dksmall

Member Since: November 25, 2009

Country: United States

  • An Arduino based on a SAMD21 is a nice idea, but I think your product description should emphasize that this is a 3.3 volt board, not a 5 volt board like the original redboard. I ran into other differences with your original SAMD21 board, such as libraries that won’t work because of the differences in the processors. And EEPROM contents (flash emulated) were erased each time you flashed the board. Does the newer boot loader improve any of these differences?

  • No anit-static mat, ESD monitors, and probably no wriststraps.

  • Without the schematic there’s no way of knowing that the solder jumper controlled VCC going into the micro. The jumper could select VCC for the I/O, while the micro always runs off a fixed supply.

    Have you been on AVRFreaks? That’s the technical forums for the AVR processors and occasionally you will see the topis about overclocking the AVR’s. From what I recall the general consensus is yeah go ahead and do it with a one-off or a proto, but don’t do it for production units.

  • Did I miss the reveal on the design flaws from part 1? Or at least the one you were hinting at that no one figured out?

  • I’ve never priced a flying probe tester, we typically pay for that type of testing with some of the proto designs, at the board house or assembly house, rather then operate one ourselves. Back when we had the full board factory in Phoenix, we had 10 or so ICT testers (board level) and lots of functional testers, many in environmental chambers to really abuse the product. I’ll work on getting some pictures up. What just occurred to me is it would be fun to show the evelotion over 35 years of using this type of test architecture. I think I did 4 designs where the first ones were total wire wrap, to the one I did 2 years ago, which almost all SMT.

  • That might be just a bad picture/angle. There are several pins that have that same dark spot that looks like a hole where there should be solder.

  • At first I thought this was going to be an article on flying probe testing, a common inexpensive test method used during proto builds. But I see you’re doing something similar to what I have in place within our test engineering group. We call it a test controller and it’s like your Flying Jalapeno on steroids. The PCB is 9" x 11" and my latest version uses an ATmega1281, 2 Xilinix XC95144 CPLDs (which can be customized for each test development), an ATtiny1634 for high speed RS-485. And there is 32 analog inputs (16-bit), 8 DAC outputs, 32 digital outputs, and 24 digital inputs. Power section has +24, +15, -15, +5, and +3.3. This is just the base configuration, 50% of the board is laid out for DIP wire-wrapping, or PCB daughter boards, so you can further customize each design for whatever you’re testing. I’d post a picture but looks like we can’t attach anything to a comment.

  • As far as design flaws, your silk is showing the part values instead of the reference designators.

  • Actually a little googling and it looks like I have a common problem that’s been fixed but not released yet. Several of the commands have a runtime error. I was using a W7 PC, going to try W10 now and see if it’s any better.

    Update: W10 has the same issues. Also tried the uno board with the same results. I guess I’ll wait for the next release of the cli tool.

  • Well I’m trying to compile a sketch for a redboard, but cli can’t find the board. If I do “arduino-cli core search sparkfun”, nothing is found. I have the redboard installed and my sketch compiles with the arduino IDE. Any ideas?

No public wish lists :(