[Coco] RGB video on CoCo3

[Steven Hirsch]

On Sat, 14 Feb 2009, Phill Harvey-Smith wrote:

> Steven Hirsch wrote:
>> On Wed, 11 Feb 2009, Phill Harvey-Smith wrote:
>> 
> [snip]
>> 
>> The main problem you're facing is that RGB SCART requires composite sync, 
>> while NTSC CoCo 3 systems (and maybe PAL, but I've never seen one) have 
>> separate H & V sync.  I ran into this trying to interface my CoCo to a scan 
>> converter that uses SCART.  If you're handy, there's a simple circuit to 
>> recombine H & V to composite.  I plan to knock this together Real Soon Now 
>> and play with it further.
>
> Ok some progress, I made up some leads to go from CoCo to veroboard and 
> veroboard to SCART so I can do some patching. So far I have :-
>
> RGB on CoCo direct connected to RGB on SCART (plus grounds connected as 
> needed).
>
> H & V Sync on the CoCo XOR'd together with an LS86, and the the output 
> inverted with an LS04 and passed to CSYNC on the SCART.

That's probabably close enough for a contemporary monitor.  The circuits 
I've seen go a bit further to condition pulse widths.

> This produces a stable picture that is pretty crisp *BUT* in shades of almost 
> white, Ii.e. the colour is all wrong :( I tried putting some resistors in 
> line incase the voltage out of the CoCo was the wrong level, seemed higher 
> than 0.7V when measured with my scope, looked more in the 2V range.

A simple series resistor is not the way to go, since it plays havoc with 
impedance matching.  The source Z looking into the monitor should be as 
close to 75-ohms as practical.  A proper pad preserves both the load and 
source impedance, but a a simple divider will probably give reasonable 
results and tell you if you're on the right track.


--