Login
Search for:


Previous | Next

From: TLC (all posts)
Date: 1998-09-14
Subject: Re: FF30, timers, FF11
Hi!

On Mon, 14 Sep 1998 15:01:04 +0100 (BST), Richard Atkinson wrote:

>> Please also measure the values in my table (I mean, setting the TED
>> Frq regs to those high values, and then also checking $10...$18,
>> $20...$28 and $30...$38 like $90...$98 and the others.). ...It would
>> give us (...me) an answer for this weird behaviour of the TED.

> This is complicated by the problem of synchronising the two oscillators.
> If they are sync-ed (by setting bit 7 high then low, AFTER setting the
> frequency) then it should be fairly easy to determine the results.
> However, if they are running asynchronously, then cycle-exact (note: maybe
> even 17MHz clock-exact) emulation is necessary.

You say, they can be sync-ed by just setting bit 7 high and then low?
..O.K. then I got another answer. These music players used to sound weird in the first seconds of the replay, anyway I suspected that it should be by bad synchronizing but I haven't exactly known it. However I'm sure you must wait a lot _after setting bit 7 and _before setting it back. I used to remember, this weird sound problem could be solved with just waiting some frames before starting to play. But this delay is needed; else, I guess the mconverter should have make the TED sounds synced right when setting the TED control reg to anything above $80.

Then I guess, it should be enough to measure the synced levels, because after the TED gets stabilized it stays in this state and never gets back unsynced.

> OK, Tibor sent a mail also about this. I'll try to clarify what I'm doing:

> The output of TED itself only has two states: high and low. In every way
> it is a digital (1 bit) output. Therefore, if we know when it is high and
> when it is low (or to be precise, the ratio of high to low) we can
> predict what the analogue voltage will be at the output after filtering,
> without having to measure it (and incur inaccuracies), and thus emulate
> the "final" analogue voltage based on what we know the digital one is
> doing.

O.K. Happy integrating :-)

Anyway I remember you asked what's up with FF. Well, if you set 03ff you'll get the lowest frequency the generator can make. ...The TED's soundgenerator seems to be quite similar to the one found in the VIC-I chip in the VIC-20.

Once I took a closer look to the generator, and got this form:

f=Phi/8/(1024-((R+1)and1023))

where f is the frequency you get (in Hz), Phi is the single clock frq of the computer and R is the value written into the frq registers (10
bits).

Levente

Ps. Whoa! I started working in a secondary school. Guess what waited for me in the deep space of a drawer? ...Well, I found a ZX Spectrum! :-)

Maybe, I should get insatiable and start discovering some new dusty caves in the school ;-).

Copyright © Plus/4 World Team, 2001-2024