In article <R7udnS5IW-xBpHrbnZ2dnUVZ_ruqnZ2d@comcast.com>,
Michael J. Mahon <mjmahon@aol.com> wrote:
Scott Alfter wrote:
My 1-Wire routines need to run at 1 MHz, so slowing down to that speed
involves switching off the ZipGS (to go from 8 MHz to 2.8) with its CDA and
then going into the Control Panel CDA to slow down from 2.8 MHz to 1 MHz.
Boy, slowing down for a while is a lot easier on a //e!
That was a brain fart on my part...just selecting "normal" instead of "fast"
is enough.
Just a quick reference to $C0E8 (slot 6 drive off), will slow down
an accelwrator for about 50 milliseconds--usually enough to get done
what needs to be done.
Hmm...hadn't considered that. 50 ms would definitely be long enough (the
longest 1-Wire transaction is a reset, and that's only about 500 �s IIRC).
If I use AN0 and AN1 on a ZipGS-equipped machine to control other hardware,
is that going to have some effect on the ZipGS's operation? Since all of
the ZipGS's configuration is done through software, was the means by which
it's configured ever documented by anybody? If the accelerator kicks in
while software is communicating with 1-Wire devices, it'll produce invalid
results, so I need to make sure that doesn't happen.
The Zip probably requires a secret dance on a softswitch to "unlock",
so you're unlikely to disturb it. On the other hand, software that
talks to the Zip may disturb your hardware.
I know it bangs on AN1 for some reason...haven't tried seeing if it does
anything to AN0. The 1-Wire interface I built pulls the bus high with a
resistor. When AN0 is off, the output of a 74HCT126 buffer is in its
high-impedance state. When AN0 is on, the '126 drives the bus low.
Speaking of 1-Wire interfaces and the Apple II, I found my stash of PC
boards I had made up a couple or three years ago. If anybody wants one to
play with, send me email. They're bare boards; you would need to add a
74HCT126 (SO-14 package), a bypass capacitor, one or two resistors, a 16-pin
wire-wrap DIP socket, and a 3-pin header for a minimal configuration. A
PN2222, a 1N4004, a resistor, and some different headers (5-pin and 1-pin)
would add the relay controller; a DS2417, a bypass capacitor, and a
32.768-kHz crystal would give you a real-time clock on the 1-Wire bus.
(Unfortunately, I forgot to add battery backup to the design, so the clock
resets when you kill the power. I also haven't gotten around to writing a
ProDOS clock driver for it.)
I think $5 would cover what I paid to have the boards made and would cover
mail). IIRC, I have six or seven that I could let go.