WIFI Drops - Guru Meditation Error: Core 0 panic'ed

I’m having trouble with my M4. I’ve got one of the new control boards with the JST connectors. I kept having issues with the stability of the wifi when using the M4 as an access point, and so I connected my handy usb-c cable up to see if I could get a better vantage point on what was going on.
I didn’t see much to explain the issues I was facing, but I noticed that sometimes the terminal would be responsive and other times it wouldn’t respond. So, I thought based upon what I saw in the forums, it might be time to try to push the full-install.bat file. That didn’t work as well as I hoped. When I ran that file I would get a messages about how the serial connection was unresponsive, or that a fatal error had occurred, but then the batch file would just continue on running the next batch file, until it finished by launching the terminal. So, I decided to run each batch file on it’s own and try to run each step until it completed successfully. I was eventually able to get that to work. At that point I was having fewer times the serial connection would stop being responsive, so I went ahead and ran the full-install.bat file again just to see what happened. It installed fine.
But all this didn’t fix my issue. The wifi was still unstable / the unit seemed to be rebooting itself.

I regularly got an info message about the nvm not being found or something like that. Since it was informational, I dismissed it. Then I started to get these warning messages one after another “Low Memory: 14248 bytes”, then 11728, 11532, 9012, 6564, 4044, 3720, 3592, 2664 bytes. Then I pressed enter to see if the terminal was still responsive, and it responded ok. Then I got this "Guru Meditation Error: Core 0 panic’ed (InstrFetchProhibited). Exception was unhandled.
It produced a register dump, then rebooted itself.
So, I’m wondering if this has been the problem all along.
I’m sure there are a bunch of details that I’ve left out that would be helpful.
Thoughts?




1 Like

I haven’t seen that exact behavior, but I did see something similar on early versions of the firmware where a loss of wifi connection could cause the ESP32 to crash.

That was well before 0.82 though.

Do you by any chance have a different device (a phone perhaps?) that you could try connecting from? That would tell us if the issue is something to do with the way that device is communicating with the machine or if it is unrelated.

Yeah, I’ve tried from three different computers. 2 Macs, 1 Windows. All using Chrome. On the Macs I also used Safari. On the windows I also used Edge.

The most stable of the three in connecting is the Mac running OS 14.6.1.

1 Like

So when I came in today, I was able to connect to the Maslow network, access the control webpage, retract the cables, extend the cables, and then start a configuration without a wifi drop. Then during the calibration, I got these errors: (full serial log attached)
[MSG:ERR: Unable to move safely, stopping calibration]
[MSG:ERR: Emergency stop! Stopping all motors]
[MSG:WARN: The machine will not respond until turned off and back on again]
[MSG:ERR: Unable to move safely, stopping calibration]

So, I plugged in my handy usb cable again to see if I could get an further info. The serial communication was responsive, but there wasn’t any other info to be had.

Then I rebooted the Maslow as directed. Very quickly the serial connection showed me that it had booted up, and the wifi settings had been loaded. The only error was this one “[MSG:INFO: Error ESP_ERR_NVS_NOT_FOUND reading from NVS!]”

What I noticed though was that while from fluidterm it looked like the wifi should be running, it was more than 5 minutes before I could actually see the wifi from my computer (refreshing my wifi regularly). Then once I was able to connect to the Maslow wifi, it was very slow to open up the webpage, when it finally did I got a popup box with a red triangle with an exclamation point in it and the error “No connection”, followed by another popup box that had an error with similar wording but asked if the firmware was up to date.

From there I tried refreshing the webpage, and it did, but the serial log on the webpage was strangely sparse and the Maslow didn’t respond to any of the commands I gave it in the setup screen.

I then got a series of low memory warnings and then this error:
[MSG:WARN: Low memory: 4532 bytes]
[MSG:WARN: Low memory: 4208 bytes]
[MSG:WARN: Low memory: 3304 bytes]
[MSG:ERR: mount_to_vfs failed code 0x0x101]

Not sure if any of this is helpful in diagnosing, but I’m at a loss as to where to go next. Thanks!

Maslow-serial (1).log (2.7 KB)

1 Like

Fortunately this looks totally normal to me.

I just put together this list here of common error messages and what they mean:

The unable to move one there is related to the initial values for the calibration process, try adjusting the initial guess for the frame and you should be good to go.

It’s good to know that page exists. For some reason that didn’t come up in my searches for the error codes. Thanks for that.

Before it errored out it calculated a fitness value of 0.5073895709726891. I’m also within 1/4" of the actual frame pin locations. (My frame is my concrete shop floor. I just put in some drop in anchors and I use knob bolts to secure.)

Even though I was able to get connected for a bit this morning, since then it continues to drop wifi connection and reboot. So much so that I can’t even complete a calibration. Any ideas what else could be causing these issues? Is there a way I can have it write a detailed log file to the sd file space, so that when it errors out and reboots I have a better record of what is going on at the time?

1 Like

I just wrote that page yesterday at someone else’s suggestion so your searching wasn’t wrong :upside_down_face:

Are you using the Ethernet controller or the JST-XH controller?

I’d like to send you a replacement to try in case it’s a hardware issue with the ESP32.

If it’s the Ethernet board we have them in stock, but the JST ones I won’t have until next week.

I’m using a JST controller that I ordered a few weeks ago.

1 Like

Since you have a second controller board it might be worth trying to plug just the encoders and power into the Ethernet control board and see if the connection is stable there. That would let us figure out if the issue is with the board or if it’s something environmental

In the UI code - specifically tablet.js - you’ve also got these:

  if (msg.startsWith('error:')) {
    const msgExtra = {
      "8": " - Command requires idle state. Unlock machine?",
      "152": " - Configuration is invalid. Maslow.yaml file may be corrupt. Turning off and back on again can often fix this issue.",
      "153": " - Configuration is invalid. ESP32 probably did a panic reset. Config changes cannot be saved. Try restarting",
    };

    msg += msgExtra[msg.split(":")[1]] || "";
  }

  //Catch the calibration complete message and alert the user
  if(msg.startsWith('[MSG:INFO: Calibration complete')){
    alert('Calibration complete. You do not need to do calibration ever again unless your frame changes size. You might want to store a backup of your maslow.yaml file in case you need to restore it later.');
  }
1 Like

I’m having this same problem. Wondering if new ESP32 fixed the problem. Maslow is 15 feet from my Linksys router. I’ve tried local AP and connecting it to my wifi. Maslow seems to drop connection.

I cannot even reliably connect to UI. Maslow blue light will blink out IP address and Linksys router will show it connected.

I have a DigUno with a ESP32 60 ft in my backyard for Christmas lights and wifi works on that.

  1. Is there a way to “flash” or reset PCB to “factory.”

Frustrated Steve

If you move the maslow out of range of the wifi, when it fails to connect it
will create it’s own network again and you can connect to it and remove the
setting.

We have seen that the maslow has very poor wifi range.

David Lang

I have moved Maslow out of range/turn off home wifi to get to Maslow WiFi and use password 12345678.

Even being next to Maslow directly connected to Maslow AP…wifi seems to fail/drop every 10-20 sec.

1 Like

Do you have any other devices that you could try connecting from? Just to confirm that the issue is on the ESP32 side

I was trying to connect from my phone or my laptop. Sometimes phone seemed to connect easier. How many devices can connect to maslow.local? I was wondering if i had too many devices trying to connect and creating an error that way.

I have since update firmware 0.87, index and yaml. With compressed air, cleaned connections. Found an old laptop that can be dedicated to Maslow and Recalibrated. It was able to print last night with only one stop in cutting. The dedicated laptop went into windows update and somehow this caused the Maslow to cut a straight line across.

Once the Gcode is loaded and clicked play, does the Maslow require computer feedback?

I think it can just handle one at a time, but it should give you a popup saying that another device is already connected if you try with a second device. That does seem like a good theory.

I should not, once the machine is moving the computer isn’t needed anymore, but it’s nice to have some way to stop it so an EStop button that cuts power is a good idea if the machine is running without a computer controlling it

Steve wrote:

Once the Gcode is loaded and clicked play, does the Maslow require computer feedback?

Currently there is a problem if the connection is lost during the cut. Trying to
fix that is on the todo list.

multiple devices connected is a problem during calibration, but should not be a
problem after that (that we know of)

David Lang

This has been fixed :slight_smile: