[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: RT.SYNTH -- anybody tried it?
Eric Smith wrote:
"Michael J. Mahon" <mjmahon@aol.com> writes:
No, each reference complements the state of the speaker toggle, with
the initial state undefined. So each reference causes the level to
switch from 0 to 255 or from 255 to 0, depending on its current level.
(The initial state may, in fact, be defined. But after loading ProDOS
<click> and getting a few "beeps", it's anyone's guess what the state
is--unless all speaker sounds use an *even* number of transitions!)
In fact, the speaker can't even tell what the state of the flip-flop is,
so there's no functional difference whether the state is high or low.
Only the transistions matter. This is because the amplifier circuit is
capacitively coupled, and thus does not pass a DC level. So leaving the
flip-flop in the "on" state results in exactly the same behavior at the
speaker as leaving it in the "off" state.
The problem is not only that the toggle is capacitively coupled
to the driver, but that the drive circuit for the speaker is not
equally competent to drive it "off" or "on". The drivers differ
between models, but in the //e and earlier, when the toggle is
switched "on", the 0.1 uF coupling capacitor causes the Darlington
pair to pull the speaker down hard through a 27 ohm resistor. As
the coupling capacitor charges, the drive current decreases with
about a 5 millisecond time constant.
If more than a few milliseconds have passed, when the toggle is then
switched "off", the capacitor is discharged very quickly through a
1N914 diode, but, since the Darlington is already off, no further
decrease in speaker current results.
If less than a few milliseconds have elapsed since the "on" state,
the Darlington is switched "off", and the speaker current is
diverted to charge a 0.1 uF capacitor, resulting in a decay to 0
in about a microsecond. (This decay time is comparable to the
current turn-on ramp caused by the inductance of the speaker and
the 27 ohm series resistor, and so is not a major source of
assymetry.)
One effect is that on these Apples, two successive PEEKs to the
speaker separated by more than a few milliseconds will be heard
as a click followed by almost no sound at all. Only when the
time between successive toggles becomes comparable to or smaller
than 5 milliseconds does the speaker waveform begin to look more
like a squarewave.
Any assymetry in the driving waveform will result in different
current waveforms (besides inversion) depending on the initial
state of the speaker toggle. This effect will become more
pronounced for low toggle frequencies.
The cassette output is not capacitively coupled, so the polarity may
make a difference, depending on exactly what you're trying to do with it.
Of course, it is capacitively coupled by a cassette recorder. ;-)
But if you planned to use it for, say, a level-sensitive network,
you would have to sense its level and flip it if the initial state
were wrong.
And, as you note, there is no internal feedback to the Apple II about
the state of either the speaker or the cassette toggle.
-michael
Music synthesis for 8-bit Apple II's!
Home page: http://members.aol.com/MJMahon/
"The wastebasket is our most important design
tool--and it is seriously underused."