[RESPONDED] HDMI expansion not working (at least on Linux)

Hi community,

before starting a second support request, I thought I might try if someone around here can help me. The scenario is the following:
I have a Framework Laptop 13th Generation (Intel) and the hdmi expansion card does not work on most external monitors/projectors I connect to. I had contact with support a few weeks ago and they replaced my expansion card, but the error is still the same. So I’m tending to say it is not a HW issue, but I’m not sure, with some bad luck, it would still be possible.

I did the following tests:

  • Various monitors/projectors all with different cabling (even hdmi to vga)
  • Tested the expansion board on all slots
  • Tested with Ubuntu, Fedora, Suse (tried with WinPE but I realized that there a close to no drivers in it, so I don’t know if it works on Windows or not)
  • Several external USB-to-HDMI adapters, which are all working

The dmesg repeatedly shows disconnects:

[  +0,012945] hid-generic 0003:32AC:0002.0092: hiddev96,hidraw0: USB HID v1.11 Device [Framework HDMI Expansion Card] on usb-0000:00:14.0-2/input1
[  +0,037929] usb 3-2: USB disconnect, device number 46
[  +2,504114] usb 3-2: new full-speed USB device number 47 using xhci_hcd
[  +0,149043] usb 3-2: New USB device found, idVendor=32ac, idProduct=0002, bcdDevice= 0.00
[  +0,000008] usb 3-2: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[  +0,000002] usb 3-2: Product: HDMI Expansion Card
[  +0,000002] usb 3-2: Manufacturer: Framework
[  +0,000001] usb 3-2: SerialNumber: 11AD1D00B21A3F1133350B00
[  +0,010402] hid-generic 0003:32AC:0002.0093: hiddev96,hidraw0: USB HID v1.11 Device [Framework HDMI Expansion Card] on usb-0000:00:14.0-2/input1
[  +0,174162] usb 3-2: USB disconnect, device number 47
[  +2,509739] usb 3-2: new full-speed USB device number 48 using xhci_hcd
[  +0,154367] usb 3-2: New USB device found, idVendor=32ac, idProduct=0002, bcdDevice= 0.00
[  +0,000009] usb 3-2: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[  +0,000002] usb 3-2: Product: HDMI Expansion Card
[  +0,000002] usb 3-2: Manufacturer: Framework
[  +0,000001] usb 3-2: SerialNumber: 11AD1D00B21A3F1133350B00
[  +0,011132] hid-generic 0003:32AC:0002.0094: hiddev96,hidraw0: USB HID v1.11 Device [Framework HDMI Expansion Card] on usb-0000:00:14.0-2/input1
[  +0,168992] usb 3-2: USB disconnect, device number 48
[  +2,572139] usb 3-2: new full-speed USB device number 49 using xhci_hcd
[  +0,154387] usb 3-2: New USB device found, idVendor=32ac, idProduct=0002, bcdDevice= 0.00
[  +0,000014] usb 3-2: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[  +0,000004] usb 3-2: Product: HDMI Expansion Card
[  +0,000003] usb 3-2: Manufacturer: Framework
[  +0,000002] usb 3-2: SerialNumber: 11AD1D00B21A3F1133350B00
[  +0,011853] hid-generic 0003:32AC:0002.0095: hiddev96,hidraw0: USB HID v1.11 Device [Framework HDMI Expansion Card] on usb-0000:00:14.0-2/input1

I even logged the USB port with usbmon, but deciphering it is out my scope:

I hope someone can provide clues?

1 Like

Hi @BBreinbauer, welcome to the community.

You’ve definitely done everything correctly. If this behavior is happening with the new HDMI card, and for the sake of triple testing the card, try all the other expansion slots, I’d say this is a board issue.

We know you have a replacement card, please test the card on another computer if you have one available. I’d like to make absolutely sure we’re seeing good behavior from the card itself.

Then from your previous ticket email, reply to it and reopen the case and link to this thread. I’d like to take a closer look at what is going on with that board.

Thanks

2 Likes

Hi Mark,
thanks for your kind reply.
I will be traveling next week, but I will try the expansion board on another Laptop when I’m back. Mind, that I tried both HDMI cards on all expansion slots (even twice), so this check was done.
Thanks

1 Like

I can also reproduce this problem on my batch #7 AMD Ryzen 7040 Series Framework Laptop 13(yes, I’ve avoided the port that doesn’t support display output as noted in the guide), @BBreinbauer have you solved the issue now?

