Use this forum to chat about hardware specific topics for the ESP8266 (peripherals, memory, clocks, JTAG, programming)

User avatar
By buzzy
#79147
torntrousers wrote:
buzzy wrote:A bit more? How much more?
Similar to you with 4.2v from a Lipo cell or occasionally connecting them to a bench power supply accidentally set to 5v.


Ok cool! Then you are yet another one that verifies the ESP works fine with 4.2v. Let's see if someone eventually shows up and tells us his ESP broke after some time on 4.2v. So far, nothing. Most people just point to the data sheet and says "it will break".
User avatar
By rudy
#79155
buzzy wrote: Then you are yet another one that verifies the ESP works fine with 4.2v.

That is not what he said. Yes his ESP seems to work fine. That is not a blanket endorsement of powering the parts beyond the ratings.

If I were to tell you that I played Russian Roulette and nothing bad happened, will you then assume that it is a perfectly safe thing to do?

Image
I don't know if it is safe. But I have never had a situation where I wanted to risk damage to the part. The cost of the part isn't the issue. All the work involved in replacing it is, for me. That is if it dramatically fails. It turning intermittently flaky would be worse yet because it could cause a lot of trouble before determining that it needs to be replace.
User avatar
By RichardS
#79156 If this is for a commercial product then for liability sake stick with 3.3V :-)

If this is running a bunch of stuff in your house then 4.2V is just fine... 18650 I find are under 4.0V in a few minutes and hold out most of its life from 3.5-3.7V which is not bad at all....

Again use LiFePo4 if your worried, they hang around 3.3-3.4V most of their life and only ever reach 3.7V maximum and again only for a few minutes.

RichardS
User avatar
By Craig54321
#83217 @buzzy: Thank you for sustaining this discussion. Your experience has been similar to mine. The use of the 18650 is so wide spread and the 'voltage life-cycle' is so within range you would think it was designed for 3.3V devices. The other points are well taken for official commercial design specifications. But I'm a hack and don't need 100% performance that costs me triple to achieve. I'll stick to 95% and call it good enough if it saves me more than it costs me. As long as the house doesn't burn down I'm good with it.

More on my experience: I have accidentally fed 5V to several ESP's. This has been either to the VCC, and more frequently, input pins. Thinking the chips to be dead, I have found that most, if not all, of them survived. I admit, magic smoke has been emitted from supporting components, but never from the ESP's. Perhaps I was witnessing self-sacrificial events to save the 'mother-ship' CPU. I have no forensic idea.

But again, I thank you for the courage to 'buck' the fear of compiance spec's to test the limits of practical experience.