Just had a thought - one which I don't know if it is realistic, since I have little experience with hardware development and my software developing mind is half asleep right now. But here it is.

Would it be possible to create an electronic stylus that plugs into the calculator link port and can be used to click on the screen? My idea for how it would work is that it would have two components, a touch button and some sort of light or color sensor, both located in the tip. When the touch button is depressed, the calculator would very quickly wipe the screen and begin to fill it with black, line by line, pixel by pixel. As soon as the light sensor registers black, the calculator would know it had found the coordinates of the stylus click. The screen would be restored to normal. This is a similar technique to a game called Duck Hunt which, although I've never played, sounds like a very clever technique.

Does anyone think this might be feasible? Would someone like to try it out? I lack the electronics and hardware skill to build something like this, but I'd be very happy if it could work for someone else.

Huh, that actually sounds pretty cool. A much faster method of searching would be to sweep two vertical lines from the edges inward to get the X coordinate (once triggered, flash the two possibilities in sequence, which saves only slightly less than half the time), then sweep two horizontal lines edges inward, again flash the two possibilities in sequence. That would work nicely, I might actually have the inclination to give this a try.

that's hot

Erm, just thought of a problem: LCD lag is disgustingly poor.

I suppose that this won't work on an emulator...

oddly enough I can't think of any reason you couldn't do it with a mouse if the capability was programmed in to the emulator.

this sounds like a sweet idea, but like Kerm said, wouldn't there be a lot of lag?

it wouldn't be real time touch screening, but you should be able to scan lines across the LCD pretty quickly in ASM.

I figure the resolution wouldn't be that great, but you might be able to discern the general area of the screen it's in. Maybe instead of doing pixel by pixel, the calculator could draw larger blocks, like 8x8.

or you could do what Kerm suggested, that would be pretty quick and pretty accurate

But you have to take into account LCD lag to correctly identify the row/pixel. I am going to draw up some plans on building this today and perhaps give it a try

Yay! If you do build one, let me know what parts and circuits you used. I mightbe able to mash one together, and then start trying out software methods.

The reason that I'm not sure how well it will work is that I'm not sure how well a passive light sensor can detect the screen changes. But I trust Kerm to know what he's doing.

Kerm will be building an active model, of course. Lightguide, 3mm LED (visual spectrum or IR? Doesn't really matter), pressure-sensitive tip to light the LED, power mooching from the serial I/O port.

Edit: on second thought, how much current can the serial port safely source? Might not be enough...

I think its 3v or 5v....you should test it though.

Current != Voltage. You disappoint me, elfprince13. The voltage is indeed 0/5v, though.

bahhh, I'm leading a deprived life. I haven't played with electricity since June.

My CS 210 class should be giving us microchips to play with any week now. I'm excited!

Circuit tested and verified for red LED and IR detector. Now I need to build it and set the threshold properly. The hard part will be finding light guides...

Hurry up and create the embedded forum so I can knock some hardware sense into these software junkies



Cool; do you know which model?

what's 210 at your school?