ESP32 lolin S2 mini REBOOT since I upgraded from 24.6.29 to 24.10.6

[SuperNinja73]

Hi all !
The hardware:
I have 2x HMS 2000-4T that work perfectly with OpenDTU v24.6.29 installed on ESP32 lolin S2 mini:

, and CMT2300A for the radio signal. It is connected to a USB port of a PC running Windows 10, 24/24 hours.

What does not work:
I updated to 24.10.6, and after several minutes of good operation, it reboots every second: we hear the characteristic sound of the USB plug unplugging and re-plugging without stopping and I lose the OpenDTU server.
I unplug/replug my USB cable and it works again for several minutes and then reboots without stopping again.

For it to work:
I put back the version 24.6.29 and no problem reported for several hours of good operation.

How:
to update the ESP32 lolin S2 mini in OTA, I use platformio to compile a firmware.bin by cloning the Git Repository, then in OpenDTU, I go to Settings > Firmware upgrade and I choose my firmware.bin

Anyone has the same problem? Or a track to explore?

[ms49434]

Please use a decent power supply, do not use a charger and do not power it by a computer.

[SuperNinja73]

I am aware of that, but why with the version before 24.6.29 I have no reboot, not even a disconnection, nothing at all for 3 months. And there with 24.10.6 after about 30 minutes it reboots without stopping.
I put the old version back this morning, and it has not crashed since 8 hours that it has been running.
I was thinking about this topic which had the same problem as me : #2314

[schneeer]

Hello.
I claim that the problem only exists for you, unless the firmware only runs on a few ESP32s.
You should definitely try the tip from @ms49434. Simply to rule out problems with the power supply.
You should also check if there is a capacitor near the CMT at the supply voltage. If not, adding a capacitor of around 100µF may help.

Doku Troubleshooting: You can try to use an USB power supply with 1 A or more instead of connecting the ESP32 to the computer.

[LennartF22]

Can you post the output on the serial console (via USB) just before the first reset happens?

[LennartF22]

Do you have a USB to serial converter (or something like a Raspberry Pi Pico, which could be used as one) that you could hook up to the ESP32 directly? The log might not be complete, since the USB serial probably resets too early, so that potential error messages are lost.

[SuperNinja73]

I have a Node Red server running on this PC and I have monitored the COM7 port:

I'm afraid I don't have more details. Here the COM7 port was deactivated 10 minutes after the end of the AP, and reboots without stopping. 😕

[SuperNinja73]

I am testing a higher version v24.8.1 than the one that works (v24.6.9): and it becomes more talkative:

it has been online for 40 minutes

I really don't think it's a power supply problem...

[LennartF22]

I don't think it's a power supply problem either. If hooking up a USB serial converter to the ESP32 is not an option, the next best thing to narrow the problem down would be to try other firmware version between v24.8.1 and v24.10.6 to eventually find the exact version where the problem begins.

[SuperNinja73]

yes i'm agree. I will have to compile the following versions and come across the defective one. For info v24.8.1 still works this morning without problems ... to be continued

[SuperNinja73]

I tried the version v24.9.26 that worked for 1 hour and then disconnected. I did a reset using the integrated key and noticed that it no longer found the saved version???

I continue to test the versions...

[SuperNinja73]

there I went back a little with v24.9.21

{
  "hostname": "OpenDTU-4DC49C",
  "sdkversion": "v4.4.7-dirty",
  "cpufreq": 240,
  "cputemp": 35.37,
  "heap_total": 165428,
  "heap_used": 106824,
  "heap_max_block": 32756,
  "heap_min_free": 49152,
  "psram_total": 2094607,
  "psram_used": 23844,
  "sketch_total": 1966080,
  "sketch_used": 1582752,
  "littlefs_total": 196608,
  "littlefs_used": 24576,
  "chiprevision": 0,
  "chipmodel": "ESP32-S2",
  "chipcores": 1,
  "flashsize": 4194304,
  "resetreason_0": "Software reset digital core",
  "resetreason_1": "Software reset digital core",
  "cfgsavecount": 17,
  "config_version": "0.1.28",
  "git_hash": "g0000000",
  "pioenv": "lolin_s2_mini",
  "uptime": 1799,
  "nrf_configured": false,
  "nrf_connected": false,
  "nrf_pvariant": false,
  "cmt_configured": true,
  "cmt_connected": true
}

Should I try a factory flash?

