ESP8266 Community Forum

- Sun Jul 26, 2015 6:51 pm #24333 Thank you Kriegste,

I forgot to mention I did put a 2 uF capacitor Vcc to ground at the chip. The tx and rx wires are very short. I was only able to connect to the esp chip at one baud rate. Interestingly, it was 230400 (and only 230400) at one time (and one flasher), but then changed to 115200 and only 115200. As I did a lot of tries, now cannot remember if it was the same or the other of the 2 chips I have, and also if it was specific to the flasher I used. Anyhow, I will follow the spirit of you suggestion and make sure the power it sable and clean, and noise is minimized.

Thank again

- Sun Jul 26, 2015 7:15 pm #24335 If you use a LD1117, for example, the data sheet suggests 10 µF.

If your current voltage regulator (no pun intended) is very weak, try 220 µF... it's worth a try until you have a proper solution. And it is very important that it is very close at the ESP!!

For programming, you can use almost any baud rate. The chip seems to auto detect it. So you could try 9600, too.

- Wed Jul 29, 2015 9:35 am #24548 Changed the wiring to very short wires and a small board with an adapter socket for the ESP8266-01. Use stable 3.3V power supply for the ESP, the FTDI serial-TTL gets its power from USB. Put 47uF capacitors on both o fate the +3.3V. Also connected the Reset pin to 3.3V, to make sure it does not reset accidentally. Found this: viewtopic.php?f=6&t=2791#p16082 and followed the recommendation (changed 2 lines on the esptool.py). Now the download goes as far as 60% or so, and breaks again: "Exception: Invalid head of packet".

The above-quoted link says the problem is a 20-second timeout of the ESP chip. Is there a timeout do you know?

While writing this, tried the same with a second ESP chip, and it worked! 100% downloaded. and exited gracefully.

Inserted the first chip again, it also worked! Wonder why only on the third try...