An interesting thing to note is that the HDMI expansion card is detected as a full-speed USB device(12Mbps), which shouldn’t have enough bandwidth to transfer the video stream (unless Framework has done some firmware hack to switch to a different USB mode when the connection is available).

Pretty sure that’s how the power save mode works. It’s supposed to only actually show up as a display when a cable is plugged into it and as a suspendable usb device the rest of the time.

The display signal isn’t going over the usb connection either way, this isn’t a displaylink adabter.

1 Like

Yes, that is what I suspected in the first place, I’ll try changing my configuration to see whether it’s actually a hardware issue.

Hi,
glad that the topic is life again. I had contact with support and they provided two (!!) spare adapters, but all show the same behavior. I was already wondering, why I can’t find similar issues in the forum.

What I confirmed in the meantime: I can plug the adapter into other Notebooks with USB-C and it will show the same symptoms. On a windows device the device manager would refresh itself every ~2secs. So I guess it’s also the same there (Don’t know how to get a decent log on Windows.)

Add: Interestingly, my monitor at home is working without problems with the HDMI card. At work, I’m using several different projectors. None of them is working with the expansion card. (I have an external USB-HDMI adapter as workaround for now, which is working flawlessly)

1 Like

Hello, I appreciate your update on this issue!

I suspect the HDMI expansion cards have a compatibility issue with certain HDMI displays. I’m reproducing this issue on a DELL P2314H monitor (SKU: P2314Ht, via an HDMI-to-DVI-D plug as the monitor doesn’t have HDMI sockets), as I’m residing in Taiwan maybe on-site debugging may be possible if Framework can do so?

I can also reproduce this similar behavior on Linux, the system can detect the monitor(with the model name, supported resolution, and other details), but whenever I try to enable it the the external monitor never appears to receive any signal while the system kinda enters an unresponsive state until I disconnect the HDMI cable.

I’m currently not on the supported OS configuration, though(currently using an Ubuntu 23.10 + 22.04 OEM-C kernel combination), I’ll try to install a supported configuration somewhere else and boot from it to see whether it relates to the problem(highly doubted so, but one can try).

Does your dmesg output show similar messages as the one I posted above? My system never gets to a state, where I can switch to the external display, as the expansion card is disconnected/connected from USB in an endless loop.

Also, I tried with Fedora and Ubuntu Live Systems, they had the same issue.

@BBreinbauer I checked to see if you had any active tickets, looks like you did get the RMA sorted.

At this stage, the next step is to get the logs what you’re experiencing with the projectors to AMD in a bug report.

I’ve seen issues with this in the past. Ideally, we want HDMI to HDMI as third party adapters are prone to issues we cannot reproduce effectively. Are you able to replicate the issue with a straight HDMI connection without the third party adapter?

1 Like

@Matt_Hartley Please note, that I’m on an Intel platform. So I guess the bug report should be directed there. Do you need any infos from my side for a bug report?

With the third party stuff, as we don’t use or support them, best course is to provide logs with the distro itself. But as you indicated, this has been happening across multiple distros.

What we can do is take a look at your logs in a ticket. If we can spot where this disconnect is happening, replicate on say, a Live USB of Fedora, then we may have a path forward for more in-depth support.

1 Like

Apologies for the ignorance, I was unable to access the display for some time and (along with other task scheduling problems) could not address this issue until now.

Unfortunately, this isn’t possible as the DELL P2314Ht display doesn’t have any HDMI sockets. However, I did have another HDMI to DVI-D adapter to test and the issue is no longer reproducible with the other adapter!

I have recorded a video of the process of reproducing this issue:

Note that the original HDMI to DVI-D adapter does work for most (if not all) other HDMI appliances so I don’t think the adapter is faulty in any fundamental way, some searches on the web also reveal some reports of similar issues:

Though this seems to be AMD graphics-specific territory, so apologies to @BBreinbauer for polluting the thread.