[LennartF22]

A factory flash would not change anything. Judging by the path of your build, you are probably just downloading the source for earch version manually, instead of using Git, right? In that case, the displaying of the Git hash cannot work. However, this is not an issue right now regarding the actual problem you have (it only affects the displaying of the version and the update checker).

If you know how to use Git, you can clone this repository and "checkout" the relevant versions to make the displaying of the version work, but you can just as well continue doing it the way you're doing it now.

[SuperNinja73]

indeed there is a difference when I compile from the Git clone or if I download the source code.zip:
with this last one the version is at 00000
I must be making a mistake somewhere? Here is how I compile from source:
1- I DL the Source code (zip) and unzip
2- in platformio I open the folder "OpenDTU-24.9.20"
3- I click at the bottom to select from the drop-down list: "env:lolin_s2_mini" :

4- it updates
5- I "build" : i got 2 "yelow attention":

and

But finaly a good compilation :

6- I update the firmware from the web interface and get no version :

To get back to the subject of this post, the version that crashes the ESP32: 24.10.6 was compiled from the Git clone, the one that displays the version correctly. I then downloaded the source code.zip and I think that's where I no longer have the version numbers.

[LennartF22]

The reason for the missing version number is that during the build process, it is generated from the current commit (or tag) that is checked out in the local Git repository. In the source archives, there is no Git information, so there is nothing to take this information from.

You are doing everything correctly. If you want the version number to be displayed, you must check out the version you want to build via Git.

[tbnobody]

Please keep in mind, that version v24.8.1 till v24.9.27 contains a error which reboots the device after ~10min if the "anonymous readonly access" was disabled. It was fixed in v24.9.30

[SuperNinja73]

@LennartF22 : ok so I'm going to choose the Git clone:
@tbnobody : ok i just upgraded from v24.9.22 to v24.9.30 and the CMT is disconnected

Websocket: [/livedata][1] disconnect
19:43:15.372 > Admin AP remaining seconds: 150 / 180
19:43:25.383 > Admin AP remaining seconds: 160 / 180
19:43:35.477 > Admin AP remaining seconds: 170 / 180
19:43:45.424 > Admin AP remaining seconds: 180 / 180
19:43:46.411 > Admin mode disabled

while I was well connected with v24.9.22

the config is saved correctly:

[LennartF22]

But it's connected when using v24.10.6 (I think that's the version you initially tried)?

[SuperNinja73]

ok , let's see what happens with the clone of Git v24.10.6:

CMT is disconected !

I was able to capture an interesting sequence on the COM7 port : it connected, then wifi disconnected, followed by MQTT, and then COM7. It went into "AP" mode and then the ESP32 rebooted every second

[SuperNinja73]

here it is! I tested all versions up to the one that crashes:
v24.9.26 is ok: the CMT is well connected and remains stable.
When I switch to 24.9.27: the CMT is disconnected and goes into AP mode to end up rebooting every second. 24.9.30 and 24.10.6 do the same thing!
@tbnobody

Please keep in mind, that version v24.8.1 till v24.9.27 contains a error which reboots the device after ~10min if the "anonymous readonly access" was disabled. It was fixed in v24.9.30

I am logged in as "admin xxxxx", no one else accesses the OpenDTU server so I shouldn't have any reboot issues ?

I am forced to stay on 24.9.26 for now. I'm waiting to see for more than an hour if it doesn't reboot... to be continued

[LennartF22]

In v24.10.6, please try changing

    // Initialize SpiManager
    SpiManagerInst.register_bus(SPI2_HOST);
#if SOC_SPI_PERIPH_NUM > 2
    SpiManagerInst.register_bus(SPI3_HOST);
#endif

to

    // Initialize SpiManager
    SpiManagerInst.register_bus(SPI3_HOST);
#if SOC_SPI_PERIPH_NUM > 2
    SpiManagerInst.register_bus(SPI2_HOST);
#endif

and see whether the CMT is connected when you build/flash the modified version.

