter. The goal is to achieve a 5V output at 0.5A load with an input of 48V DC. However, I’ve been facing some serious challenges:

  • The circuit works fine at lower input voltages (around 30V DC).
  • When I increase the input voltage to 48V DC (the target input), the IC either shuts down or gets damaged after applying a load.

Circuit Specifications:

  • Input Voltage: 48V DC

  • Output Voltage: 5V DC

  • Load Current: 0.5A

  • Components Used:

    • Inductor: 47µH, 1A saturation (XAL5050-473)
    • Input Capacitor: 22µF, 75V ceramic (Murata GRM32 series)
    • Output Capacitor: 47µF, 10V ceramic (Murata GRM31 series)
    • Feedback Resistors: 40kΩ (R1) and 10kΩ (R2)

  • Has anyone faced similar issues with the MP4541 when operating at 48V input? If so, what was the solution?

  • Would a higher current-rated inductor (e.g., 2A) help prevent IC damage in this case?

  • Is adding a TVS diode at the input for spike suppression a common solution in such designs?

  • Are there any layout tips I should consider for minimizing noise and improving stability at higher input voltages?

2

what frequency is it running at? Slower will be worse from a inductor saturation perspective.

1 Like
3

XAL5050 series doesnt seem to go to 47uh are you sure about that?

4

Hi Shivam,

Thanks for reaching out and welcome to the forum. Sorry to hear you’re having challenges with the part.

Doing a quick google search, I could not find a 47uH inductor in the XAL5050-473. Looks like the XAL5050 series only goes up to 22uH, as jshannon mentioned. Is this a custom part for you? Doing a quick simulation, a 47uH inductor with the given parameters, should have a peak current of around 620 mA. Could you double check the inductor value and current ripple you are seeing?

Assuming you are using an XAL5050 series part (from my google search), would be a 4.7uH inductor (XAL5050-472). If it has a saturation current of 1A as you mentioned, the peak current at 48V input is closer to 2.5A which would definitely damage the component.

To answer your other questions:

  1. Yes, a higher rated inductor would help and add a layer of redundancy. Our evaluation module goes with a 1.4A saturation, 2A would be more than enough. A BOM for recommended components can be found in the evaluation board datasheet.

  2. What does your startup/transient behavior look like? With the values you have, the transient overshoots/inrush would not be my first concern.

  • Input capacitor rating of 75V could result in some capacitor value derating which could affect the performance of your component. The DC bias characteristic can be found in the capacitor datasheet. Typically, voltage rating should be around 2x the voltage seen at the capacitor to avoid derating. I would recommend a 100V rated capacitor if possible.
  1. For layout I can recommend following the guidelines in the datasheet. You can also refer to our evaluation board for layout guidance.
    image
    image960×784 96.5 KB
1 Like
5

Hi Stephen,
Thanks for your response.
Yes, I am using a 47µH inductor as an alternative to the one given in the reference design, but I am still facing the issue. Additionally, when I mount the circuit on a zero-size PCB, it works, but the inductor heats up.

6

It is running at approx. 410kHz frequency.
And I am using alternative inductor of 47uH .

7

I am attaching schematic of that particular circuit and layout.

image
image1693×498 71 KB

8

An inductor is allowed to have a temperature rise of 40C it may seem hot to you but be actually OK

DC-DC Optimizer | Coilcraft Is a super handy calculator and spits out a whole range of parts and their temperature rises.

It looks like you have an Aluminum Electrolytic cap as the input filter. I think if you look at the EVAL circuit or the datasheet they want a ceramic.

9

also what is the value of your bootstrap cap (C33) it should be 0.1uF bigger isn’t better

If there is an EVAL board I really suggest getting one to see how the experts do it.

10

Thanks for your valuable response.
What is the recommended soft-start capacitor value for handling a 0.5A load at a 48V input? Is there a specific formula or guideline to follow for calculating the optimal soft-start time based on load and input conditions? How can I connect it externally?

11

This part doesn’t seem to have soft start, I am talking about C33, it should be 0.1uF I don’t see a value on the schematic. Most likely not your problem, get an EVAL board and copy it as much as possible

12

Thanks for providing your schematic and layout. I have a few notes:

  1. Your enable is tied directly to Vin. We recommend using a pull up resistor of 100k to pull enable high. Enable should be a logic signal that senses voltage and should not take in current. At 30V it may be OK, but 48V would increase the input current seen at that pin.
  2. What is the purpose of R37? It is not needed.
  3. As JShannon mentioned, I would recommend ceramic capacitors for input and output. Electrolytics have higher ESR and can dissipate more heat. If ceramic capacitors are not available, a parallel combination will reduce ESR of the capacitor network.
  4. You can also add a feedforward capacitor to improve transient response, an 820pF capacitor connected across your top feedback resistor is recommended.

This part has an internal soft start; therefore an external soft start capacitor is not needed.

For layout:

  1. Your high current paths should be as close to the device as possible with short, direct, and wide traces or copper pours.
  2. You have a very large hot loop around the inductor/SW node. You want to keep those traces short and away from the feedback lines.
  3. The input and output traces can be much larger to accommodate the current, looks like ou have a good amount of space. Again, a wide, short trace should be used.
  4. If heat is a concern, add additional vias to dissipate heat through the board rather than just under the IC itself. Specifically, Vias around the GND pour.
  5. Try to replicate the EVAL/datasheet layout as it is optimized for this part

I agree with JShannon, obtaining an EVAL board will allow you to compare the operation of your design to a working design. From there you can swap out specific parts compared to your design to see if there are any differences and can give insight into the issue.

Here is a link to the evaluation board:
EV4541-N-00A | 80V, 0.8A, Synchronous Buck Converter Evaluation Board | MPS (monolithicpower.com)

Hope this helps!

13

Thanks to all for your response!