in the "getting started" pages wifi section it is mentioned that the ESP-01 would be 5V tolerant on gpios THAT IS NOT TRUE! That is why the adapter boards all have 2 little mosfets as buffer (see schematics) and even espressif warns see item 5.14 on page 25 from their faqs
Well, I have seen contradicting informaton if the RX can cope with 5V not not. I have so far run the GPIO pins (to enable and to flash) at max 3.3V and the RX at 5V directly from the Arduino and it has not blown up on me yet. Btw, even the FAQ even contradicts itself... 4.1: Peak analog circuit current: 350 mA 4.3: The maximum analog power (instantaneous) may be considered to be 500mA .. In my world "maximum instantaneous" is "peak". But we should probably add a note/disclaimer to the info in the getting started pages. Thanks for reporting this, Harald.
over at esp8266.com there were some incidents where modules developed strange behaviour over time. The paragraph about not being 5V tolerant was definitely changed and was positiv in earlier documents. As most of your users will go for the shields, there is no harm. As good measure a hint on using a series resistor with plain ESP-01 wouldn't hurt. But I read in the thread that someone is making a pcb, at least that should be "according to specs". p.s.: when using bare ESP-01 soldering a 470ยต/10V electrolytic across vcc and gnd on the module solved some "odd errors".
Everyone has their opinion and there is nothing like debates between engineers. However, I'll go with the CEO of Expressif, Teo Swee Ann, and I quote: "i can reply officially here: it is 5V tolerant at the IO. while the supply voltage is at 3.3V." Happy to point anyone who needs it to engineering testing data that backs this up as well. Lots of things can destroy these boards, like the greater than maximum 430mA currents bursts that can occur in certain situations and are responsible for failures on startup. Have good, low resistance connections and a solid voltage source! That said, we will be adding suggestions to our paths for Tinkerers and Engineers about added components to all the different hardware items to "take things further". We want to balance the simple approach with the smallest need for tinkering or confusing people with too much information with side tracks that can have people design the "perfect" system or add things for an abundance of caution. Same with a troubleshooting section on the effect of wiring, filter capacitors, etc.
That may or may not have been the CEO and it may have been taken slightly out of context. I'll trust the data sheet for the ESP8266 chip itself over what the CEO may have said in a tweet.. In section 5.1 (Electrical Characteristics) the maximum voltage listed is 3.6 V maximum. Nowhere in the Espressif provided documentation does it state that the GPIO pins are 5V tolerant. However, they are generally tolerant to 5V for some time, *BUT* I would not encourage it to be used in that manner. It has the potential to shorten the lifespan of the ESP8266 considerably. With the cost of a UART TTL converter being as cheap as they are it is a very inexpensive way to ensure the ESP8266 lasts for some time. Where is this coming from? I didn't see it in the linked data sheet or tech manual (though I didn't go hunting for it either). I'd expect that sort of current draw to only occur during WiFi radio operations and only in short bursts (primarily occurring during the connection process).
and that is exactly why I suggested the electrolytic on the esp-01 because the 3.3V from arduino mega is all but a solid voltage source ...
@FlightRisk Very interesting. The spikes every 100ms is as stated, wifi beacon ping. This lines up I believe with the esp8266 tech docs as well regarding power consumption. There is also a new ESP32-C3 on the horizon that is intended to replace esp8266 as well, it will have a peak draw similar to the esp32-s2 (350mA peak with radio active and all GPIO pins HIGH/3v3).
