Question regarding HDMI Audio support on i.MX6 with vanilla kernel

[Russell King - ARM Linux]

On Tue, Nov 07, 2017 at 10:24:37PM +0000, Luís Mendes wrote:
> Hi,
> 
> On Tue, Nov 7, 2017 at 8:07 PM, Russell King - ARM Linux <
> linux at armlinux.org.uk> wrote:
> > 2. there is an (optional?) TXS0102 level converter on the DDC signals
> >    which, although marketed for I2C buses, doesn't actually conform to
> >    the I2C bus spec.  The I2C bus spec calls for 30% and 70% of the
> >    bus voltage for input thresholds, which would be 1.5V and 3.5V.
> >    The TXS0102 datasheet gives more stringent levels, which are 0.15V
> >    and 4.6V assuming a 5V "B" supply.
> >
> >    (Note: HDMI calls for the DDC bus to conform to the I2C bus spec.)
> >
> > So, I think this Wandboard design has some question marks over the
> > hardware design, and so I wouldn't expect HDMI to be reliable.
> >
> 
> This chip must be the culprit, it really does a bad job at translating I2C
> signals. The logic levels are completing wrong then.
> If the chip is not populated because iMX6 may have some 5V tolerant pins,
> then it is barely able to drive the logic 1 level for the monitor, although
> it can receive properly

I just looked at what was done on the Hummingboard, and it uses a
N-channel mosfet with the gate connected to the 3.3V supply, source to
the imx6 and drain to the HDMI connector, with 4.7K pull-ups on both
sides - obviously to 3.3V on the imx6 side and 5V on the HDMI side.

IOW, a standard single-mosfet level shifter for open-drain signalling.

I haven't heard of that causing any problems, and as it's on the uSOM,
and remains the same on the more recent rev1.5 uSOM, I suspect that
no one has had a problem with it.

It's probably cheaper than a TXS0102!

-- 
RMK's Patch system: http://www.armlinux.org.uk/developer/patches/
FTTC broadband for 0.8mile line in suburbia: sync at 8.8Mbps down 630kbps up
According to speedtest.net: 8.21Mbps down 510kbps up