ESP8266 shuts down some time later, why ?

Using Arduino General Electronics
1

Hey falks,

I have problems with my ESP8266-12E. It sometime shuts down. First it runs for some hours. But than it shuts down spontaniously. The error in detial:

I can measure a voltage on the VVC and GND Pins, but the IC is dead. Even the SSID of the ESP is not visible. No onBoard-LED. Nothing works. But if I press the reset button, than it works again for some minutes or hours.

Could it be, that the BuckConv sometimes failes ?

Or did I do something else wrongly ? Maybe the cap ?

Could it be because I put resin on the regulator to make it fixed on position 3.3V ?

I have turned the regulator exactly at 3.3V Output.

Unbenannt
Unbenannt1038×931 95.3 KB

2

most of the schematics you find in the internet are using resistors for the Pullups and Pulldowns on EN, 00, 02, 15...

http://sigmaprj.com/download-esp12e-adapter.html

please show a real picture of your wiring.

1 Like
3

The 1117 should be powered by your 6V, not the output of the 2596 (as shown).

1 Like
4

I am sorry, actually it is connected to the 6V... the first shematic will be corrected right now. wait...

but it still don't work dispite the new shematic.

Unbenannt
Unbenannt1038×931 95.3 KB

5

Naturally.
What about those missing pull-up resistors? Or is that another schematic oversight?

What about a simple blink sketch? Does it crater, too?

1 Like
6

As I said: it works very well, but it shuts spontaniously down completely.

It worked without those Resistors very well too since I am using a BuckConv.

Before BuckConv, I used an AMS1117-3.3 (see shematic). But I had to change it because of the USB-DEVICE.

I suppose the Problem on the BuckConv. I guess the ESP can't be run by an BuckConv... a friend here told me to use an BUCKCONV @HotSystems

Unbenannt2
Unbenannt2853×684 6.24 KB

7

You're on about your sketch while I'm asking about a simple thing. I'm trying to isolate a problem.

In the first place, that's how you see it; in the second, the latter has nothing to do with the former.

And everything was wonderful till you switched to the 'BUCKCONV'?

1 Like
8

I have to try the blink sketch... later today.

yes, it worked for years without any problem and without any resistors before switching from AMS1117-3.3 to a BuckConverter.

But I have some news: I guess there was a short on the circuit. I simply forget to cut off those legs from these electrolytic caps. And on the back of the PCB, there are many "Through-Holes" which connects some coppertrails with the front-pcb. I will test it out today and tommorrow and than I will report it here. Thanks.

9

I've a small update. I'm not done with testing it out for 24 hours or more. But in the mean time, I've heard some noice on the BuckConv:

YOU HAVE TURN INCREASE THE VOLUME TO HEAR IT. Before the videos ends, I pulled the USB-Device (Wifi Repeater) from the board, and you can clearly hear that the noise drops a little.

Note: I have put my phone mic right on top of the DC-DC-BuckConverter.

To be clear: the more current flows through the BuckConv, the noisier it gets. I don't know whether this is normal, or this is my problem.

I will wait until tommorrow or the next day and see what is going to happen.

10

Well, there you have your smoking gun. If everything else remained the same, that's where you should look.

Hearing a little sound from a buck converter sounds normal considering the way it works. Though I expect it to be too high a frequency to be audible, most work at frequencies such as 60-300 kHz.

Another issue: in your schematic I'm missing the required capacitors around the AMS1117-5.0. Also the 1V overhead seems to be too little, I recall it needs 1.3V minimum drop for regulating. Details on this you can find in the datasheet.

1 Like
11

Missing caps ? should it be like this ? But do I need caps at all ? Because I am connecting a TP-Link Wifi-Extender on it. I am asking not because I know better, but because I really don't know :smiley: I only thought the these caps only needed for powering ICs.

5v
5v873×279 2.41 KB

And here is the datasheet of the AMS:

12

Bad News. It has crashed again. I will try it out like this. Could maybe someone agree with my solution?

To be clear: it is the same code that worked for year. The only difference with these chrasehes is: I have replaced the AMS1117-3.3 with the BuckConv you see on this shematic. And again: I have to use a BuckConv because of the USB-Devices.

Now you can see my full plan. Yes, there is also an ESP-Cam. And no, I have to use both IC's.

crashed again
crashed again1206×1002 45.7 KB

13

Try an LM2596-based BuckConv.
Why can't that be adjusted to 3.3v and get rid of the linear 1117?

1 Like
14

I've adjusted it to 3.3V. To be exact: 3.34V. But it still crashes. I really do not understand, why it crashes. It worked many many hours after I fixed the problem with the legs which maybe caused a short. I thought that it caused a short. But it turned out, that this was not the problem. I really think, that the problem is the BuckConverter.

If you click on the following AliExpress, you can choose between a "good converter", and a "bad converter". I mean, is this for real ? Why are they marking it with "good" ?

https://www.aliexpress.com/item/1005005393995240.html

15

Wondering what the "6V Source" is.
Too bad you don't have a decent power supply ("bench supply", "lab supply", etc) that could be set for 3.3V

I think that these are better than what you are using
https://www.aliexpress.us/item/3256806021101003.html
I've used these modules successfully, but I haven't bought any from aliexpress.

I have an ESP8266 project that runs 24/7, but once a year or so it has a hiccup (that I cannot explain).

1 Like
16

Looking back at the thread and I really don't understand one thing.

You say you need 200 mA for some USB devices - which you try to power through an AMS1117-5.0 linear regulator.

To accommodate this 200 mA at 5V, you try to replace your 3.3V supply, but not the 5V supply. That part just doesn't make sense. It would make more sense to keep the 3.3V supply as you had it (with the AMS1117-3.3 regulator - which as you said worked just fine), and use a 5V buck regulator for the USB supply.

That said, most 5V regulators can deliver 200 mA just fine. If that 1V is enough overhead for the regulator (from 6V to 5V) you only dissipate 200 mW, which again is no problem for most regulators. Most won't even need a heat sink for that.

1 Like
17

It's a normal 6V 1A Power supply like this:

netzteil

May I ask you what kind of voltage regulator you are using ? Do you use this LM2596 Converter ?

18

Yeah, i thiught the same. The AMS 5.0 makes no sense. It would be better to have a 5V fixed BuckConverter which powers the AMS1117-3.3 and than to the ESP-12E.

But the problem is: I am using the ESP-Cam and also the ESP8266-12E. The are communicating. But the problem is: if I plug in both to the AMS1117-3.3, than the video-stream of the ESP-Cam will flicker like this:

27904a17b536fff941f4f85546230531a2873dac_2_366x500

I also opened an topic here for the flickering:

So, there has to be a buck converter :smiley:

19

@runaway_pancake

I guess the noise was the problem. My ESP-12E is now damaged. Now I've placed a "5V-Fixed-Buck-Converter". Now there is no noise. I will test it out constantly for 3 days, including a Camera-Stream which will also run for 3 days. Also the on-board-LED. Also the Wifi Repeater which consumes arraund 200mA.

So far, there is no flickering on the ESP-CAM. And I hope the ESP8266 will not shut down within these 3 days

But clearly: there is absolutely no noise on the Buck-Converter. I have placed this on my PCB (the 5V):

https://de.aliexpress.com/item/1005003332040548.html?

20

Where ?