Your new topic does not fit any of the above??? Check first. Then post here. Thanks.

Moderator: igrr

User avatar
By Barnabybear
#25345 Hi, your devlopment board work looks very neat. The only comment is that the ground from the PSU to the ESP passes through 10 or more connections. For flashing with little current drawn this may be fine, however during normal opperation the ground line could be being drawn high. Bear in mind that if the ESP can draw up to 250mA when WiFi is in use, as I understand if it has been connected to an access point in the past it will attempt to connect on power up. I would try a link between the top and bottom grounds on the right side of the board (as shown in your earlier photo).
Edit: With the current drawn it might be worth taking both VCC and ground directly from the PSU to the ESP as least for testing.
User avatar
By PaulRB
#25363 Thanks BB, I will give that a try.

I'm looking for any suggestions at the moment, I was doing really well, with the ESP-01 reading a temp/humidity sensor over i2c bus and posting the results to a web server, writing my first real PHP scripts and so on. I was about to start writing an i2c slave sketch for an attiny, so it could monitor wind speed, direction & rainfall using those inexpensive Maplin/Fine Offset sensors, using minimal power while the ESP is in deep sleep. Then it could wake the ESP every few minutes using the CH_PD pin. The ESP would then read the various sensors via i2c (temp/Hum/Pressure etc) and the attiny's wind/rain readings and post them all to the web server.

But then the ESP just stopped running the sketch. Rats.

I did try using the screen command with a 74880 baud rate to see if I could make any sense of the garbage I get over serial from the ESP when I press reset. It was still garbage at that baud too.
User avatar
By PaulRB
#25586 I've tried a few things this evening in an attempt to get the ESP sketches running again.

Following BarnabyBear's suggestion, I directly linked the power supply to the ESP's power pins with jumper wires. This reduced the breadboard connections, from ground pin of the power supply to ground pin on the ESP, from around 10 down to only 2. Similarly for the 3.3V connection.

I tried a different power supply. Up to now, I have been using a dc-dc convertor, set to 3.3V, running off 4xAA NiMh cells. Instead, I fed 3.3V from my USBasp adaptor, which has an ams1117 3.3V regulator on-board, which should provide more than enough current for the ESP.

I even combined the two above ideas. No luck.

I am now attempting to decipher the "garbage" that the ESP sends out when it gets reset. Previously I tried to use the command "screen /dev/ttyUSB0 74880", but the result was still garbage. This evening, I experimented with an Arduino Nano 3 running a sketch to send serial data at 74880. The result seen by the screen command was still garbage, so I don't think the screen command is interpreting the 74880 baud data correctly. The Nano and screen commands work perfectly at the more commonly used baud of 115200.

I have also been reading and attempting to understand what sounds like a similar problem on this thread on the Adafruit forum. I may have to try booting my PC into Windows 7, installing Arduino 1.6.5, ESP boards etc and give it a go that way.
User avatar
By Barnabybear
#25590
The result seen by the screen command was still garbage, so I don't think the screen command is interpreting the 74880 baud data correctly.

You should only get 10 to 20 char of garbage on the serial whilst the baud rate is established, more than that sugests that it has booted into SD card mode (which keeps slowly producing garbage char). I've booted into this myself a few times whilst working on beadboard. RST is one of the pins that sets boot to SD mode (see here), to prevent this I connected RST (or GPIO 15 as its sometimes silk screened on boards, but visably doesnt appear to be) directly to VCC (infact on most of my boards its soldered to VCC the pin next to it to prevent random resets) and used CH_PD or disconected the power to force a reboot. Basicaly just swap the connections to the ESP pins 4 & 6 on the schematic on the first page of posts.