After analyzing my kernel log dumps I can confirm this is not the symptom I am experiencing, the following is the kernel log generated while I am doing the following tasks:

  1. Connect the HDMI expansion card to the Framework Laptop(23:18:56)
  2. Connect the HDMI cable to the HDMI expansion card(23:19:08)
  3. Wait a few seconds while noticing the internal display stops displaying and the external display is stuck in “NO SIGNAL” state(23:19:33)
  4. Disconnect the HDMI cable to the HDMI expansion card(23:19:33)
12月 30 23:18:56 brlin-fw13 kernel: usb 7-1: new full-speed USB device number 3 using xhci_hcd
12月 30 23:18:56 brlin-fw13 kernel: usb 7-1: New USB device found, idVendor=32ac, idProduct=0002, bcdDevice= 0.00
12月 30 23:18:56 brlin-fw13 kernel: usb 7-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
12月 30 23:18:56 brlin-fw13 kernel: usb 7-1: Product: HDMI Expansion Card
12月 30 23:18:56 brlin-fw13 kernel: usb 7-1: Manufacturer: Framework
12月 30 23:18:56 brlin-fw13 kernel: usb 7-1: SerialNumber: 11AD1D002C9740171C2F0B00
12月 30 23:18:56 brlin-fw13 kernel: hid-generic 0003:32AC:0002.000F: hiddev0,hidraw2: USB HID v1.11 Device [Framework HDMI Expansion Card] on usb-0000:c3:00.4-1/input1
12月 30 23:19:08 brlin-fw13 kernel: usb 7-1: USB disconnect, device number 3
12月 30 23:19:11 brlin-fw13 kernel: [drm] DP Alt mode state on HPD: 1
12月 30 23:19:11 brlin-fw13 kernel: usb 7-1: new full-speed USB device number 4 using xhci_hcd
12月 30 23:19:11 brlin-fw13 kernel: usb 7-1: New USB device found, idVendor=32ac, idProduct=0002, bcdDevice= 0.00
12月 30 23:19:11 brlin-fw13 kernel: usb 7-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
12月 30 23:19:11 brlin-fw13 kernel: usb 7-1: Product: HDMI Expansion Card
12月 30 23:19:11 brlin-fw13 kernel: usb 7-1: Manufacturer: Framework
12月 30 23:19:11 brlin-fw13 kernel: usb 7-1: SerialNumber: 11AD1D002C9740171C2F0B00
12月 30 23:19:11 brlin-fw13 kernel: hid-generic 0003:32AC:0002.0010: hiddev0,hidraw2: USB HID v1.11 Device [Framework HDMI Expansion Card] on usb-0000:c3:00.4-1/input1
12月 30 23:19:11 brlin-fw13 kernel: Registered IR keymap rc-cec
12月 30 23:19:11 brlin-fw13 kernel: rc rc0: DP-3 as /devices/pci0000:00/0000:00:08.1/0000:c1:00.0/rc/rc0
12月 30 23:19:11 brlin-fw13 kernel: input: DP-3 as /devices/pci0000:00/0000:00:08.1/0000:c1:00.0/rc/rc0/input26
12月 30 23:19:33 brlin-fw13 kernel: usb 7-1: USB disconnect, device number 4
12月 30 23:19:36 brlin-fw13 kernel: usb 7-1: new full-speed USB device number 5 using xhci_hcd
12月 30 23:19:36 brlin-fw13 kernel: usb 7-1: New USB device found, idVendor=32ac, idProduct=0002, bcdDevice= 0.00
12月 30 23:19:36 brlin-fw13 kernel: usb 7-1: New USB device strings: Mfr=1, Product=2, SerialNumber=3
12月 30 23:19:36 brlin-fw13 kernel: usb 7-1: Product: HDMI Expansion Card
12月 30 23:19:36 brlin-fw13 kernel: usb 7-1: Manufacturer: Framework
12月 30 23:19:36 brlin-fw13 kernel: usb 7-1: SerialNumber: 11AD1D002C9740171C2F0B00
12月 30 23:19:36 brlin-fw13 kernel: hid-generic 0003:32AC:0002.0011: hiddev0,hidraw2: USB HID v1.11 Device [Framework HDMI Expansion Card] on usb-0000:c3:00.4-1/input1

Although the USB device of the HDMI extension card appeared to be unexpectedly reconnected at the moment the HDMI cable was inserted(23:19:11), it didn’t end in a connect-disconnect-connect endless loop like you have been experiencing.

Thanks for the infos. I’m currently in discussion with support again. I’m going to report results here, as soon as I have some.

1 Like

Good luck! As your problem doesn’t seem to be widely acknowledged I suspect it might be rather in your mainboard instead of the HDMI expansion cards.

That said it’s a wild guess so I leave the Framework support to figure out themselves.