Date   : Thu, 19 Jan 2006 11:17:52 +0100
From   : "W.Scholten" <whs@...>
Subject: Re: Master to SCART

Sprow wrote:

> I thought they took sync off the composite input. That would suggest a 0.3V
> sync if terminated into 75R wouldn't it?
> Sprow.

Might not matter if there's just a sync signal?

Regarding
   http://www.mdfs.net/Info/Comp/BBC/Monitor/BBCtoTV.htm

I emailed the author of that document in 2001 about a few things which I 
didn't agree with. My webpage at that time was a little incomplete as 
well (not mentioning the BBC's output impedance).

I have the following problems with it and the original doc file:
  - TTL is not 5V, but approx >=3.5V counts as a '1'. Outputs are thus 
 >=3.5V, but usually not much higher.
  - Pin 17 is ground for composite video in/out. Composite is not used 
so that is wrong. Even though internally all ground pins are usually 
connected, this is not always the case and for example a SCART plug my 
dad made long ago wouldn't work with a new TV. That's why I connect all 
relevant grounds to 0V.
  - Resistor values are too low (but see below for more thoughts on this).
  - Resistor values don't correspond to the image shown (5V & 1V <-> 220 
Ohm & 75 Ohm)
- "the SCART inputs are required to be 1v peak to peak" I presume is 
meant as "1V peak to peak or __less__" (nitpicking). Also, see below 
about eff. values being the standard (for a long time).

Sync output is higher than RGB output, about 4V. SCART specs say sync 
input should be below 1V, so a resistor of at least 225 Ohm (I use 330 
Ohm and have used much larger ones as well, all working). 270 Ohm looks 
  safe from the calculations, but I've not tried it.

See my webpage on monitors which has moved to:

   http://www.xs4all.nl/~swhs/whs/8bit/varia/monitors/monitor.html

Note: The SCART specifications seem to have changed at some point, 
changing from peak values for say RGB inputs, to effective values. 
Someone told me this was done as this was easier to measure.

So old SCART docs say RGB is 1V max, whereas newer ones say 0.7V maximum 
(where they implicitly mean the effective value).
Where does that leave square inputs? Can one use the peak value 
associated with a sine wave? If so, this would give an eff. value of 1V 
which is too high. Etc. A TV engineer might be useful at this point :)

I measured the BBC RGB outputs long ago and got to about 3.6V eff. (Wave 
shape? Not sure, my dad used to have a scope long ago but unfortunately 
I don't at the moment, so can't check now).

In any event, using 3.6V eff vs. 0.7V eff. maximum I get Rrgb >=243 Ohm
and I recommend using about 330 Ohm. It depends on your TV settings, I 
use 330 because it gives me nice bright picture similar to my TV 
settings so I won't have to mess about with brightness when I switch to 
the SCART input and back...

-- 
Wouter
---
BBC microcomputer information: http://www.xs4all.nl/~swhs/whs/