View source for Support/Braswell

Intel processors with Braswell SoC's are the replacement for the previous-generation Bay Trail family.

Braswell support incomplete, but being tracked at GitHub: [https://github.com/GalliumOS/galliumos-distro/issues/270 Braswell Platform Validation]. You can subscribe to that ticket for the latest updates.

Braswell models do not include firmware support for booting legacy OSes, but custom firmware is available -- see [[Firmware]].

<div style="margin:3em 0; padding:1em; border:1px solid #ff8888;">
<span style="background-color:#ffcccc;padding:0.2em 0.4em">'''Important&nbsp;Note'''</span> '''(Now RESOLVED -- see below for details)'''

We have received several reports of '''hardware speaker damage''' in Braswell models with the Realtek ALC5650 audio chip. This includes most Braswells, '''except''' the Acer R11 Chromebook (CYAN), which has a Maxim MAX98090 chip instead.

This issue is not limited to GalliumOS or Linux, but it seems to happen more frequently when the machine is ''not'' booted into ChromeOS (i.e. at the Developer Mode screen, in firmware, in GRUB, or running Linux).

When your Chromebook is powered on or rebooted, the ChromeOS bootloader (depthcharge) is loaded first, regardless of what operating system you will be running. Depthcharge performs the first initialization of the audio chip. The chip is re-initialized later by ChromeOS, but if the Chromebook is booted into another operating system instead (or not booted at all), no further initialization is performed. It is ''possible'' that depthcharge is leaving the chip in an inappropriate state, but that should not be an adequate explanation -- unless some percentage of the Realtek chips are faulty.

<div style="border:1px solid #cccccc; padding:0 0 0 0.5em; margin:0 1em;">
'''Update 20170118''': GalliumOS now includes audio drivers for the Realtek ALC5650, and so ''does'' perform additional initialization after depthcharge. This ''might'' reduce the risk of speaker damage, but of course the additional initialization is not performed until GalliumOS is booted -- so there remains a period of time between power-on and fully-booted where the chip is only initialized by depthcharge, and might be at risk. Other Linux distros '''do not''' include these drivers yet, but they should be in mainline kernels around version 4.11 or 4.12.
</div>

Full details and current speculation can be found on the GitHub [https://github.com/GalliumOS/galliumos-distro/issues/270 Braswell Platform Validation] ticket.

The cautious approach is to disconnect your Chromebook speakers until and unless the ultimate cause of the problem can be identified and avoided.

A more casual approach is to ''carefully'' observe your Chromebook for unusual noises/heating around the speaker area while in use (or powered on), and be prepared to return any defective hardware within ordinary return windows if problems occur. We estimate the failure rate at about 2-3% of all devices, and the failure is expected to happen within a short period of use (minutes to single-digit hours).

At this time, and for reasons detailed at the aforementioned link, we suspect a hardware problem with the Realtek ALC5650 chips. We are still investigating. If you have a failed device that you have decided not to return (or that has adequate time remaining on the return window), and which you are willing to let us borrow for a few days, please get in touch (contact. at. galliumos.org or Libera.chat [https://galliumos.org/irc #galliumos]). Thank you!

<div style="border:1px solid #cccccc; padding:0 0 0 0.5em; margin:0 1em;">
<p>'''Update 20180723''': We've seen no new reports of this issue since before January 2017, so it's safe to consider this '''fully resolved'''.</p> <p>Additionally, Braswell models now have the option of full ROM firmware from MrChromebox (see [[Firmware]]), which replaces the ChromeOS firmware entirely, removing the last small gap of exposure.</p>
</div>

</div>