Connection Reliability Problems

[Michael-Equi]

Hello, I have been working with a pycubed V5 running CircuitPython 7 using both a Windows and Mac computer and am experiencing issues getting it to show up as a drive on my host computer. It seems to show up once every 10-20 times it is plugged but somedays I just cant get it to connect at all. Additionally it seems to save files only occasionally out of the sessions in which it does show up as a drive. That is, if I make some changes when it is properly connected and proceed to save everything + verify it is saved before unplugging, then unplug the pycubed and repeatedly plug it back into the computer until it shows up as a drive again, sometimes those changes I made in the last session are not saved.

I am not sure where to start debugging these issues or if they are even related, any advice is greatly appreciated!

[maholli]

Hey @Michael-Equi. I've never had this issue in the 4+ years using PyCubed boards (i.e. I've made over 100,000 file-saves to a single board without any filesystem or mounting issues) so I'm sure we can get it figured out.

Here are a ton of questions that cover a whole host of things I can think of.

• Who built the board?
  • If it's the v5 I built for you guys, I worked with it extensively over the course of ~5 days before it made it to you (it was my only board on-hand, you guys got to have it since it was an emergency) and had zero issues with it.
• Have you tried multiple different USB-C cables?
• Has the USB-C connector on the board been damaged?
  • Does it connect/disconnect when you have it plugged into the computer and then gently articulate the USB-C-to-PyCubed connection?
• Is the board being used standalone on the bench is it inside a structure?
• Any external connections to the board?
  • Battery connected?
• What's your RBF jumper configuration?
• (edit) I personally never eject the device on Windows/Linux and have never had an issue, but I'm told Mac's can cause filesystem corruption if not ejected properly
• (edit) What text editor do you use? I've heard notepad++ has some weird autosave behavior that can cause filesystem weirdness as well
• (edit) With the board plugged into your computer, use a multimeter in "dc voltage" mode and measure between any of the 3V3 and GND through-holes. Is it getting power?
• (edit) rename your main.py and/or code.py files on the board (so that they don't get overwritten) and put the "default" main.py file from here on the board. Unplug/plug again and wait for ~30s to 60s. Anything happen? Does the RGB cycle but the board not mount as a mass storage device?

This sounds like textbook "bad USB cable" behavior, but once I know more we can further troubleshoot further 😄.

[Michael-Equi]

Im trying to avoid unsoldering the payload board at this point and it seems to be working fine now, although nothing has changed which is a bit frustrating. I grabbed a couple snapshots of the power on voltage and it looks fine to me. Maybe you can compare it to another pycubed that doesn't have a payload board on it to see if it is irregular enough to potentially sometimes cause problems.

[maholli]

@Michael-Equi thanks for the scope pic! It's very revealing.

That's a ton of overshoot (and ripple) on your board. I measured 4 separate pycubedv05 boards I had laying around (USB powered, scoping 3.3V rail) and they all exhibit this behavior:

• Are your scope pics from USB or battery power?

[Michael-Equi]

Those are from USB power

[maholli]

Strange... Will you take the highest-res photo you can of both the top and bottom of the board and I'll look for anything that may be amiss.

Email them to me at support@pycubed.org

[Michael-Equi]

Oh ya, that is quite a difference. Not sure why I thought it looked ok.