MK64FX1M0VLL12 USB internal LDO burned out

[Oliveira]

Hi all,

This project at https://easyeda.com/MARCODEOLI/te0123-v2 is running for about three years until now, and suddenly at one of the boards has been detected an exponential increase of mcu´s body temperature, nonetheless all peripherals seem fine but USB. Windows stop detecting COM port and no error is shown at Devices Manager. When the board is sourced with in circuit power which excludes USB power input I can see the temperature goes down, so, it looks like an internal LDO fault. Am I correct about this? The process to buy some mcu has started although this will take some time as you already know about silicon shortage.

Any clue of what may be the cause would be appreciated.

Marco 

[Oliveira]

Hi Gustavo,

Yes, there was one event. USB is used as a normal COM port to send / receive data logs, when these input came in a burst, sometimes this specific board loses USB config and I have to reset it. I thought the cause could be the buffer length, although other boards are working fine. I increased the buffer anyway, then send a long stream of data and noticed the Windows losing it. Started checking cables, plugs, tried another PC, even erased and re-flashed the MCU with no avail to just in the last minute noticed the over temperature.

Have you seen the circuit? I cloned some parts from freedom boards. USB input has ferrite beads, fuse resistor, capacitors, TVS diode, all MCU´s input pins have 100nF capacitors. I am designing a new board. Perhaps I might include some more protection?

Thank you. 

Marco

[gusarambula · Accepted solution]

Hello Marco,

The schematic looks okay, and the protection included should be adequate under normal circumstances. If anything, I think the rework to run the USB is not trivial. It is doable but given the size of the board and their location I would consider adding the USB connections by default.

Regards,
Gustavo

[Oliveira]

Hi Gustavo.

I´ll do that. Thank you.

Marco

[gusarambula]

Hello Marco,

The USB module is somewhat isolated from the other modules in that it uses the VREGIN (VBUS) input at 5V and can be unpowered if not being used. From the symptoms you are seeing it could be that there was electrical induced physical damage to the USB module, but determining this would require to make an analysis on the faulty part. Was there any event previous to this issue that could explain damage to the MCU?

Regards,
Gustavo