Board reliability

I’ve been working for about 1 year on a potential product using the Astar 32U4 SV. The first board failed “out of the box” on the gyro port. The 2nd board has worked for a few months, but suddenly has also failed unexpectedly on the programming interface; can no longer program the board using the usb port. Have tried all listed approaches to no avail. Upon experimenting, if I “tickle it” in just the right way, I can first set it up using the USB port, followed by programming it using the programmer. Sadly, this does strange things to the software and so is useless. I’ve also verified that the first board which failed on the gyro port programs normally as well .
There is just no way I can justify continuing with these components for a project that is eventually intended to be sold to a specific market segment.
I have no choice but to “pull the plug” on this project. These components are great for home labs, hobbyists etc, but in my experience, are not recommended for product development intended to be sold in any reasonable quantity.


I am sorry to hear about your experience with our A-Stars. We manufacture the A-Stars with the same process we use for our other products, including various robots such as the 3pi and Zumo, and in general, these products seem to stand up to a reasonable amount of use and abuse.

Is there anything you can pinpoint on your A-Stars from a manufacturing or design perspective that you think could be the problem? You can see the full schematic, and most of the microcontroller I/O pins are just brought out without any additional protection circuitry, so perhaps those pins are getting damaged in your application. I am not sure what you mean by “tickle”, but if it has to do with jiggling around the connector or flexing the board, do you see any mechanical (soldering) issues?

- Jan

Jan: Thanks for your response. In all fairness, I have discovered that some
of the problems are indeed due to connection problems, which I will
address. As nearly as I can determine without further analysis, I’ve
actually had one failure, apparently a gyro board. If it persists, it may
have been due to static damage, though I do take precautions. The most
significant issue seems to be the unexpected inability to reprogram a
controller board, which sometimes can be corrected, sometimes not, despite
attempting all the various methods documented on your web site. Hopefully,
the failure problem/failure rate is not as bad as I first thought. Time
will tell.