USB Resets from touching the case with the charging cable

Framework Laptop 16 Linux
1

So I noticed recently that just touching the charging cable to the outside of my FW16, next to the charging port, causes the external HDMI monitor to blank out for a second sometimes.

This is correlated with messages in the log pretty similar to those that I’ve gotten during other transient problems. One example:

Aug 03 09:50:51 simba kernel: ucsi_acpi USBC000:00: unknown error 256
Aug 03 09:51:03 simba kernel: usb 1-4.1: reset full-speed USB device number 9 using xhci_hcd

That 09:50:51 is from just touching the connector to the laptop case, next to the charging port. 09:51:03 is from starting to plug it in. In this case the monitor blanked for a second at 09:51:03 (the instant the cable touched the machine, before it was fully inserted), but I’ve also seen it blank just from touching the case near the charging port.

I’m not sure what this means, but it seems like a useful data point for tracking down the USB problems.

Which Linux distro are you using? NixOS

Which release version? 26.11.20260702.6517942 (Zokor)

(If rolling release, last date updated?) 2026-07-02 I guess

Which kernel are you using? 7.1.2

Which BIOS version are you using? 04.04

Which Framework Laptop 16 model are you using? AMD Ryzen™ 7040 Series / FRANMZCP07

2

Use a grounded charger

3

I am using the three-prong charger that Framework sent me.

Also, why would the grounded-ness or not of the charger impact what’s going to happen when I touch it to my (ungrounded) laptop? There’s going to be a potential between the two in either case. Maybe I am missing something.

4

The act of them becoming the same potential is likely what causes the issue.

5

I’m aware. My point was (a) I don’t think having a grounded or ungrounded charger would impact that, since the laptop itself is ungrounded at the moment before it’s plugged in (b) isn’t this an issue? I only bring it up because I have been dealing with USB disconnects (currently it manifests as crashes of the whole audio system apparently connected with USB issues on the headphone expansion card for example). If potential in the laptop case has the ability to work its way inside, to the USB subsystem, and then to give voltage spikes to the USB controller when you touch the case with a charger cable or your fingers or anything else that has potential relative to the laptop case, that seems not ideal. Right? I am not 100% sure but it sounds like one possible explanation for what is happening here.

I mean, is this normal in your experience? Do your other laptops cycle their external display when you plug them in, or disconnect their audio or from their network connection?

6

That is exactly the point, it has a floating potential that then on contact with the grounded charger all flows to ground really quickly. The potential itself isn’t doing much but the act of it flowing to ground quickly is what can cause temporary disruptions.

No not particularly ideal, might want to ground your display.

given grounded usb-c psus are actually quite rare it is not that common but it can happen, but if it is that severe you might want to check the psu of your display.

7

I feel like you guys are missing the point lol

Touching the outside of my laptop should not lead to kernel log messages at all, nor to USB failures. I am happy to see if this happens without the monitor connected, just in case it is somehow relevant anything at all about the monitor or the HDMI connection, but my focus is more on “why is the outside of the laptop not fully insulated from the USB controller, and how do we make those two things start to be insulated from one another.” To me, as a non expert, it kind of seems like they should be.

8

Now that is werid.

Fully isolating stuff is hard and has drawbacks.

9

Now that is werid.

I can confirm that yes, just now I touched the cable to the case (next to the USB port, so maybe that part is electrically connected through that USB port specifically? I don’t really know), and got the kernel messages with no monitor or anything else plugged into any port:

Aug 03 23:01:19 kitchen kernel: usb 1-4.1: reset full-speed USB device number 9 using xhci_hcd

… and then the “unknown error 256” one when actually plugging it in, so maybe I don’t really care about that one.

Fully isolating stuff is hard and has drawbacks.

I don’t do EE and honestly I have no idea. That’s why I phrased it as just maybe a useful data point for as long as I am chasing down weird USB headphone-jack problems. To me it seems very surprising that touching the case can cause USB kernel messages though. That’s not how I generally expect laptops to work.