[SuperNinja73]

that's better :

I let it run to see the long-term stability...

[LennartF22]

@SuperNinja73 As it turns out, on the ESP32-S2, SPI3 shares the DMA controller with the ADC/DAC, which causes issues when both are used in parallel. To confirm my theory it would be great if you could revert the change above, make sure that the CMT connection is broken again (by temporarily flashing it with the original code), and then do the following change in lib/SpiManager/src/SpiBus.cpp:

Before:

    ESP_ERROR_CHECK(spi_bus_initialize(host_device, &bus_config, SPI_DMA_CH_AUTO));

After:

    ESP_ERROR_CHECK(spi_bus_initialize(host_device, &bus_config, SPI_DMA_DISABLED));

[LennartF22]

Oh, and you must also change

        .max_transfer_sz = SPI_MAX_DMA_LEN,

to

        .max_transfer_sz = 0,

in the same file.

[SuperNinja73]

ok I'll go back to the original 24.10.6 and change the lines you told me tonight when I get home from work... I'm going back to the SPI2-SPI3 modification you made me do: which worked well early in the evening and unfortunately it crashed last night!
Do you think there is a link with the conflict between the SPI and the ADC/DAC?

[LennartF22]

I created issue #2343 for this.

[SuperNinja73]

@LennartF22
1 - I went back to the original 24.10.6 and I confirm that no Rx Tx is displayed on the console
2- I will change the lines in
lib/SpiManager/src/SpiBus.cpp . Should I also change the if /endif// Initialize SpiManager above ? Come on! I'll change that too!

Result:
it works: the CMT connects, with a little difficulty at the beginning:

Again, I'm not going to cry victory too quickly and wait a few hours of operation to break out the bottle of champagne. 🍾 lol

[SuperNinja73]

after 1 hour of operation, I did not expect a RX fail rate:

... to be continued

[SuperNinja73]

after 1 hour of additional operation, it crashes. It lasted 2 hours with all the modifications you asked me to ... I am desperate !
I will go back to 24.9.21 which worked well 😞

[LennartF22]

I'm afraid there is not much we can do right now without someone who also has an ESP32-S2 + CMT and can hook up a serial adapter or JTAG probe to it. Do you mind posting you config.json (with the WiFi SSID and password removed) and pin_mapping.json here, so that we can (eventually) try to replicate your issue as closely as possible?

@SuperNinja73 Thank you for your efforts so far. It helped us finding at least one bug, even though we are not fully there yet apparently.

@tbnobody @schlimmchen @stefan123t Maybe someone of you has an ESP32-S2?

[schlimmchen]

Negative.

[SuperNinja73]

