2
\$\begingroup\$

In my boost converter I am seeing the inductor burn after some uses. I am still completing some bench testing but I am hoping my theory of why it is burning is correct.

My theory is since this MOSFET can handle only 4A, during the turn-on (short), the current can be as high as 128A. 5V/0.039 = 128A. How is it possible for that MOSFET to handle that over time? I will assume it will fail too eventually and the inductor too. Please let me know if you questions.

VCC is 5V. Load current is 1A.

The inductor use is XAL5050-223MEC, the DC/DC switch controller is LM3478, and the MOSFET is FDS3512 . Snip it of my schematic

Here is the layout. Green is the ground right underneath the top layer.enter image description here

\$\endgroup\$
6
  • 1
    \$\begingroup\$ You're ignoring the effect of the inductor. The FET is never turned on long enough for the current to get that high, unless something has gone very wrong with the controller. It's strange for the inductor to be what dies; show us your PCB layout. You may have a control issue. \$\endgroup\$ Commented Nov 3, 2023 at 14:54
  • \$\begingroup\$ @Hearth I edited my question with they layout. \$\endgroup\$ Commented Nov 3, 2023 at 15:53
  • \$\begingroup\$ Yout R/C network at the COMP pin has very unusual values. Where do they come from? \$\endgroup\$ Commented Nov 3, 2023 at 16:07
  • \$\begingroup\$ You haven't said what the load current is for your testing. \$\endgroup\$ Commented Nov 3, 2023 at 16:36
  • \$\begingroup\$ @Jens I used the document "AN-1286 Compensation for the LM3478 Boost Controller" chrome-extension://efaidnbmnnnibpcajpcglclefindmkaj/ti.com/lit/an/snva067d/… \$\endgroup\$ Commented Nov 3, 2023 at 16:56

2 Answers 2

Reset to default
3
\$\begingroup\$

during the turn-on (short), the current can be as high as 128A. 5V/0.039 = 128A. How is it possible for that MOSFET to handle that over time?

This won't happen because the LM3478's current sense input has a threshold of 0.156 volts and, you have a 39 mΩ resistor for Isense (R12). Simple math reveals that the MOSFET peak current is 4.00 amps but, your inductor will saturate at 3.6 amps. Data sheet extract: -

enter image description here

In my boost converter I am seeing the inductor burn after some uses.

This will cause the inductor to get quite hot and may indeed fail. The MOSFET may get a little warm too. Personally I think you need a bigger inductor; one that doesn't saturate until 5 amps and can handle the RMS current passing through it.

That RMS current cannot be determined with the details you have given but, fixing the inductor saturation current may be good enough. Also take note of the Irms testing box in the data sheet extract above; they tested the inductor with a fair amount of copper (which will act as a heatsink).

More detail from the OP

The OP revealed that the load current is 1 amp and, looking at his circuit, it looks like it will be operating at 500 kHz. So, with an input voltage of 5 volts and an expected output voltage of 26 volts (1 amp) we can expect an inductor RMS current of greater than 5.2 amps and, considerably greater than the 2.5 amp limit: -

enter image description here

Image taken from my basic website. The image is a calculator for boost converters but, it assumes ideal components for the inductor, diode and, MOSFET. I fully expect that the RMS current may even approach 5.6 amps in a real circuit.

Also think about this: the DC resistance of the inductor is typically 0.091 Ω (see data sheet extract above) so, the \$I^2R\$ losses will be about \$5.6^2 \times 0.091\$ = 2.85 watts so, no wonder it gets hot!

Elephant in the room

The other problem is this: with a 4 amp current limit, the output of 26 volts and 1 amp (26 watts) can never be met. Think about it; if the input power is (at best) 5 volts and 4 amps (a power of 20 watts) then how can you possibly get 26 watts to the output?

\$\endgroup\$
1
  • 1
    \$\begingroup\$ thank you. I will rework my calculations and design. The elephant in the room, good one. \$\endgroup\$ Commented Nov 3, 2023 at 17:28
3
\$\begingroup\$

How is it possible for that MOSFET to handle that over time?

There are other ratings that need to be followed to prevent the part from dying. The part needs to stay below a max temp and also not exceeding voltage and power ratings.

So you can send 4A through it as long as you don't make it too hot and don't exceed the power rating. This means making sure the average power through the device does not exceed 2.5W and also making sure that the part does not reach over temp (175C)

enter image description here Source: FDS3512 datasheet

Make sure that you either use the mosfets that are in reference designs or simulate the design or do the math to make sure the controller doesn't drive the fet too hard.

\$\endgroup\$

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.