New camera sensor SIV121C

[James Cameron]

On Sat, Jun 07, 2014 at 08:22:23AM +1000, James Cameron wrote:
> On Fri, Jun 06, 2014 at 09:26:31AM -0400, Paul Fox wrote:
> >  > On Wed, Jun 4, 2014 at 9:38 PM, James Cameron <quozl at laptop.org> wrote:
> >  > > We have begun qualification of a new camera sensor SIV121C for
> >  > > production of the XO-1.75 and XO-4 laptops.
> >  > ...
> >  > 
> >  > > The new sensor does a little bit better at low light levels, as a
> >  > > result of a different analog to digital conversion method.
> > 
> > i noticed on trac that you're supporting the new sensor on 1.5.  for
> > replacement purposes?
> 
> yes, that's my hope.  i'm assessing that now.
> 
> i've an XO-1.5 motherboard, display, and sensor stacked, and have been
> fiddling with new bits in the sensor registers; output drive
> capability select [1].
> 
> the XO-1.75 and XO-4 work best with 3mA for the pixel clock and "other
> outputs", which are the data lines.  at 8.2mA there are occasional
> corruptions from bright parts of a scan line to the rest of the line,
> with the condition clearing for the next scan line.
> 
> the XO-1.5 works best with 8.2mA for both.  at 3mA the data is certainly
> corrupt, at 5.6mA it is occasionally corrupt, and at 8.2mA it is
> rarely corrupt.

turns out the XO-1.5 shares the USB clock with the sensor, at 48 MHz,
and this is way above the sensor maximum clock of 27 MHz.

so this siv121c camera sensor won't work on XO-1.5.

the siv120d on the other hand has a maximum clock of 48 MHz, so it
works.

-- 
James Cameron
http://quozl.linux.org.au/