@LennartF22 sure ! pin_mapping.json

[
    {
        "name": "NRF, LEDs, Display",
        "nrf24": {
            "miso": 19,
            "mosi": 23,
            "clk": 18,
            "irq": 16,
            "en": 4,
            "cs": 5
        },
        "display": {
            "type": 3,
            "data": 21,
            "clk": 22
        },
        "led": {
            "led0": 25,
            "led1": 26
        }
    },
    {
        "name": "CMT, LEDs, Display",
        "nrf24": {
            "miso": -1,
            "mosi": -1,
            "clk": -1,
            "irq": -1,
            "en": -1,
            "cs": -1
        },
        "cmt": {
            "clk": 18,
            "cs": 4,
            "fcs": 5,
            "sdio": 23,
            "gpio2": 19,
            "gpio3": 16
        },
        "display": {
            "type": 3,
            "data": 21,
            "clk": 22
        },
        "led": {
            "led0": 25,
            "led1": 26
        }
    },
    {
        "name": "NRF, Display",
        "nrf24": {
            "miso": 19,
            "mosi": 23,
            "clk": 18,
            "irq": 16,
            "en": 4,
            "cs": 5
        },
        "display": {
            "type": 3,
            "data": 21,
            "clk": 22
        }
    },
    {
        "name": "CMT, Display",
        "nrf24": {
            "miso": -1,
            "mosi": -1,
            "clk": -1,
            "irq": -1,
            "en": -1,
            "cs": -1
        },
        "cmt": {
            "clk": 18,
            "cs": 4,
            "fcs": 5,
            "sdio": 23,
            "gpio2": 19,
            "gpio3": 16
        },
        "display": {
            "type": 3,
            "data": 21,
            "clk": 22
        }
    },
    {
        "name": "NRF, LEDs",
        "nrf24": {
            "miso": 19,
            "mosi": 23,
            "clk": 18,
            "irq": 16,
            "en": 4,
            "cs": 5
        },
        "led": {
            "led0": 25,
            "led1": 26
        }
    },
    {
        "name": "CMT, LEDs",
        "nrf24": {
            "miso": -1,
            "mosi": -1,
            "clk": -1,
            "irq": -1,
            "en": -1,
            "cs": -1
        },
        "cmt": {
            "clk": 36,
            "cs": 34,
            "fcs": 33,
            "sdio": 35,
            "gpio2": -1,
            "gpio3": -1
        },
        "led": {
            "led0": -1,
            "led1": 15
        }
    },
    {
        "name": "NRF",
        "nrf24": {
            "miso": 19,
            "mosi": 23,
            "clk": 18,
            "irq": 16,
            "en": 4,
            "cs": 5
        }
    },
    {
        "name": "CMT",
        "nrf24": {
            "miso": -1,
            "mosi": -1,
            "clk": -1,
            "irq": -1,
            "en": -1,
            "cs": -1
        },
        "cmt": {
            "clk": 18,
            "cs": 5,
            "fcs": 4,
            "sdio": 23,
            "gpio2": -1,
            "gpio3": 15
        }
    }
]