10

Like with nothing else connected at all? Is this just esd? Are there carpets involved/do you frequently get zapped touching door handles and stuff?

Still I’d assume if you hear an audible zap you’d probably have mentioned it.

Yeah at this point I also got used to that, no idea when/if they’ll ever fix that.

If you actually are zapping it with static electricity, it is mandated to survive some pretty unreasonable esd but not necessarily keep running through them without interruption. You are basically setting off a mini emp right next to it’s metaphorical head.

Given the device resetting is a full speed (usb1) it’s probably either (one of) the keyboard(s) or the fingerprint reader which are both coupled to the case (and in case of the fingerprint reader likely needing that very much to function).

1 Like
11

There’s a pin in an HDMI cable that indicates the cable has just beennl plugged or unplugged, to force the screen to rescan(often involves reboot) the signal. However interference can trigger it inadvertently. I think you are mostly right as I can’t find a difference between using grounded or ungrounded adapters. Try using another HDMI cable?

12

What the heck

No, I don’t have carpets. No, I’m not zapping my laptop with static electricity. Yes, this happens when the HDMI cable is not connected at all, like I said above. The only reason the monitor is even relevant is that it’s a signifier of when the USB bus resets. It’s not the HDMI cable.

I just wanted to share one data point that I thought might be related to my headphone troubles; like I say I don’t know that much about the EE side of things so maybe I would be able to learn something. I wasn’t looking for people to hallucinate various things that might be happening, but aren’t, and then want to talk to me about them.

Man, never mind. Like I say in the headphones thread, I’m currently testing a couple of things with an eye towards bringing that up in the GH issue for USB headphone troubles. I think I’ll just include this business in there as a side-note, in case it’s useful for the qualified people when they’re taking a look at it.

13

It is not at all clear what you are doing. Can you post a link to a video showing the whole laptop and what cable you are touching to the laptop and what is connected to the cable.

1 Like
14

I don’t think that would be a good use of my time, no.

One thing that all this stimulating commentary did cause to occur to me, though - I think a lot of my symptoms (including the USB troubles) would fit pretty precisely for what would be happening if there was something inside my laptop that was supposed to be grounding some component (a mainboard screw for example) that wasn’t seated right, or wasn’t forming the connection it was supposed to be forming. So when the ground potential of the laptop as a whole changes – because I’m touching it after going to turn off the light or some other completely normal activity that any functioning laptop should be able to cope with for example – it causes the voltage to equalize instead across some component that’s not designed for it to equalize across.

I might open the thing up and take a look around and make sure I can’t see anything obviously disconnected or loose. Like I say, I have no idea about that side of things generally speaking but it sounds plausible enough to be worth investigating at least.

15

Without perfect information making a hypothesis and then confirming it is a way to get more of it. For just touching the case to do anything electronically there would have to be some flow of current and that would have to come from somewhere, since you say the laptop is not connected to anything that pretty much leaves static electricity. So if the laptop actually is not connected to anything else just touching the case with a grounded or even very not grounded leaky power supply should not produce enough current for anything to happen.

Given pretty much everything is soldered on and the modular components (wifi/ram/ssd/keyboard) would straight up just not work without common ground I doubt that’s it.

[TRACKING] Audio expansion card connection sometimes unreliable
16

I can tell you all from experience that even a shock so hard as to be painful [0] shouldn’t in and of itself cause issues.

EDIT: on top of that charge generally accumulates on the surface of conductive objects as electrons repel each other until they cannot move.

[0] This one time I spent 24h in a building, where the electrical installation was poorly made, so earth potential would vary significantly between areas.

Turns out even though consumer devices use floating chargers which are galvanically separated, some current inevitably leaks.

Every time I went to the bathroom the faucet would zap me and after I returned to my (aluminium) laptop, I would get zapped again.

17

Without perfect information making a hypothesis and then confirming it is a way to get more of it.

That’s actually not true. Or… I mean it is technically true that that is a way. But it’s far from the best way. I’ve observed this a few different times at various software shops; I actually recommended not long ago that someone make it a topic for a work presentation they had to give because to me it is such a key thing for being able to solve problems in engineering.

