Re: TI-H: EEPROMS
[Prev][Next][Index][Thread]
Re: TI-H: EEPROMS
From: "Simon Trimble" <gingerno1@hotmail.com>
> This is a very complicated expander you have got in mind!! Analogue
> sensors!! RTC!! Although ive not come across the 8535, so it could be
> quite easy to do.
It is actually quite simple. The 8535 has 8 10 bit A to D converters on
board that are quite easy to use. A temperature sensor would be as simple
as connecting a linear temp sensor to the input, running an A-D conversion
and then scaling the output. A thermister could be used to, but would
require a bit of correction for its nonlinearites.
> Being able to implement the expander in a BASIC program
> would be pretty good aswell, especially for people like me who dont know
ASM
> but are have good BASIC programming skills.
Using the basic communication scheme I outlined, writing a BASIC program to
act as the driver would not be terribly difficult. The program would just
have to send a request for the index page, then format it for display to let
the user choose what to transfer (possably first validating that space was
available), then request the file from the expander.
Such a scheme would allow simplified operation of the expander on
calculators like the 85 that don't directly support assembly. It would not
be necessary to have a shell on the calculator to have a nice interface to
the expander.
> Someone could probably quite easily tackle the idea of having an accurate
> ref ocillator fed into the clock input of the calc so that you can then
have
> accurate timings within programs.
Yep, thats one of the things I've been meaning to try, should be fairly easy
to set up, maybe I'll check it out.
> Im not so sure about the colour display though. I think major surgery on
> the calc would be in order to get that up and running!
The most difficult thing about a color display is the video timing, a game
would have to switch display planes at just the right time to pick up the
correct color. Synchronizing this over the link port might be possible,
using one of the link port lines as a clock so the controller knows when to
change the lamp color.
Also, setting up the screens in a game so that they looked good might be
difficult. I've been trying to come up with a scheme to do this using three
LED colors, red, green and blue. Since you get 4 display planes you could
get 8 colors, red, blue, green, purple, yellow, cyan, black and white. That
leaves one plane unused though. Possibly it could be used as a secondary
'gamma' channel, to intensify one of the colors.
Thats just off the top of my head though, anyone know of a better scheme
that uses only primary colors (RGB) for lighting?
An advanced scheme could involve half-powered or multiplexed LEDs, to
provide a wider range of colors. The color palette could be sent to the
controller before the game started, to allow different stages of the game to
use different color schemes.
One of the problems here is that when lit from behind the LCD tends only to
be about 50-75% opaque to begin with, so hidden colors wouldn't really be
hidden, just dimmed. This would complicate designing screens.
I keep intending to simulate this on the computer, but have yet to get
around to it.
For more information visit my home page and find the fiberoptic backlight
page.
http://www.genetech.net/~dknaack
(note, the servers internet connection keeps going down, so it might not be
easy to get to.)
DK
Follow-Ups:
References: