[mythtv-users] Hardware for Best TV-Out?

Cory Papenfuss papenfuss at juneau.me.vt.edu
Mon Oct 25 21:36:38 UTC 2004

On Mon, 25 Oct 2004, Dan Lanciani wrote:

> Cory Papenfuss <papenfuss at juneau.me.vt.edu> wrote:
> | 	It cost me about $10 in parts to build one.
> I should have added "off the shelf" to the cost.

 	Fair enough... I don't think that such a thing exists.  It's 
basically a really cheesy NTSC converter.  As you said, most equipment 
that allows for someone bright enough to program the RGB to the right 
frequency probably wants other niceties like TBC or at the very least 
genlock.  I did get ahold of a ancient Macintosh/RasterOps doodad that 
does more or less the same thing.  It's called the "Video Expander" and 
encoded the RGB coming from the mac (that was told to output 480i timings) 
to NTSC.  It also had genlocking capabilities and was apparently 
quite expensive in its day.  Now it's $10 on ebay.  Although I made it 
sorta work, I think it requires clocking the RGB vid card (makes sense 
since it can do external genlock).  It didn't want to free-run so I 
couldn't make it sync.

> |No rate conversion,
> |just VGA->NTSC encoding (color subcarrier with UV modulation, etc).  Aside
> |from having to set the VGA card up for 480i timings, (read: no console,
> |only X), it works great.
> How did you phase lock the color sub-carrier to the existing sync?
 	Easy... I didn't... :)  I *did* adjust the crystal's trimmer cap 
to be as close as I could to the 3.58MHz standard frequency.  I don't know 
how temp stable it is, but it looks excellent on the TV.  I suppose it's 
possible that on a suitably advanced TV (adaptive comb filters, etc) that 
the frequency interleaving would be off a bit with a slightly non-standard 
color subcarrier.  I suspect that the VCO's in most TV's will lock onto it 
within 1000ppm or so anyway... should be well within the tolerance of a 
tuned crystal circuit.

> | 	Check out the AD724 chip.  A few passive components, one chip, and
> |a crystal is all you need.
> I did look at that chip, but it wasn't clear that it solves the hard problem
> of locking a free-running color sub-carrier to the sync in the incoming RGB
> signal.  The documentation seems silent on the issue, but if you look at the
> block diagram there is no obvious path from the sync inputs back into the PLL.
> It will lock to an external carrier clock, of course, but if you have that then
> you have already solved the problem. :)
 	Again... is it strictly necessary to do this?  So long as the comb 
filter gets its own subcarrier lock from the actual signal, the 
frequency interleaving won't cause much trouble.  All I know is that it 
works great for me (not too bad even before I tuned the crystal).


* Cory Papenfuss							*
* Electrical Engineering candidate Ph.D. graduate student               *
* Virginia Polytechnic Institute and State University 			*

More information about the mythtv-users mailing list