[Intel-gfx] TV detect

[ykzhao]

On Tue, 2010-01-05 at 10:17 +0800, Peter Clifton wrote:
> Does anyone know what the expected (s-video output) electrical waveforms
> for a "correct" load-detect look like (with / without output connected).

We don't care the electrical waveform for the "correct" load-detection.
In fact this is handled by hardware. And we only use the TV status
register to decide which TV type is connected.

> I can probably hook up a scope in the lab to verify whether they are
> getting out of the chip OK.
> 
> Do the DAC outputs sit at 0.7V DC? What are the timing requirements for
> the sensing?
Using 0.7V DC in DAC output is defined in spec.

The doc can be found in
http://intellinuxgraphics.org/documentation.html
> 
> 
> I see a few possibilities for the miss-detection..
> 
> 1. Not waiting long enough for the load detect pipe's PLLs to come up
> before trying to use it to detect the TV-connection.

Have you tried to add more waiting time and see whether the TV is
detected correctly on your box?
> 
> 2. Not waiting long enough for the DACs to switch on / stabilise ->
> under voltage level when we detect -> misdetect.
> 
> 3. Laptop specific hardware issues (presenting some phantom load or
> "noise" which is near the detection borderline)
> 
> 4. Correct detection depends on a more comprehensive register set-up, or
> other bits of state within the TV-out circuitry. I guess the hardware
> guys would know more about how the load detection works inside the chip,
> hence what prerequisites there are.
> 
> (Although, as far as I can see / infer, there is a full mode-set to NTSC
> 480i before detection commences).
> 
> 5. Evil penguins flipping bits inside the chip.
> 
> 
> 
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/intel-gfx