Keyboard and trackball unresponsive

I recently (last couple of days) experienced several occasions of the keyboard and trackball becoming completely unresponsive, leaving no way to control the device.

I was only able to reboot by using the hardware power switch in the side.

Anyone else seeing this? And is there any way I can debug and maybe address such an issue?

I think I also experienced this, but I don’t have any insight as to why. I think I powered off by attaching a usb keyboard. I’ll try to investigate a little more next time.

I occasionally get the opposite, where the keyboard doesn’t respond to pull up the menu on the system controller to power up. Not sure I’ve had this problem though.

It seems to be happening with increased frequency so that as me worried. I don’t usually carry a spare keyboard with me so the only remedy is power cycling which leads to losing unsaved work.

It might be a partially damaged RP2040 controller or just loose cables. The current keyboard firmware has a watchdog feature to recover from most crashes… which keyboard FW are you on? In any case you can also request a replacement keyboard via warranty.

how do I determine the keyboard firmware version? Is it the same as what the OLED display shows under “system status”?

Yes, there are two fw versions written there normally, you can also show a photo if you don’t wanna type that all in.

Just chiming in (from pocket) but my keyboard does this as well. MNT has been very good about trying to get to the bottom of it, but in my case MNT felt that the RP is bad or going bad. Especially since it seems that the frequency of having the kb freak out and not be controllable seems to increase.

I actually drilled a hole in my kb cover so that I could manually reset the kb. It is getting to the point now where even manually resetting the controller is not resolving the issue.

I am very much looking forward to the replacement kb.

My personal experience with this has me thinking that controlling the RGB kc backlight is one of the possible triggers.

In particular I would allow the psuspend script pulse the kb backlight to indicate when the script was being activated and when it was being resumed from. It was when trying to resume that the kb would just stop resuming. This meant that I needed to manually reset the controller more frequently.

Here is a picture of the hole in the kb top cover that I made:

All in all the kb is the only issue I am having. The Pocket is great!

How do I do that? The problem is persisting and becoming more frequent. It happened two times just today.

You can just write a line to support@mntre.com incl your CS order number and ask for warranty replacement of the keyboard, and link to this post here for reference.

In the meantime, I recommend to upgrade both sysctl and keyboard firmwares to 2025 versions. I’ll make a new release page and instructions later today (longstanding TODO).

Update: tomorrow, because I went to an event for longer than expected and didn’t take a computer.

4 Likes

I have a suspicion: the issue seems to happen mostly when connecting or disconnecting a charger while the device is running.

Not sure if this is helpful though.

Edit: of course right after writing this, the keyboard died on me while running on battery power as I was giving a presentation :anxious_face_with_sweat:

Maybe unrelated but (dis)connecting a charger while running reboots my pocket 1 out of approx. 5 times and that seems to be a new and more frequent thing for my pocket.

I also have keyboard issues where it stops responding and sometimes leaving a T (IIRC) on the menu display but most of the thime I’ll get it to work again by doing a keyboard reset from the menu. Occasionally, when the menu is dead, I’ll need to ssh into it to reboot it and the keyboard works again.

Edit: ugh.. just had the non responsive problem exactly as you described. I had the keyboard displaying the battery status, was away for a couple minutes, came back and all was unresponsive, opened the keyboard menu, it opened, then froze and would not let me reset using the R. Reboot via ssh did not help, power switch cycle did help.

Argh, I’m very sorry that this happened to you! I’ve now written 2 firmware updater scripts that you can execute on device. I’ve tried to cover all eventualities but it’s still possible I didn’t think of something, so if anything fails, please report back and I’ll fix it.

First, the System Controller update (on old FW, this probably requires toggling the standby switch after the update to get everything back to normal, and if the charger doesn’t want to charge, unplugging and replugging the batteries once):

curl -sfL mnt.re/update-pocket-sysctl | sudo bash -

Then, the Keyboard/Input update (probably after a power cycle):

curl -sfL mnt.re/update-pocket-keyboard | sudo bash -

Note that the dash (-) at the end is important. Also, disclaimer: it’s dangerous to pipe stuff from the internet to sudo bash, it’s best to read the script first. These one-liners are merely for convenience.

This will update everything to the latest release 20250429. In the future, from here you’ll be able to update through fwupd / Gnome Firmware as well.

The sysctl updater runs fine but the keyboard updater claims “Error: MNT Pocket Reform Input device not found or more than one found.”

Script started on 2025-04-29 20:44:02+02:00 [TERM="linux" TTY="/dev/tty1" COLUMNS="120" LINES="37"]
[remp 20:44 ~]$ sudo bash update-keyboard.sh 
[sudo] password for me: 
Model: MNT Pocket Reform with i.MX8MP Module

MNT Pocket Reform host detected, assuming you want to flash the system that
is running this tool.

Ensuring dependencies: picotool reform-tools curl...

Downloading firmware to /tmp/pocket-keyboard-tmp.RGQDPTgF/pocket-hid.uf2...

  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
0100  112k  100  112k    0     0   241k      0 --:--:-- --:--:-- --:--:--  240k

Firmware downloaded to /tmp/pocket-keyboard-tmp.RGQDPTgF/pocket-hid.uf2.

Error: MNT Pocket Reform Input device not found or more than one found.
[remp 20:44 ~]$
exit

Script done on 2025-04-29 20:44:27+02:00 [COMMAND_EXIT_CODE="1"]

For reference, this is the output of lsusb on this machine:

[remp 20:54 ~]$ lsusb
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 001 Device 002: ID 0451:8142 Texas Instruments, Inc. TUSB8041 4-Port Hub
Bus 001 Device 003: ID cafe:4004 MNT Pocket Reform Input
Bus 002 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
Bus 002 Device 002: ID 0451:8140 Texas Instruments, Inc. TUSB8041 4-Port Hub
Bus 003 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 003 Device 002: ID 1209:6d07 Generic Pocket Reform System Controller 1.0
Bus 004 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
[remp 20:54 ~]$ 

Ahh, many thanks for the lsusb output. Can you try again? I changed the USB ID that the keyboard update script is looking for. I had forgotten about cafe:4004 :sweat_smile:

Great that the system controller update worked.

I had the same issue but re-running the script worked for some reason

Glad to hear! Maybe the second try was exactly after I pushed the update? ^^ In any case, can you confirm that both date strings in the OLED menu’s System Status page now show 20250429?

Updated! Thanks @minute!

Btw, keyboard status first gave a “syntax error” message but all was fine after a reboot.

1 Like

Since it goes a bit off-topic, I started a new thread about the update issues: Issues with April 2025 firmware

1 Like