avatar

I Dream of JNE

Member Since: September 10, 2011

Country: United States

Profile

Check out what you can do with a Beaglebone and other open-source project ideas at http://knek-tek.me

  • Note that the switch on the board doesn’t cut power from the USB cable to the 5V rail, it only enables the 3.3V rail. So if you’ve got stuff hooked up to the 5V pin, that will be powered as long as the USB cable is in.

    I would have preferred the power switch to be placed just after the diode so that it cuts both the 5V and 3.3V rails. Stuff can get damaged if the 5V rail is on but 3.3V is not, for example if this board is hooked up to a circuit using both voltage levels.

  • No it is not connected to anything, including the GND plane. You could attach a jumper on the board to GND though if you want to use it like that.

  • Make sure not to use D0 (pin 17, GPIO0) for an active-low signal, because if you do, you won’t be able to reset the board if that net has a low voltage on it. That low voltage on D0 will prevent bootup. I experienced this intermittently where my board could boot up fine (because the net attached to D0 floated to a high voltage) then in software I set D0 to GND, then when I reboot it can’t start (because now the net is grounded). Took me all day to figure out.

    D15 also needs to be at 0V during bootup or the device won’t start because it’s looking for a non-existent SD card. Sure would be nice to have that on the graphical datasheet!

  • Just taking a wild stab at this, the VCC net in the schematic is the 3.3V net so an external 3.3V power source would be attached anywhere there, however, one thing to check out is the datasheet for the AP2112K-3.3V voltage regulator on the board and see what happens if you power the 3.3V side externally and don’t have 5V on the other side- the voltage regulator may have severe problems if you power it “backwards” with 3.3V on it’s output, and GND or floating voltage on it’s input. For example, power might flow backwards through it and burn it out.

  • 12mA per pin max.

  • According to the datasheet I found for the ESP8266 chip itself:

    3.1.1. General Purpose Input/Output Interface (GPIO) There are up to 17 GPIO pins. They can be assigned to various functions by the firmware. Each GPIO can be configured with internal pull-up (except XPD_DCDC, which is configured with internal pulldown), input available for sampling by a software register, input triggering an edge or level CPU interrupt, input triggering a level wakeup interrupt, open-drain or push-pull output driver, or output source from a software register, or a sigma-delta PWM DAC.

  • The SPI pins as defined in the SPI.h library for this board are: MOSI = D13, MISO = D12, SCLK = D14.

    And then you should be able to set CS pin to be whatever you want in software, if it’s an available GPIO on the ESP8266 dev board and not used for something else (don’t make my mistake of using D0 for CS! See my post above).

  • You are clearly burning out the on-board power capability by drawing too much current. You will not be able to power your servo from this board based on what you described. To get your system working you’ll need another power source for the servo’s power, then you need to connect that GND to the ESP8266’s GND, and you can still keep your signal wire on the ESP8266.

    Strange that the voltage regulator was hot, since you pulled current from upstream of it. Probably what happened is that your servo drew so much power that it caused voltage droop on the 5V line, which made the LDO (LDO = voltage regulator) have trouble and go into a very inefficient mode where it could not supply the 3.3V out at such a low voltage. Perhaps you even caused the 5V line to temporarily droop below 3.3V, at which point current would flow backwards through the LDO unless it was protected.

  • I was thinking the same thing, here are the initial thoughts I was going to pursue: * How much voltage can a solar panel the same size as the edison boards provide? Is that voltage enough for this battery? * How much more power management circuitry does this require? Adafruit sells a board for that but it’s got some size to it due to inductors and capacitors and plugging 3 devices in (solar panel, battery, load). * For full sunlight how many minutes of sun are needed to power 1 minute of computing and wifi?

  • I was going to suggest the OLED screen board but I guess if the Edison has no way, via software, of knowing the battery level it’s a moot point. Even with some clever software to try and monitor power usage and estimate, that’d still be quite a bit of work.

No public wish lists :(