Your new topic does not fit any of the above??? Check first. Then post here. Thanks.

Moderator: igrr

User avatar
By tinus88
#45270 Hello,

I am experiencing problems when performing a system_restart();
Most times the system will boot fine, but about 1 or 2 out of 10 times it will freeze after a restart.
On the terminal I see the following:

Code: Select all2 Minutes have passed, time to restart system out of precaution  <-- my own code
 ets Jan  8 2013,rst cause:2, boot mode:(3,6)
 ets Jan  8 2013,rst cause:4, boot mode:(3,6)
wdt reset

Then nothing happens and I have to repower my hardware.
Please help!

Background:
- For testing purpose I have set a timer for 2 minutes to restart the system. In the field this will probably be several hours.
- I am using software prepared for OTA. I am using boot_v1.5.bin as bootloader in the field. For testing I have also tried boot_v1.4(b1).bin with no difference.
- I am using a ESP-12e module, I have a 15k pull-up on GPIO0 (ofcourse with a button to ground for programming), 15k pull-up on GPIO2, 15k pull-down on GPIO15, 15k pull-up on reset, 15k pull-up on CH_PD with 100n to ground (as a power-on-reset function). On all of these pins nothing else connected than described.
- For the power supply I use a LDO which can source 800mA, decoupled with two 100n and one 2u2 capacitors.
User avatar
By tinus88
#45364 Hi,

I am afraid the link you posted describes a very similar, yet another issue.
They are having a problem with a WDT reset freeze after a system_restart() without a power-cycle after serial flashing.
This specific problem is acknowledged by igrr and clarifies that this isn't the case with OTA updating.
I do not have a problem with a freeze after serial flashing, but after OTA updating.

I have several hundreds of modules, all showing the same result.
I also have a nodeMCU dev board v1 (https://github.com/nodemcu/nodemcu-devk ... T_V1.0.PDF), which hasn't showd the problem for about 4 hours (system_restart()-ing every 2 minutes).
In hardware I don't see any major differences, except for:
- a bigger 3V3 capacitor (they use 100uF). For testing I have put a 220uF on the 3V3, problems still exists;
- the pull-up resistors (they use 12k, I am using 15k). I don't expext this to be the cause;
- power-on-time-hardware. They use a pull-up with 100nF in series to ground, and feed this into reset-input. I do the same but feed it into CH_PD-input.