New Model F: Intermittent Fault

[Coeus]

I have a new Model F which is demonstrating an intermittent fault in which a whole bunch of keys stop working. Here is the output of the VIal Key Matrix Test:

Vial Matrix Test

Screenshot From 2026-02-26 14-55-29.png (10.4 KiB) Viewed 335 times

This first happened a few days ago. At the time I didn't have time to do anything about it, so I unplugged the keyboard, put it aside, and used another one. Yesterday, I reconnected it with the intention of seeing what diagnosis I could do and, to my surprise, it worked perfectly - all keys working. Later the same day, the fault had reappeared. Unplugging it for a few minutes and plugging it in again solved the issue and, as I left it last night, it was working. This morning, the fault was back.

Are all the keys in that test result on the same row or column of the matrix and, if so, what could be the reason that line stops working after a while, only to work again after the power has been off for a few minutes? I know there is an auto-calibration but why would that drift after the KB has been on for a while?

P.S. I would try to see what the levels are, but I can't find a working version of the the tool. First of all, it isn't clear which controller this keyboard has, though it seems like it is probably xwhatsit. Then, the only PC I can install arbitrary software on is running Linux and the package linked from the manual seems focused on Windows and MacOS. QMK is an open source project but on that website, it does not seem to know about the New Model F.

[Green Maned Lion]

I know absolutely nothing about this subject, but based upon my grasp of computers in general... I'd reflash the firmware.

[Ellipse]

Please take a look at the manual on the project web site to diagnose this issue. You refer to reading the manual but you use the key tester that the manual specifically notes you should not use for key testing, instead of the proper signal level monitor tool.

Intermittent issues can be due to the need to touch up the solder on both ends of the ribbon cable, or a grounding issue. The manual notes which controller each model has.

However, since the keys with issues do not affect every key of the matrix row or column, it could be an installation issue (need to reinstall the keys/springs/etc.).

Yes, please flash the latest firmware as it fixes issues relating to key sensing.

[wobbled]

Recommending to touch up solder is wild on a new product.
Send it back for a refund or replacement mate, yet more QC issues.

[Coeus]

So I finally found a link from the manual to someone else who had helpfully worked out how to compile the pandrew utility*

Running that, even before pressing any keys, it seems to suggest there is a fault with one of the matrix lines, not with individual keys. See:

Signal Monitor, upon connection, layout view

Screenshot From 2026-03-04 20-18-29.png (79.46 KiB) Viewed 220 times

The keys that don't work already have a noticably different signal strength at around 104-106. Switching to matrix view shows a complete line that appears faulty:

Signal Monitor, upon connection, matrix view

Screenshot From 2026-03-04 20-18-46.png (85.92 KiB) Viewed 220 times

Doing as the manual describes and pressing each key gives:

Signal Monitor, after every key pressed, layout view

Screenshot From 2026-03-04 20-22-00.png (99.62 KiB) Viewed 220 times

and then in matrix view:

Signal Monitor, after every key pressed, matrix view

Screenshot From 2026-03-04 20-22-12.png (111.46 KiB) Viewed 220 times

That's as far as I can get at the moment, as I don't have a T8 Torx, either as a standalone screwdriver or a bit.

* Back to the pandrew utility, it would be far preferable that, rather than have a separate distribution of firmware and tools for the New Model F, that where existing open source software has been used, the specifics for the New Model F (for example details of how the layout corresponds to the matrix) were contributed back to the open source projects concerned. That way, the open source projects concerned would do the maintenance of making sure the tools continue to compile and work as compilers and the libraries depended on evolve.

[Ellipse]

So you are saying that the manual provided the correct instructions to compile the utility that you wanted? If anything was no longer accurate or if any details were missing please write them in your reply so I can update the manual.

Regarding keeping all the code in one place, I have just messaged Rico, ploxiln, and NathanA to see if everything can be combined into one repository on the project's github https://github.com/ModelFKeyboard

For now, the latest xwhatsit release and source code is noted in the manual below:

Additionally, as of 10/27/25 there are updated (beta) precompiled Linux and Mac pandrew utilities available here: https://github.com/ploxiln/vial-qmk/rel ... newfxx_r5c

Good detective work with the signal level monitor! Looks like you'll need to pick up a screwdriver kit and touch up the solder for that particular trace on both ends of the ribbon cable. So that we do not further beat a dead horse here once again, it is important to emphasize the manual's reminder that these keyboards do require maintenance and that end users are strongly cautioned against any expectation that the keyboard will run without maintenance; for years we've discussed why I don't take back entire keyboards to solder one wire or fix any other functionality that the user can fix themselves by following the manual.

[wobbled]

Seriously?
A brand new board has issues with solder from factory and you’re suggesting people fix this themselves?
How about you stop shipping out broken junk?
Coeus you should strongly consider doing a chargeback against this scamming asshat.

[Ellipse]

I forgot to mention, please email me instead of trying to send a PM on the forums as sending a PM is currently not working for me.