config.json

{"cfg":{"version":72704,"save_count":17},"wifi":{"ssid":"Bbxxxxxxxx","password":"8xxxxxxxxxx","ip":"0.0.0.0","netmask":"0.0.0.0","gateway":"0.0.0.0","dns1":"0.0.0.0","dns2":"0.0.0.0","dhcp":true,"hostname":"OpenDTU-%06X","aptimeout":3},"mdns":{"enabled":false},"ntp":{"server":"opendtu.pool.ntp.org","timezone":"CET-1CEST,M3.5.0,M10.5.0/3","timezone_descr":"Europe/Paris","latitude":51.16569901,"longitude":10.45149994,"sunsettype":1},"mqtt":{"enabled":true,"hostname":"192.168.1.54","port":1883,"clientid":"OpenDTU-16229348","username":"","password":"","topic":"solar/","retain":true,"publish_interval":5,"clean_session":true,"lwt":{"topic":"dtu/status","value_online":"true","value_offline":"offline","qos":2},"tls":{"enabled":false,"root_ca_cert":"-----BEGIN CERTIFICATE-----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\n-----END CERTIFICATE-----\n","certlogin":false,"client_cert":"","client_key":""},"hass":{"enabled":false,"retain":true,"topic":"homeassistant/","individual_panels":false,"expire":true}},"dtu":{"serial":28147069391688,"poll_interval":5,"nrf_pa_level":0,"cmt_pa_level":0,"cmt_frequency":865000000,"cmt_country_mode":0},"security":{"password":"bxxxxxxxxxxxxxxx","allow_readonly":true},"device":{"pinmapping":"CMT, LEDs","display":{"powersafe":true,"screensaver":true,"rotation":2,"contrast":60,"language":0,"diagram_duration":36000,"diagram_mode":1},"led":[{"brightness":100},{"brightness":100}]},"inverters":[{"serial":0,"name":"","order":0,"poll_enable":true,"poll_enable_night":true,"command_enable":true,"command_enable_night":true,"reachable_threshold":2,"zero_runtime":false,"zero_day":false,"clear_eventlog":false,"yieldday_correction":false,"channel":[{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0}]},{"serial":19123647046167,"name":"cuisine_droite","order":0,"poll_enable":true,"poll_enable_night":true,"command_enable":true,"command_enable_night":true,"reachable_threshold":2,"zero_runtime":false,"zero_day":true,"clear_eventlog":false,"yieldday_correction":false,"channel":[{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0}]},{"serial":19123647046160,"name":"cuisine_gauche","order":1,"poll_enable":true,"poll_enable_night":true,"command_enable":true,"command_enable_night":true,"reachable_threshold":2,"zero_runtime":false,"zero_day":true,"clear_eventlog":false,"yieldday_correction":false,"channel":[{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0}]},{"serial":0,"name":"","order":0,"poll_enable":true,"poll_enable_night":true,"command_enable":true,"command_enable_night":true,"reachable_threshold":2,"zero_runtime":false,"zero_day":false,"clear_eventlog":false,"yieldday_correction":false,"channel":[{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0}]},{"serial":0,"name":"","order":0,"poll_enable":true,"poll_enable_night":true,"command_enable":true,"command_enable_night":true,"reachable_threshold":2,"zero_runtime":false,"zero_day":false,"clear_eventlog":false,"yieldday_correction":false,"channel":[{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0}]},{"serial":0,"name":"","order":0,"poll_enable":true,"poll_enable_night":true,"command_enable":true,"command_enable_night":true,"reachable_threshold":2,"zero_runtime":false,"zero_day":false,"clear_eventlog":false,"yieldday_correction":false,"channel":[{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0}]},{"serial":0,"name":"","order":0,"poll_enable":true,"poll_enable_night":true,"command_enable":true,"command_enable_night":true,"reachable_threshold":2,"zero_runtime":false,"zero_day":false,"clear_eventlog":false,"yieldday_correction":false,"channel":[{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0}]},{"serial":0,"name":"","order":0,"poll_enable":true,"poll_enable_night":true,"command_enable":true,"command_enable_night":true,"reachable_threshold":2,"zero_runtime":false,"zero_day":false,"clear_eventlog":false,"yieldday_correction":false,"channel":[{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0}]},{"serial":0,"name":"","order":0,"poll_enable":true,"poll_enable_night":true,"command_enable":true,"command_enable_night":true,"reachable_threshold":2,"zero_runtime":false,"zero_day":false,"clear_eventlog":false,"yieldday_correction":false,"channel":[{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0}]},{"serial":0,"name":"","order":0,"poll_enable":true,"poll_enable_night":true,"command_enable":true,"command_enable_night":true,"reachable_threshold":2,"zero_runtime":false,"zero_day":false,"clear_eventlog":false,"yieldday_correction":false,"channel":[{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0},{"name":"","max_power":0,"yield_total_offset":0}]}]}

[devdems]

@SuperNinja73 i also have one S2 mini that i would like to use in openDTU. Can you share wiring diagram?

[SuperNinja73]

here is my pin mapping :
CMT : lolin S2 mini:
11 clk gpio36 SPI clock pin
13 csb gpio34 SPI register accesser
14 fcsb gpio33 SPI FIFO accesser
12 sdio gpio35 SPI data pin
you still have to add the 3.3v and the ground

after putting a yellow protective film, it looks like this:

[SuperNinja73]

@SuperNinja73 i also have one S2 mini that i would like to use in openDTU. Can you share wiring diagram?

@devdems where are you? have you wired your S2?

[devdems]

@SuperNinja73, I ordered the fusion board, which already includes antennas so I don't need to buy them separately: https://shop.allianceapps.io/products/opendtu-fusion-community-edition. I will put this in an IP65 box outside, close to the inverter.

