On Mon, 24 Jul 2000 13:27:50 +0200, Laurens Holst wrote:
>> This is just a speed up. (-: It doesn't generate garbage, and once you
>>switch back to 3.57, everythig is ok again.
>Well I certainly think it sounds like garbage. Especially when it switches
>back to 3.5MHz all the time. The thing I was talking about by the way is why
>they didn't output a 3.5MHz clock to the devices, it wouldn't be hard and
>cause insynchronization (???) since the clock speed is exactly the double.
To sync you'll have to slow Z80 down, because PSG, pe, doesn't control the
speed the data arrives. Even making PSG work at 3.57Mhz, it will sound faster
if the sound routine is not interrupt-attached.
So, the only solution is: if data is sent to PSG and another that is sent
before a resonable time, a circuit should "halt" Z80 until PSG is ready to
play the new data.
BTW, this is what TR does with V9958.
>The real solution is letting the engine control the timing, just like the
>turboR does. If you output a byte to a port, and then output another byte
>too fast after the first, the engine holds that byte and sends a wait signal
>to the processor until the preset wait-time expired.
Yes, this is the trick. I think this is the better
type of control, but I also think it would be nice to
be turned on and off... Something we would like to use
"different speeds" to play data. (^=
The bad thing on TR is there is no way to turn this
control off.
----- Abra�OS/2, Daniel Caetano ([EMAIL PROTECTED])
/| | | |\
\| ___ |/ OS/2 Sites: http://www.quasarbbs.com/daniel/
\/ ----- \/
http://www.geocities.com/SiliconValley/8752/os2hp/os2index.html
| | MSX Sites: http://www.fudeba.cjb.net/
-- -- Drawings: http://www.djgallery.tsx.org/
...Programar e' a arte de organizar zeros e uns de forma que eles produzam trabalho
util!
****
Problems? contact [EMAIL PROTECTED] See also http://www.faq.msxnet.org/
****
