Member Since: September 20, 2011

Country: United States

  • I love the look of the colored PWM headers! That should make hooking things up that much easier.

  • In silicon valley, Sparkfun would be considered a "lifestyle" business, one where the immediate owners can pay bills with profits, but the growth is not there to significantly impact the returns of the venture fund. VCs dislike these businesses. However, most of the "grass roots" value to society from businesses comes from those lifestyle businesses, not from the megacorps that silicon valley is set up to conceive and grow. In fact, because of this structure, it's actually harder to build a small, sustainable business in SV than on the heartland, unless your business revolves around providing value to those trying to build megabusinesses.

  • "This supply uses two 200V caps in series, but I don’t understand how it works yet."

    Two equal capacitors in series will double the voltage tolerance, but halve the capacitance. Two capacitors in parallel will keep the same voltage tolerance, but double the capacitance. It doubles the voltage tolerance (each capacitor sees "half") as long as the capacitors are equivalent (same ESR, leakage, capacitance, etc) but the equation gets more complex (similar to parallel resistors) if they are dissimilar.

  • The later versions of the ESP8266 have up to 12 GPIO pins. (ESP-8266-12)

    If you already have a TTL UART adapter for your computer (FTDI, CP2102, or whatever) then all that's needed to program this device with the Arduino IDE is soldering a "programming mode" button/switch between GND and GPIO0.

    The specificaion sheet says the imp has a Cortex M3, but doesn't say what the frequency is, so I can't tell if it's faster or slower than the 80 MHz of the ESP.

    The server-side solution for the imp is something the ESP doesn't have at all.

    In the end, is the difference worth $40? That's up to you to decide. I like the ESP, but I already run my own servers anyway!

  • I'm developing a rover, and I was using a Sirf-III USB puck that would not get a lock indoors. After about 15 minutes of warm-up, this device actually does get a lock, although somewhat intermittently. I wired its UART to a Atmega32u4 and forward the NMEA stanzas to gpsd over USB. I will have 20-60 second periods of "no fix" and then as much time of "3D fix" as viewed by xgps. So, yay for better sensitivity! But it's not yet perfect.

  • I don't think so. I think the rise/fall is already included in the "poles." 8 poles means 4 magnets. 4 magnets timesd two poles/states each (north and south, high-and-low) times two encoders equals 16 transitions/counts. I could be mis-understanding how the poles are counted, though.

  • 8 poles and 2 encoders should give you 16 counts per revolution. 360/16 is 22.5 degrees per count. With 100 mm wheels (4") that would give you about 2 cm per count. I'd like to know what the output is -- is it totally raw Hall sensors that need conditioning, or is there any clean-up of the signal to turn it into something nice? (Answered my own question: The outputs are open-drain, so all the clean-up needed is a pull-up resistor.)

  • I have one of these, but actually want to just do 3.3V I/O. Can I use this board as a simple break-out board, by connecting to the IN/OUT pins on the sides, rather than the DIN/DOUT pins at the bottom? And if I only have a 3.3V rail, would connecting this to both 3.3V and 5V make sure the regulator doesn't break? (back-feeding regulators is a bad idea in general, but equal potential should be safe)

  • Tantalums have a lot longer lifetime than electrolytics when treated right, though. If you want to build for the next 10 years, not just the next 1 year, consider tants your capacitor of chocie.

  • The live geiger counter feed seems to have flatlined.

No public wish lists :(