Hi everyone, I’m having an issue with an Arducam 108MP USB 3 Camera Evaluation Kit. It was working fine at first, but the USB port on the back has completely fallen off the board. When I first got it, the port was already a bit wobbly, but it worked so I didn’t worry about it. It ran without any problems on Friday, and then I found it broken the next morning. The camera had been shoved into the part of the device I use it in, so that probably didn’t help.
Thanks for posting. That really sounds like a physical/mechanical failure rather than a firmware or driver issue. Could you take a close look at the area where the port was mounted? I’d like to know whether the solder pads are still intact on the board, or whether anything lifted off with the connector. Pictures of the board and the detached port would be a big help before we go any further.
Sure, here are some photos:
Looking at it up close, the two square solder pads that were meant to hold the USB port in place look bad. It seems like the power/data connections were actually what was holding the port onto the camera. One of those connections from the port to the chip came away with the port itself, so the shield didn’t break — the trace tore off instead. I think the stiff USB cable that came with the kit put too much strain on a weak solder connection.
Thanks for the photos, that’s really useful. The key detail is that a power/data trace lifted off with the connector rather than the connector’s metal shell just cracking. That means the mechanical anchor pads weren’t doing their job and the signal traces were taking all the strain. Have you tried placing the port back onto the board and checking whether the camera is detected? Even seeing what the LEDs do while you hold it in place would tell us a lot.
I did try holding the port back in place by hand. When I do that, the red LED comes on and the green LED flashes on and off, but I can’t get any connection at the software level. I tested with the EXE from the Arducam GitHub repo and also with the Python code from GitHub, and neither one sees the camera. I don’t have any logs to share.
Okay, that’s a good data point. Red LED on plus green flashing means power is reaching the board while you’re holding the connector, but the data lines aren’t making a reliable connection. That fits with what you found earlier — the data trace tore away with the port. Even with the port held in place, the data pins probably aren’t making solid contact with the lifted trace, so USB enumeration fails. Gluing the port back would only fix it mechanically; the data path would still be broken.
That makes sense. So the LEDs are basically just showing power, and the line that actually tore off is the data one. Would re-soldering just the shell be pointless then? I’m on a tight timeline for a project, so I’m trying to figure out if this is something that can be repaired or if I’m stuck.
Right — re-soldering only the metal shell won’t help, because the issue isn’t the shield, it’s the lifted power/data trace. The port’s anchor pads were weak, so the signal traces took the mechanical load, and at least one data trace came away with the connector. The LEDs confirm partial power while the port is held on, but USB can’t enumerate without the data path. To get it working again, the torn trace(s) would need to be repaired with fine SMD rework, and even then there’s no guarantee. In the meantime I’d stop holding the connector on and powering it, to avoid making the lifted pads worse.