Re: TED colors (was: vic-ii pal colors)

Hi!,


On 2020. 06. 08. 19:03, Istvan Hegedus wrote:
> Thank you Levente for the information!
> 
> I have already noticed in plus4emu source code that the NTSC color
> phases are different while the standard says the color phases are
> rotated with 33 degree which is the case in the documentation (just the
> wrong direction). I will check YAPE too and use color angles from the
> emulators. 
> I am investigating the color and luma signals via oscilloscope directly
> on the TED pins and sometimes on the composite output after the RF
> modulator. The luminance levels are also strange and not the same as in
> the documentation...

As to why the actual palette is different from what the specs say, I
know of no evidence. (As a wild guess, the palette hasn't been a
critical factor; that is, whatever colours finally came out of the chip
were "fine" as long as they didn't look particularly horrible... and
IMHO they really didn't). Another guess: the specs were made while the
actual design was still being worked on. Haven't heard of fundamental
differences between HMOS-I and HMOS-II i.e. 7360 and 8360 TED palettes
so far.)

If you make measurements on the signals coming out of the video out
port, be sure to terminate them with 75 ohm resistors to ground. That
way, you're supposed to get value ranges comparable to those defined in
the standards. (I assume, there's no equipment connected to the signals
being measured, other than the oscilloscope probe.)

There's also a small design flaw in all 264 series machines: there is a
ferrite bead in series with the common ground point of the video out
port, which (i.e. the common point that has non-negligible resistance
above the 1-2MHz mark) causes crosstalk between the signals. (This is
the main reason why the separate luma-chroma output image of these
machines looks bad... there is so much crosstalk between luma and chroma
that it introduces visible moire on the image). All in all... if you
want to get rid of a disturbance factor, find and short the FB between
the machine's and the socket's ground.


Best regards,


Levente