Re: [gentoo-user] U.V.C. Webcam Disconnects shortly after Registration

[Michael <confabulate@kintzios.com>]

[-- Attachment #1: Type: text/plain, Size: 3367 bytes --]

On Friday, 25 December 2020 03:56:58 GMT Ashley Dixon wrote:
> How's it going dudes?
> 
> I bought a Logitech C270 webcam from Argos about a year ago, which,
> according to the LKDDb [1] and U.V.C. driver website [2], should work well
> on Linux.  Despite working flawlessly until recently, some strange
> behaviour has appeared.

Have you changed anything recently on this PC, coincidental with this problem 
arriving?  Anything related to kernel/drivers, webcam related apps and audio?

Besides dmesg, do you get any other errors in syslog?

Have you disconnected all other USB devices on other ports hanging off the 
same MoBo USB hub controller?


> When I connect the webcam, `dmesg` registers it as would be expected
> (timestamps omitted):
> 
>         usb 3-7: new high-speed USB device number 14 using xhci_hcd
>         usb 3-7: New USB device found, idVendor=046d, idProduct=0825,
> bcdDevice= 0.12 usb 3-7: New USB device strings: Mfr=0, Product=0,
> SerialNumber=2 usb 3-7: SerialNumber: 9EEFE4E0
>         uvcvideo: Found UVC 1.00 device <unnamed> (046d:0825)
>         input: UVC Camera (046d:0825) as
> /devices/pci0000:00/0000:00:14.0/usb3/3-7/3-7:1.0/input/input27 usb 3-7:
> set resolution quirk: cval->res = 384
> 
> However, about two seconds later, me having done nothing, the kernel ring
> buffer suggests it has been physically unplugs, and thus deregisters the
> device...
> 
>         USB disconnect, device number 14

I don't have the same camera, but it could be something related to power.  
When the device is detected and initialised by udev + kernel driver, it starts 
drawing more current to deal with features like autofocus/exposure.  Once it 
settles the current consumption reduces.


> I have tested this on another very similar machine, and this problem is  not
>  to be seen, and everything works perfectly; thus ruling out a hardware 
> issue  with the camera.  Furthermore, no other devices---even other U.V.C. 
> webcams---do not present with the same symptoms on my machine, in the  same
>  U.S.B.   port;  thus ruling out any issue with the physical port or
> motherboard controller.

Not necessarily.

Stating the obvious, other machines are not the *same* machine.  Differences 
in mV/μA between MoBos could cause a marginal case to manifest itself.

Other UVC webcams, are not the same webcam in terms of sensitivity and 
electrical consumption.


> I have no idea how to triage this further.  I feel like my head has been 
> nailed to a coffee table by Dinsdale Piranha.  Any tips?   My  kernel  is 
> 5.9.13-zen1, although this trouble presents identically on a vanilla
> kernel.

Do you have an external USB hub, with its own separate power supply, to see if 
connecting the camera to it overcomes with this problem?

Check if the driver has any power saving features with 'modinfo' and if 
switchable, disable them and try again.

Camera sensors and chips can degrade over time and change their performance 
with differences in temperature.  I appreciate a year is not long enough, but 
I had camera sensors fail in a matter of months - thankfully still under 
warranty.


> Cheers.
> 
> Hope everyone is remaining well.
> 
> [1] https://linux-hardware.org/index.php?id=usb:046d-0825
> [2] https://www.ideasonboard.org/uvc/#devices


[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 833 bytes --]