[F1zye]

@SuperNinja73 : you could try increasing "poll_interval", e.g. from 5 to 20

[SuperNinja73]

can you provide the lines before your snippet above?

it was with the PC COM port (not the usb/uart adapter). In this capture we see that it suddenly stopped, losing the wifi connection and rebooting

[tbnobody]

This does not help. Please provide the serial output of the USB/UART adapter before the reboot. Otherwise anything else is just guessing without any result

[SuperNinja73]

I increase the buffer of my terminal

[SuperNinja73]

here is the log of the transition from normal operation to crash:

Fetch inverter: 116492306210
TX RealTimeRunData 865.00 MHz --> 15 92 30 62 10 80 12 93 48 80 0B 00 67 0F 98 F2 00 00 00 00 00 00 00 00 66 7E 1D 
Interrupt received
RX 865.00 MHz --> 95 92 30 62 10 80 12 93 48 01 00 01 01 AD 01 A5 00 9A 00 BB 02 97 03 14 00 05 A2 | -62 dBm
Interrupt received
RX 865.00 MHz --> 95 92 30 62 10 80 12 93 48 02 0C 1B 00 05 67 FA 00 6F 00 82 01 A4 01 A5 00 BA D7 | -61 dBm
Interrupt received
RX 865.00 MHz --> 95 92 30 62 10 80 12 93 48 03 00 BA 03 11 03 12 00 05 69 80 00 05 6C 8F 00 82 3E | -61 dBm
Interrupt received
RX 865.00 MHz --> 95 92 30 62 10 80 12 93 48 04 00 82 08 9E 13 89 0B 39 00 02 00 82 03 E7 00 FE 2E | -61 dBm
Interrupt received
RX 865.00 MHz --> 95 92 30 62 10 80 12 93 48 85 00 01 85 D9 D4 | -61 dBm
RX Period End
Success
Guru Meditation Error: Core  0 panic'ed (Interrupt wdt timeout on CPU0). 

Core  0 register dump:
PC      : 0x400f32e8  PS      : 0x00060134  A0      : 0x800f34cd  A1      : 0x3ffe1370  
A2      : 0x01c19000  A3      : 0x01000000  A4      : 0x80000020  A5      : 0x00000005  
A6      : 0x00000001  A7      : 0x00000000  A8      : 0x3f408800  A9      : 0x3ffce720  
A10     : 0x00060123  A11     : 0x3ffdf3d0  A12     : 0x80028504  A13     : 0x3ffc55e0  
A14     : 0x3ffc56d4  A15     : 0x00000001  SAR     : 0x00000009  EXCCAUSE: 0x00000005  
EXCVADDR: 0x00000000  LBEG    : 0x80028504  LEND    : 0x3ffc55e0  LCOUNT  : 0x40026b49  

Backtrace: 0x400f32e5:0x3ffe1370 0x400f34ca:0x3ffe1390 0x400d1401:0x3ffe13b0 0x401444f6:0x3ffe13d0 0x4002b1fb:0x3ffe13f0 0x4013bc2b:0x3ffe1410 0x40172107:0x3ffe1450

ELF file SHA256: 3bf00f81b45b9d4d

E (6437) esp_core_dump_flash: Core dump flash config is corrupted! CRC=0x7bd5c66f instead of 0x0
Rebooting...
ÿTest 123456

Starting OpenDTU
Initialize FS... done
Reading configuration... done
Reading PinMapping... found valid mapping done
Initialize Network... Guru Meditation Error: Core  0 panic'ed (Interrupt wdt timeout on CPU0). 

Core  0 register dump:
PC      : 0x400f32ee  PS      : 0x00060734  A0      : 0x800f34cd  A1      : 0x3ffe12e0  
A2      : 0x01c19000  A3      : 0x01000000  A4      : 0x01050b6b  A5      : 0x0000006b  
A6      : 0x00000b00  A7      : 0x00050000  A8      : 0x3f408800  A9      : 0x3ffce720  
A10     : 0x00060720  A11     : 0x00000001  A12     : 0x00060021  A13     : 0x6000e130  
A14     : 0x6000e138  A15     : 0x6000e140  SAR     : 0x00000000  EXCCAUSE: 0x00000005  
EXCVADDR: 0x00000000  LBEG    : 0x00060021  LEND    : 0x6000e130  LCOUNT  : 0x40026b49  

