XIAOs faulty/dying

[omeroguz45]

I bought 7 XIAOs and 3 of them died.

On march 13, I ordered 3x nrf52840, 3x rp2040 and 1x ra4m1 xiaos directly from seeed. I had soldered the headers like a week ago, and now (apr 10) i had the chance to sit down and try them.

I was going one by one with the nrf52840s first, and the third one I plugged completely turned off after like 10s, and now it only shows the charging led when usb is connected. one of them worked with no issues. but the other, always had issues with it’s usb, connecting and disconnecting in a repeating pattern, unless it was in bootloader mode. i’ve tried replacing the firmware, with the one from the seeed repo, and one from adafruit, but the usb was always hit or miss. then sometime later, i realized that the rgb led’s red led was lightly on, and was fluctuating when i was trying to upload firmware/code. i couldn’t turn it off, the other (b,g) were working with no issues. as i tried throughout the day, it started to stay disconnected more, and now it won’t connect pretty much.

while i was working with the nordic xiaos, and i was writing the issue on the discord, i decided to check the rp2040s.

back when i had soldered them i’ve had two of the rp2040s to be hit or miss with the usb connection. i was at the lab (where i had soldered the headers), and i didn’t have a proper type c cable with me, so i decided it must be the cable and haven’t touched them since. now, when i had decided to check the rp2040s, i’ve had usb connection issues with two of them again. i left one connected, and went to write about my issue on discord, then i heard a crackling sound, saw the led lights dying on the board, and immediately unplugged it. it was very hot to the touch. it’s what prompted me to write about this on the forum. i’ve already contacted seeed support when i had issues with the first nordic board, but now the issue is bigger as 3/7 boards are pretty much dead.

i’ve ruled out short circuits in the soldered headers; i had soldered them with a high end jbc station, under a microscope, and i checked twice after each board for shorts. i now checked again with my bare eyes, and i still cannot see a short. first chance i get, i’ll test the headers for shorts, but i doubt it.

there were no issues with the packaging when i had received them.
can anyone help?

[PJ_Glasso]

Hi there,

And welcome here…

SO , can you provide a picture..? what is the environment WIndows, Arduino IDE ?
The Cable Matters and NOT all USB-C cables are alike.

I always test them without touching them(soldering) initially, over 100pcs. never a failure. FWIW

What were the settings on this

@omeroguz45

a high end jbc station

They do over cook easily, so Beware. (the traces/pads) will lift off
The dim LED is a dead giveaway, short…

HTH
GL PJ

If you are seeing windows pop ups USB bad device etc. check the drivers you are using, Zadig maybe needed.

[omeroguz45]

nrf52840-board-2-images.zip (783.1 KB)

rp2040-images.zip (1.5 MB)

nrf52840-board-1-images.zip (890.6 KB)

The images of the respective boards are attached as .zip files.

I’ve used three type c to type a usb 2.0 spec cables that I’ve used for years with no issue for many devices. The cables are not the issue.

I’ve checked, none of the boards are shorted between the pins, housing or usb c connector.

RP2040 board is dead. Just heats up, up to 70C. Internal short.

The nordic board that was weird with the usb and had it’s red led be always on dimly:

It’s dim led issue seems to be gone. But might come up after longer testing, as the previous dim light was an indicator of an internal short, and it might come up again with heat. It does not get recognized via usb unless it is in bootloader mode. No issues in the 3v3 and 5v power pins.

The nordic board that was non responsive with only the charging led on:

Still, the issue persists. Does not connect to usb at all, normally or in bootloader mode. Cannot even know if it goes into bootloader mode. Power pins have no issues, both 3v3 and 5v.

I’ve also attached my soldering settings as an image. JBC DDE, T245 iron, set to 300C, 150C sleep temp.

I’ll also post the images directly in following posts for the boards as there’s a 5 image limit.

[omeroguz45]

nrf52840 board no 1, turns on, usb issues.

[omeroguz45]

nrf52840 board no 2, charging led only, seems dead

[omeroguz45]

rp2040 board, dead and burns

[PJ_Glasso]

Hi there,

WOW, nice Work.

So I zoom in and I think the USB traces have some questionable solder flow there, comparing the two photos, The can on the RP looks off also? I would BAG up the whole LOT and Send them Directly to Seeed! , they look poor quality IMO.

I’ve only cooked one intentionally once as part of a Crash test. it still worked but non of the battery charging circuits survived, weather over voltage or direct short(internal charge chip is supposed) to handle that.

You brought the receipts my guy, the thermals’ NEVER LIE Sweet.

Find the picture of the Naked Xiao on here( no @cgwaltney ) not you
and see which chip is on that side , this will further your findings maybe ,

It’s a hard find , but seeed needs to know if there’s a production problem coming or what , for ONE guy to get all these bad devices, That’s an Outlier AFAIK. never happened before. (one maybe) chalk it up to new user kinda thing. THAT is NOT this!

I would request a pre-soldered unit and a plane one. (retest) is the only way forward I feel that gives any kind of confidence in the hardware.

well that my swing and bunt,

HTH
GL PJ

I compared the two lower USB-C backside photos ,soldering traces, one is clean the other not so much. Hmmm.
please keep us posted.

[cgwaltney]

Ask AI to generate the Naked XIAO i hope it is not Sharkface

[omeroguz45]

The closeup of the USB connector pads of the boards.
I’ve already contacted seeed, talking with them.

Thanks for the help!

[omeroguz45]

Talked with seeed and they are sending replacements.

@PJ_Glasso Thanks for the help!