The main point is this: There are broadly speaking two ways of approaching problems. Method A goes as follows:

  1. Decide, more or less at random, what is “probably” wrong
  2. Act on that assumption
  3. If your action didn’t work, then come up with a new “probably,” rinse and repeat

Method B, to me, is much preferred:

  1. Investigate with the primary aim of figuring out definitively what is wrong
  2. Fix the problem

I’m sure I don’t have to lay out the rest of the presentation. The summary is, I don’t agree with the “make a hypothesis at random” strategy. Now I’m not stupid enough to go out and buy a new HDMI cable or whatever weird thing people are telling me very directly to do here, meaning to act on their “hypotheses” to help them very slowly gather nuggets of information when each one doesn’t work. But yeah it does irritate me a little here. I’m short-tempered about this one specifically because I’m already irritated because it involves my computer being not completely working. And while I do actually have a couple of theories now after looking at it today, I still don’t really understand what’s going on or how to fix my computer, which is its own type of frustrating.

(I should say – this whole thing with touching the case is important to me because my audio system is crashing sometimes when playing over headphones, and I think now that this is related. I’m not bothered about it just because my computer is doing something surprising.)

So if the laptop actually is not connected to anything else just touching the case with a grounded or even very not grounded leaky power supply should not produce enough current for anything to happen.

Completely agree. You need to go up and speak to the people higher in the comments who seemed to think it was not an especially big deal (“not that common but it can happen”). Those guys I definitely didn’t agree with. This behavior, to me, is very weird.

Given pretty much everything is soldered on and the modular components (wifi/ram/ssd/keyboard) would straight up just not work without common ground I doubt that’s it.

That’s also not completely true. So I have been reading up; I am still not really that up to speed on how the EE works, but it honestly sounds to me like some failures I’m experiencing including this particular bizarre behavior would be explained if there was a problem in the “chassis ground” in some components. It’s different from “signal ground” or the main ground plane of the PCB; without those, yes, nothing would work.

Think of it like the third pin of the thing you plug in the wall. If you cut off the neutral pin (the bigger of the two “main” pins), nothing’s going to work. That’s the signal ground. But if you cut off the ground pin (the circular third one), then things will “kind of” work, but not safely and you may have weird problems sometimes. I’m saying that my current working theory is that some component is missing its connection to chassis ground, the “third pin” in plug-socket terms.

That’s a sad theory because I don’t think I am really qualified to diagnose or fix it, if that’s what is happening. And quite possibly I am wrong, I’m just throwing around EE stuff that I don’t understand. But I would be happy to share the details of my totally-unqualified-at-EE thoughts on it and how I am proposing that they could relate to the problems I am having with my headphones, if you want to hear.

18

I got the same-ish thing. With a grounded monitor already connected via HDMI, touching the ungrounded USB-C charger to the case caused my wired keyboard to disconnect momentarily

19

It is good that you can reproduce the problem. We already know what the solution is. Use 3 pin power adapters on all things connecting to the laptop and connected displays etc.

Another possible cause is faulty house wiring

A guess another option is for someone to come up with a design for 2 pin power adapters that does not cause this problem.

And another option, design some feature into the laptop design that prevents this happening on 2 pin power adapters.

If you wish to further gather evidence, take a volt meter and compare the voltage difference between the laptop ground and the usb sheild before connecting it. Some people have seen 50V difference.

20

I got the same-ish thing. With a grounded monitor already connected via HDMI, touching the ungrounded USB-C charger to the case caused my wired keyboard to disconnect momentarily

Fascinating.

Shot in the dark, you don’t happen to use a USB headphone jack expansion card in your machine? Just random curiosity.

I also was pleased to be able to find one other person sometime in history who had an issue like this, and one of the guesses back then was a poorly grounded motherboard. I’m actually a lot less sure about that explanation for my issue, now, but I saw it as confirmation maybe that my EE reading and guesswork led me vaguely in the right direction.