Backtrace: 0x400f32eb:0x3ffe12e0 0x400f34ca:0x3ffe1300 0x400d1401:0x3ffe1320 0x401444f6:0x3ffe1340 0x40146a07:0x3ffe1360 0x40102313:0x3ffe1380 0x4010236e:0x3ffe13b0 0x400ddccf:0x3ffe13d0 0x4012ad46:0x3ffe13f0 0x4012b4f3:0x3ffe1410 0x40129b46:0x3ffe1430 0x40172095:0x3ffe1450

ELF file SHA256: 3bf00f81b45b9d4d

E (1125) esp_core_dump_flash: Core dump flash config is corrupted! CRC=0x7bd5c66f instead of 0x0
Rebooting...
ÿTest 123456

[SuperNinja73]

Ich meine im Text herausgelesen zu haben, das das eine oder andere ähnliche Board funktioniert.
Daher wollte ich den Topicstarter darauf hinweisen, doch ein anderes Netzteil zu probieren.
Ich habe übersehen, das du das eigentlich geschrieben hattest, was ich zitiert habe.
Trotzdem: man weiß erst hinterher, ob der Netzteiltest sinnlos war.
Mit etwas Bastelei bekommt man ein Netzteil auch mit, bzw. trotz UART-Adapter angeschlossen.

still on the same software v24.10.6 modified by @tbnobody , I power the ESP with a usb-c cable + quality mains transformer :

and I can't even launch OpenDTU, the USB/UART disconnects after a few seconds after booting :

console didn't display nothing

and nothing happen on the dashboard...Another WIFI/MQTT disconnection

i reset the ESP and working 15minutes then crash again

I reinstalled v24.9.21 and, of course I no longer have the uart on pin1, but OpenDTU works normally: the console indicates the first returns from the HMS in 30s and no crash!

It's definitely not a power problem.

[tbnobody]

Could you please test this version?

20241016_lolin_s2_mini.zip

[tbnobody]

Again... The ESP cannot never ever have any influence in your usb/uart adapter. Thats something completly different. But lets see if it crashes again

[SuperNinja73]

Again... The ESP cannot never ever have any influence in your usb/uart adapter. Thats something completly different.

I would be curious to see if activating the uart on pin1 in "my favorite lol" version v24.9.21 disconnects or not?

But lets see if it crashes again

👍 It's been 1 hour playing without crashing... wait and see.

I notice that the temperature is not uploaded to the MQTT broker...

[tbnobody]

I notice that the temperature is not uploaded to the MQTT broker...

Yes, thats by intention... Seems to have something todo with the reading of the temperature.

[F1zye]

For me, above version works without crashing for more than 11hrs (lolin s2, NRF24L01+, 2 HM), while any other version since v24.9.22 did crash after a couple of hours.

[SuperNinja73]

@F1zye Great! This latest version worked all night for me too. @tbnobody I think we are on the right way. 🥳

[SuperNinja73]

@tbnobody a question: v24.10.15 was released 3 days ago, does it include the fix for our problem on the S2? Because I just tested it and it doesn't work: it doesn't connect to the HMS.
In the meantime v24.10.15-2-g217192b has been working perfectly for 2 days.

[SuperNinja73]

ok i'm waiting for the next update ? perhaps ? 😋

[SuperNinja73]

No, there is no "full" fix in the master yet. There are actually multiple issues, one small change arrived in the master already, but that alone won't help.

And this release v24.11.7 ?

[F1zye]

I'd say no - at least v24.11.7 crashed on my lolin s2 mini.
Would it make sense to remove temperature readings for the s2 as a temp fix to master?

[SuperNinja73]

I'd say no - at least v24.11.7 crashed on my lolin s2 mini.

Thanks for this feedback. We still have to wait.

[F1zye]

I recompiled v24.11.7 with "CpuTemperature.read()" commented out in "WebApi_sysstatus.cpp" and "MqttHandleDtu.cpp" (all occurrences) and it works without crash - just in case you want to move to the latest version.