Failure uploading ESP8266 firmware through Pico

[espruino-discuss2]

Posted at 2016-08-16 by JoeAndrieu

I have a Pico, an ESP8266-ESP1, and the rev2 shim. Actually, I have three of each, thankfully. But after hardware errors on the first batch, I've apparently botched the firmware on the ESP8266 twice.

The behavior is the same on #2 and #3, and I was double-extra-careful with all my soldering by that point.

Here's what I did. FWIW, I'm on windows 10, using Python 2.7.

1. Update firmware on the Pico with web IDE.
2. Test AT baud rate with code from tutorial, get back reasonable, but old version of ESP8266 firmware at 115200 baud

AT+GMR\r\r\n00160901\r\n\r\nOK\r\n

3. Load Espruino FTDI code (broken into functions so I can reset independently)

    digitalWrite(B9,1);
    digitalWrite(A1,0);
    digitalWrite(A10,0); // pulse reset
    digitalWrite(A10,1);

    Serial2.setup(115200, { rx: A3, tx : A2 });
    Serial2.on('data', function(d) { USB.write(d); });

    USB.on('data', function(d) { Serial2.write(d); });
    Serial1.setConsole();

4. Disconnect IDE from COM port and run esptool.py to upload recommended v25

  esptool.py -p COM4 -b 115200 write_flash 0 "c:\Users\Joe\Downloads\ESP8266\ESP8266_AT25-SDK112-512k.bin" esptool.py v1.1 Connecting...
 Running Cesanta flasher stub...

 A fatal error occurred: Invalid head of packet ('\x13')

5. Try initial AT test again. With NO valid result. Powercycling, reburning Espruino firmware has no effect. I can't recreate the initial firmware test. Although I do get visible output when toggling the reset GPIO.

Note: I have tried with both 74480 and 115200 as the baud. With 74480 baud, the reset of the ESP8266 gave a readable output (although it seems a bit brief)

ets Jan  8 2013,rst cause:2, boot mode:(1,7)

If I change ath A1 setting to 0, I get something different at boot

 ets Jan  8 2013,rst cause:2, boot mode:(3,6)
 
load 0x40100000, len 24444, room 16
tail 12
chksum 0xe0
ho 0 tail 12 room 4
load 0x3ffe8000, len 3168, room 12
tail 4
chksum 0x93
load 0x3ffe8c60, len 4956, room 4
tail 8
chksum 0xbd
csum 0xbd
�`jÔNø

And a different error from esptool

esptool.py -p COM4 -b 74880 write_flash 0 "c:\Users\Joe\Downloads\ESP 8266\ESP8266_AT25-SDK112-512k.bin"
esptool.py v1.1
Connecting...

A fatal error occurred: Failed to connect to ESP8266 

It didn't seem to help to set A1 to 1, but it was one of the few knobs I had to play with.

I am powering solely through a USB port on my PC. I've tried both USB 2.0 and 3.0. Both give the same results.

Suggestions?

[gfwilliams]

Posted at 2016-08-16 by @gfwilliams

Did you change ESP_RAM_BLOCK and ESP_FLASH_BLOCK in esptool.py first? That could be one cause of problems...

If you've had the ESP8266 reporting version numbers then it looks like your wiring is absolutely fine, so it's just software that's the issue.

[espruino-discuss2]

Posted at 2016-08-16 by JoeAndrieu

I did. Not before the first one, but by the time I got to the second one I had.

I'll try with a direct FTDI link...

[gfwilliams]

Posted at 2016-08-17 by @gfwilliams

By default, all the Pico's pins will be floating - so if you do use an FTDI cable there's no need to disconnect the ESP8266 from the Pico - just tack RX, TX and GND wires on top of the module.

It's annoying - I'm not really sure what could be causing the issue, as I've updated loads of ESP8266s this way... Unless there was some change to the uploader recently that broke things.