Member Since: January 12, 2011

Country: United States

  • Well this is kinda weird -

    I didn't realize that these were 5 volt devices, but I figured since I was prototyping on an Uno I could get the systems working together and then worry about getting 3.3V components.

    I set it up so that this was running off of the Uno's 5V pin, and everything worked fine.

    Then I decided, "What the heck, I'll put it on 3.3V and see what happens". Apparently, it works ok - the time seemed correct - so I guess that's nice.

    The particularly strange part, though, is that the thing I'm working on is time-critical, so I had code timing my communication with the RTC. At 5V it took ~1060 microseconds, but with the RTC running off the 3.3V pin it took ~290 microseconds - more than 3x faster.

    I figure hey, the RTC has its own battery, maybe it's ignoring the 3.3V line and just using its internal reserves to talk? Nope, if you don't provide it with anything it'll refuse to chat.

    So that's weird I guess. The datasheet says that the min for Vcc should be 4.5V, but it actually seems to work faster at 3.3V. It might be slowly breaking on the inside or something.

    edit: aha, found the downside - when you cut power at 3.3V, the RTC forgets what time it was despite the backup battery being present and charged. There had to be a catch!

  • If all you're doing is 5V to 3.3V (or 2.8V or 1.8V), the logic level converter is a better deal. What this thing does is translate between a range of voltages on either side; for instance, you can use it to translate between, say, a 1.8V sensor and a 3.3V microprocessor, which is out of range for the logic level converter.

  • Probably, but I imagine that it would only work well if you were being yelled at by giants.

  • I just got mine, and here's what I measure:
    The straight extension shaft head is a smidgeon more than 0.25".
    The bendy extension shaft is a bit wider at .44".
    Depending on how deeply the bolts are recessed, you could also stick the torx bit inside the 4.0 M hex socket, which is pretty long and skinnier than the extension shaft. Might be hard to recover afterwards unless you can tip the enclosure over.
    Though I have to say, if you're really worried about the tool being too wide, you just can't go wrong with a set of plain old torx screwdrivers.

  • Oh man, I might have to order one of these just to see if I can reverse-engineer that nifty "program it from the website" thing. It looks like it's all done client-side and they haven't obfuscated their Javascript code, so it shouldn't be too difficult.

  • Huh, I had no trouble getting this to work with a breadboard using the tutorials at Gooligum Electronics (as recommended by the PIC forums here on Sparkfun), and that's for PICkit 2 with the PIC Starter Kit.
    The only weird thing is that this comes with a poster that seems to claim that the pin numbering on the PICkit is in some really odd order, when in reality they're in numerical order.

  • This stuff looks awesome, but I have to ask - judging by the ribbon soldering tutorial, it seems like the only thing keeping the ribbon and the brick together is the solder. Does that actually last? I would imagine that solder would be unhappy with all the stretching and bending that fabric does. Would it be useful to reinforce the connection with thread, or would that not actually do anything?

  • What I liked the most about this episode was that Pete talked about some of the engineering reasons behind the setup on a kit I've assembled (the through-hole breadboard power supply). I liked finally understanding why I had to solder in two different capacitors that looked pretty much the same :)
    I think it would be interesting to go over the reasons why some other SparkFun kits are assembled the way they are, especially for ones where it might not be completely obvious.

  • Hmmm... if you get those LEDs individually it's nearly $50 for 27 - but there's a pack linked down there with 25 of these guys for $20. Maybe this kit's price should be reviewed?

  • arbarnhart:

    I am hoping for an "n" degrees of freedom board that only reports "n-1"; a little bird told me that is one of the most common things in "bad" sensors...
    Your little bird must be pretty accurate, that's almost exactly what I got! I received a 5 DoF board, hooked it up to an oscilloscope, and discovered that 4 DoFs work - I get nice clear signals for X, Y, X-rotation and Y-rotation. Even Z (which I assume is the broken one based on comparing its waveform to X and Y) still kinda works - it's just very very fuzzy instead of nice and smooth like its siblings.
    Now all I need to do is order more stuff in order to make something neat with it - that's how they get you I guess :)
    Edit: Chuh this was supposed to be in reply to arbarnhart up there as indicated, it seems that got lost when I previewed the post?

No public wish lists :(