
Pete Scargill has posted a note to let people know that there’s another ESP-12 variant available from AI-Thinker. It’s designated as the “ESP12E” (the “E” presumably denoting “Extended” “Enhanced”) and it differs from the previous versions in that even more of the ESP8266 chip pins are broken out to pads along the bottom edge of the module. There is a paucity of information available on just how to use these new I/O pins as yet and, to be honest, early indications are that this module might not be too useful for the average user. It looks as though the pins are the multiplexed serial data and clock for the on-board memory, which might make them unattractive unless you really, really need I/O at the expense of everything else. However, time will tell and I’m sure there will be some novel implementations using this module popping up soon.
A definite plus point for the ESP12E though, is that the flash chip appears to be (a lot!) bigger. Here’s a post from Koelie2 on the ESP8266 forum with his findings after removing the RF shielding from an ESP12E module.
UPDATE: Link to Eagle library file temporarily removed — I just discovered an updated layout diagram on AI-Thinker’s shop page which shows that the new pins are not symmetrical about the centre-point after all. The link will go live again once I’ve had time to update the library file. [ In the meantime, here’s an Eagle library file for the new ESP12E version. Note that this is untested and the pads for the new pins were placed with the assumptions that they are symmetrical about the centre point of the module and that they share the same spacing and dimensions as the existing pads. ] Please let me know in the comments section if you spot any (other) obvious errors.
It’s also worth noting that the silk-screen on the new ESP12E modules shown on the supplier’s site (see Pete Scargill’s original post for the links) show the original, incorrect labels for GPIO4 and GPIO5. According to the board overlay on the same page, those silk-screen labels on the physical board are reversed. The Eagle library file (above) has this corrected (that is, the library file matches the board overlay diagram, not the silk-screen).
There also seem to be at least two, conflicting versions of what the new pinout assignments actually are (one layout on the Aliexpress site and another on the AI-Thinker shop page on Taobao). Given that the units all have “AI-Thinker” as the maker’s name engraved on the RF shielding plate, I’m going to use the Taobao version of the pin assignments for now. Hey, it’s good to see that we have such a track record of consistency with all things ESP8266… the documentation is consistently fubar-ed.