Member #134773

Member Since: May 19, 2010

Country: United States

Profile

I started playing with electronics in the mid-1960s, and with computers shortly after Neil Armstrong took "one small step". I got a degree in CS in 1980, and started working then as an engineer.

  • I hope that projects like this can inspire an interest in "STEM" in your daughters! At least it teaches them that interesting and useful stuff can come out of "DIY" projects!

  • I'd like to see a version of this with the 16 button keypad -- it could make for easier-to-navigate "menus", e.g., the programmer/maker can assign one button to be "up" and one "down", one for decimal point, one for "enter", one for "backspace", and still have the 10 digit keys for direct data entry.

  • The Qwiic OpenLog looks good, BUT (and frequent readers of my comments will be able to guess what's coming next) it would be a lot better if it had an RTCC (Real Time Clock Calendar) on board. Many of the data logging "use-cases" (at least in my experience) are dependant upon reasonably accurate time stamps. The reprogrammability of the '328 would allow for a very small (and reasonably low power and low cost) data collection systems.

    I can hear the screams "but what about the Qwiic Real Time Clock Module?" Going to that item, though, besides the banner saying "You last purchased this item" (so yes, I know about it), it's also on Backorder, and there's a banner saying "We do not currently have an estimate of when this product will be back in stock."

    There's another advantage of putting the RTCC onto the same board -- the "INT" output pin could be connected to one of the '328's IO pins so that the user could set up to have the RTCC produce "alarm clock" interruptss and the '328 act upon them.

    Another thought: The I2C is a bus where only the "master" can initiate a data transfer, I've seen several I2C "slave-only" devices, including RTCCs, IMUs, and even GPS, and other sensors, that can generate an INT (interrupt) signal, usually set up so that it can be "wire-ored" together, to get word to the "master" that it has something "interesting". I'd like to see a "Qwiic Plus" that adds a wire for interrupts. Since we typically have fairly few things on the Qwiic/I2C bus, it wouldn't be a major problem to have to poll the ones that can generate an interrupt to find out which one(s) have actually asserted the interrupt.

    If I recall correctly, the ATmega328 doesn't have a "sleep mode", only waking up on an interrupt. Too bad -- that can really help the bottom line in a power budget!

  • Neat project, Rob!

    First thought is that when I glanced at the B&H link you provided, I did notice that they say the Kaiser ringlight is "special order", and expected to take 7 to 14 days (I sure wouldn't be surprised if it took longer), while all of the items in your parts list are "in stock" (at least as I type this), meaning that if someone with a bit of skill, could be taking pictures with yours sooner than with sprining for the commercial one!

    Admittedly, I didn't look thoroughly at the commercial one, but one thought about your ring light is that it could do a LOT more than at least the less expensive commercial ones. For instance, since the LEDs are individually addressable, it could be programmed to illuminate one side one color, and the other side with another color. Or if your camera can take videos, experiment with illuminating with a "moving" light.

    I'd have been inclined to put the battery on an easier-to-disconnect connector than the JST connectors, so that you can have multiple batteries in your "gadget bag" -- I've had way too many times when I was in the midst of a remote photo session when the battery on some key "gadget" (including the camera!) died! (Even "once" is "way too many times", IMHO.)

    I just wish tilt & shift lenses weren't so expensive... it's tempting to go back to large format film cameras...

  • I'm sure the QC tests the microcontrollers, but first, I've seen times over the past half century or so when once in a while, one will "slip through", and second, I've also seen "post test failures" in systems (wayward static electricity, and even wayward cosmic rays, can cause such).

    As for the AVR programmers, the Pro Mini isn't as convenient as say, the Uno, for use of one, at least not without a special jig which us "mere mortals" are unlikely to have. I bought the aforementioned Pogo Pins with the intent of making a "jig" that I could use with either the 5V flavor or the 3.3V flavor of the SparkFun FTDI Basic Breakout which works well wiith the corresponding flavor of the Pro Mini, and also works well with Adafruit's HUZZAH ESP8266, and save having to solder in the 6 pins for programming on each, but that adapter, like so many other projects, has succombed to the shortage of "round tuits"...

    Just as an aside, I happen to have JTAG programmer that I haven't used in years, and a Pi AVR Programmer HAT that I've never gotten around to attaching to a Raspberry Pi.

    BTW, I posted my comment above in hopes that some might benefit from my time in the "school of hard knocks", without taking the "courses" themselves. ;-)

  • One other comment: I strongly advise that before you solder any CPU module onto another board, you program each one with at least a "blink the LED" program. Although I've only had this happen once, I had a CPU module that wouldn't program and I only found out after building up the hardware. It wasn't worth the effort of trying to desolder it, especially since OSHPark has a 3-board "minimum order", and I had spares of all the other parts. Anyway, ever since, I've always checked that the CPU module can be programmed before soldering it onto another PCB. (For ProMinis and others this often involves soldering the programming pins onto it. I've never gotten around to setting up a Pogo Pin adapter that I could use to program them without soldering anything to them.)

  • I forgot to mention that I used digital 3DOFs from Adafruit for both these projects. They are easy to deal with under I2C, and even have address select line, allowing me to put both IMUs on the same I2C bus.

  • Maybe 3 years ago I built a "flashy Santa hat" that had, if I recall correctly, 11 Neo-Pixels that responded to head movements. (Unfortunately, it has managed to "disappear" -- I think maybe I took it along with several other items to a "demo" and it didn't manage to come home.)

    Another project, although it didn't involve "flashy lights", did involve two 3DOF IMUs. When recovering from a broken wrist, I was prescribed this "brace" that would apply (gentle) pressure to twist my wrist to try to regain movement. I found the directions rather nebulous, and couldn't come up with a practical indication of how much "twist" I'd applied that I could read while wearing the thing. (You put the brace on at "zero twist", and then you start cranking it.) I figured out how to use two IMUs, one attached to the "fixed" portion of the brace and the other attached to the "rotating" portion to calcuate the "twist angle". For a CPU and display, I used a MicroView. It displayed the angle, and included a timer to help with knowing how long I'd been wearing it this session. I approached the brace manufacturer, and they were NOT interested, saying that it would have to go through FDA approval -- your tax dollars hard at work! Oh well...

  • We can always add bags onto the side...

    BTW, the SparkFun computers are showing me (a) that I last ordered the RTC on July 23, 2018, (b) they're on "back-order", and ( c) there's no estimate of when they'll be available again. Seems to me, that makes that solution a non-starter, at least for the moment.

    The Teensy 3.5 has the RTC built in (although it takes a little bit of doing to connect the coin cell), as I type this it's showing "in stock" and has a built-in uSD slot. Admittedly it's not directly compatible with standard "shields" (and lacks a Qwiic connector -- partly because it pre-dates the Qwiic system), but it DOES demonstrate that a chip with a built-in RTC can be used in an Arduino-compatible board.

  • I got a bit excited when I saw "RTC crystal" in the RedBoard Turbo description, and investigated further. When I downloaded the Datasheet for the SAM D21 and saw the Microchip logo, I got even more excited, as I've fondly used the RTCC built into PICs. (I'd somehow missed that Microchip bought Atmel.). However, getting down into it, it's a "Real Time Counter", with a clock calendar mode -- it lacks one KEY feature, namely, the ability to run off a separate power source (pronounced "coin battery" or super-capacitor), even when the rest of the device is completely powered down. It's not SparkFun's fault, but it's still frustrating. Oh well...

No public wish lists :(