On 3/18/06, James Richard Tyrer <[EMAIL PROTECTED]> wrote:
>
> Timothy Miller wrote:
> > On 3/17/06, Jack Carroll <[EMAIL PROTECTED]> wrote:
> >> On Thu, Mar 16, 2006 at 09:27:40PM -0700, James Richard Tyrer wrote:
> >>> No, if you are trying to display square pixels, the rise time is 
> >>> important.
> >>>
> >>         And displaying square pixels is _really_ important.  We can't get a
> >> fair evaluation of OGA without that.
> >
> > Understood.  But I think we'll have bandwidth that exceeds the 3rd
> > harmonic anyhow.
>
>
> We are talking about the HiRes head here.
>
> I hope that I now have this correct.  The pixel rate is 500 MHz so the
> maximum fundamental video frequency is 250 MHz and the 3rd harmonic is
> then 750 MHz. The frequency response should be basically flat to 750 MHz.

This is something I don't understand all that well.  But I believe
that it is 500 million pixels per second.  In order to have reasonably
good pixels, the bandwidth has to be higher than that.

>
> So, if we intend to have 500 MHz dot clock frequency, the video buffer
> amps do not really have enough bandwidth with a gain of 2.  The ths3201
> spec is 850 MHz -3dB.  This could be improved by increasing the input
> current of the DAC and reducing the gain of the amp.

We're probably going to have to do that tweak.

> With a current feedback amp, you can also do peaking by reducing the
> value of the feedback resistor slightly.
>
> However, it appears that the Burr Brown OPA695 would be a better choice
> for a gain of 2 -- the spec is 1.4 GHz -3 dB.
>
> We could do a better by running with a gain of 1 and differential input.
>   This is what is recommended for the DAC.  But, this is problematic
> with a current feedback amp.

I'll defer to Howard on this.  :)
_______________________________________________
Open-graphics mailing list
[email protected]
http://lists.duskglow.com/mailman/listinfo/open-graphics
List service provided by Duskglow Consulting, LLC (www.duskglow.com)

Reply via email to