SparkFun Forums

[adamjvr]

Hi, I have an ESP8266-Thing that I bought as a nifty dev board for the ESP chip. After adding headers the first thing I've been trying to do is to test program it with an example sketch with the Arduino IDE. I made sure I had the board support properly installed and setup by test programming several ESP-01 modules and a ESP-12E module breakout all successfully. For some reason I cannot get the ESP8266-Thing to program, the IDE attempts to load the compiled file, I found this by looking at the UART signals going from my FTDI breakout to the ESP8266-Thing with my logic analyzer. It appears the ESP chip is not responding to the programming routine, since I am new to this chip I'm not sure what that means. I hope I didn't get one with a dud IC.

[Valen]

Did you read the hookup guide? And connected the 2 gpio pins of it to either vcc or ground to enter bootloader mode? (I forgot which way it has to be for each ... it's in the guide)

[adamjvr]

Did you read the hookup guide? And connected the 2 gpio pins of it to either vcc or ground to enter bootloader mode? (I forgot which way it has to be for each ... it's in the guide)

I did follow the hookup guide, I am using an FTDI breakout like they have in the guide, the DTR pin is tied to a pullup circuit on the thing which pulls the GPIO0 pin low for programming, this is supposed to be triggered by the DTR signal. IDK what else I can try.

[Valen]

Is it the exact same FTDI breakout? There is a 3.3 volt version and a 5 volt version. The Thing product page specifically mentions the 3.3 volt is required. [EDIT] In hindsight it should not matter which version you have. It's 3.3 volt on the RX/TX/DTR pins anyway. Only the voltage on the power output is either 3.3volt or 5 volt, and it's not connected through on the Thing board.

Also how do you have the Thing powered? The FTDI has insufficient current supply to power it (infact the Thing does not use it by default). You need external power supply through the USB connector or Lipo connector.

Since you had to solder the headers in, or a way to hook up the FTDI Basic you may have bad solderjoints on it. It's going to be really hard to test the connection of the RX and TX line, as they go directly to the chip. There are no traces to other component solder-pads from which you can check conductivity. You'd have to MacGyver some needles to your multimeterprobes and test on the proper ESP8266 pins. The GND pin can be checked to the other GND pin on the opposite corner of the board. The DTR pin solder connection can be checked by testing for conductivity between it and the SJ1/SJ3 jumper pads on the back. The Vcc pin on the FTDI connector is not used to power the Thing, so they left the solderpad jumper on the back open. There is no point in testing this one (for this purpose).

[jhedup]

Had the same issue with the new DEV version of the Thing. You can manually connect GPIO0 to GND before loading the sketch. Alternatively use the "Generic" ESP board BUT change the "Reset Method" to "nodemcu" instead of the default "ck". This works as the nodemcu reset circuit is effectively the same as that of the Thing-DEV. I have reported this issue with the json board manager file default settings to the Sparkfun technical department.