ESP8266 Community Forum

- Wed Aug 12, 2015 3:45 am #25763

Your statement that the GPIO2 pin can be left unconnected is correct indeed, but recommending this practice against advises from several experienced electronic engineers saying it is safer to add an external pullup resistor to prevent the chip from booting into a deadlock is not responsible.

This is expanding the question. The argument is either right or wrong . My statement is right , his is not . Simple as . I am not recommending any practice ,i am proving that what I said is correct and what he said is wrong . Simple as that .

Get you modules here:

https://www.banggood.com/custlink/KvGGhGF4wG

- Wed Aug 12, 2015 4:20 am #25767 Please consider that English may not be the native language for most of us here, and that there are many inexperienced users reading this forum.

Everything is not just black or white, stating that GPIO2 must be pulled up externally may not be strictly correct or wrong, but this is probably the easiest way to have people avoiding troubles with strange ESP8266 behaviors.

There are a number of reliability issues with this chip, mainly due to bad power supplies, lack of bypass/decoupling capacitors, missing pull resistors, etc. that are difficult to solve because of lack of clear directions and incomplete bug reports.

Pushing people to use pull resistors is a good thing, feel free to remove them if you think it works without them in your setup, although it is not considered as good engineering practice.

Going back to the OP, it looks like we both agree that the problem may be due to a bad power supply, do we?

- Wed Aug 12, 2015 9:37 am #25790

This is expanding the question. The argument is either right or wrong . My statement is right , his is not . Simple as that. I am not recommending any practice ,i am proving that what I said is correct and what he said is wrong . Simple as that .

This isn't a court of law or a debating society, it's a forum that (hopefully) enables us to help each other get the most out of the ESP8266. Whether a particular person is right or wrong is less important than whether the information will help anyone else.

"GPIO2 can be left floating" - Yes, it can. My first EP-01 test board has been like this for about 9 months. I think it's established that it CAN be left floating. Your statement is correct. The operative word is can. (Martin, c'mon you can concede this much, yes?)

Should it be left floating? I would say no, and when designing/building a board for real use, it would be a false economy to leave off a 2 cent pullup resistor. All the schematics I've seen for commercially made test-boards for the ESP8266 include a pullup on GPIO2.

We have a fairly new, famous make clothes-dryer in the basement that I've added 3 or 4 extra pullup resistors onto the control board, because the designer relied on internal pullups, and the dryer control became erratic after a few months. It's been a very common failure - google "e68 error".

Please - let's consider this GPIO2 dispute resolved.

- Wed Aug 12, 2015 10:03 am #25792

This is expanding the question. The argument is either right or wrong . My statement is right , his is not . Simple as . I am not recommending any practice ,i am proving that what I said is correct and what he said is wrong . Simple as that .

Yes, it is as simple as that : you've NEVER quoted anywhere any evidence that shows that what I said is wrong. Since the beginning 7 weeks ago I said "leaving pin floating is simply not a good pratices" and you still argue like you did before ... Here are some of YOUR quotes (those that you didn't deleted like you did to remove traces) :

It appears that both might work and maybe the truth table should be ammended to "x" floating ?

GPIO2 - pull up (4k7)

I think this is wrong is should be floating or used for something else . When high on boot it pulls the chip into UART mode and thats not where you want it . See the wiki above on pin modes and connections

This last one is obviously ridiculous, it is contradicting the facts themselves from https://github.com/esp8266/esp8266-wiki ... ot-Process

And when I quoted above what Cal answered you 7 weeks, and recently, you are arguing that his english is not good enough to understand the basics of this discussion and come up now to the conclusion that Cal is wrong too ?

Moin tytower,

please accept that you are wrong with this. There is a difference between floating and connecting.
There is a difference between "don't care" (X) and "floating".
You are trying to help a lot of people but here you are indeed misleading.
A floating pin especially with some trace and possibly even a connector works like an antenna.
Using pull-ups/pul-downs help to work against catching a signal from the environment.
Take an oszilloscope to get an impression what difference a resistor makes.
Or build a darlington based detector to amplify the antenna effect to some led lighting.
If your hand is just getting nearer/away from the antenna you see a difference.

Please don't mislead people. Works for me is no replacement for application notes and rules.
I have been corrected recently here on an issue I believed I know and I based my tipps on.
I was wrong because the espduino guy did a clever adaption of the arduino stuff to espduino.
There is a lot to learn outside and the internet provided great new way of doing things.
But it does not change the rules of physics.
It frightens me if I see people just starting with electronis "designing" PCB layouts that are connected to the
house mains current without proper fusing, distances, etc. Or switching machines on and off using
an esp8266 with just an relais and no hardware/software protection schemes.
People will die because of this.

Cal

Simply Amazing !!!