Ok im reading up on digital logic basic .. AND, OR, NOT etc.. I see now exactly how a multiplexer and demultiplexer works.
Why in the world are we not using a demultiplexer as a port expander? For output only...
I mean im sure we waste some time doing it but it can be done. Using a simple buffer im sure. Never tried it but i will simulate a test in proteus later.. Will try to build a 8 bit expander for output
A few reasons, but easily number one is that we usually want latched outputs, which demux will not give you. While you have control of one output, the other outputs all return to either high or low depending on chip.
Shift register is much simpler, allows much greater expansion without necessarily needing any extra control pins, comes in more varieties with more output options and is often pretty cheap. 74HC595 is popular for output expander (if I remember correctly) and has a powerful enough latched output to drive an LED (through a resistor) from every output.
What will you be latching? One output will have what you want, whilst all other outputs will latch the 'deselected' state of all highs, or all low. It doesn't work.
Plus you are using two chips, where one would do.
EDIT: But if you are bored..... why not give it a try. You will learn more from thinking about why it does not work, than if it did.
yeah. I was reading some flip flop and stuff.. these are so cool. I dont have the money also but i plan to start on the cheapest CPLD i can find then move my way up. This way i understand the basics.
You can create some cool things....
when you say "video" you mean VGA i assume ? (or NTSC or PAL)
Yes, VGA to an LCD monitor 15 way D input. I use a lot of PICs, but have not found an affordable way to use VGA monitor for display.
Best I found so far was uVGA for about 25 euros, but the display quality is terrible (only because the designer got the timings wrong), and is still half the cost of a typical small hobby gadget.
When you learn enough about FPGAs to make a uVGA equivalent that works, I'll be the first to buy your finished product.
Too many unfinished projects on too many different MCUs already..... can't get distracted again....
Plus I always end up wanting the faster device, the better programmer, the more expensive design tool. Always ends up costing a fortune with the only end result being fun.
Let me know how it goes. Sounds like you will be having great fun messing about.
yeah. i know what you mean. I have played with PICs(8bit,16bit), AVRs, NXP ARM, ST 8bit , Parallax Propeller and those are the ones i can remember heh...
Something new would be fun. I will order the $50 one this week and give it a go. Its best to learn from doing i hear so might as well shell out $50 for some fun ...
Microcontrollers have integrated peripherals, prior to that microprocessors used the data lines as a bus and then decoded the address lines for latch, enable, read/write & clocking. That actual hard logic depended on which peripheral chip.
Modern microcontrollers, with easy bit manipulation instructions and high speed cores, serial data/clock to shift registers are an easy choice for parallel expansion. ST (596c..) and Toshiba (6270x) (& probably others) make constant current serial input drivers. Most any other peripheral combination is available in Some flavor in your series. I'm spoiled by working with Freescale products >10 years.
Building sophisticated controllers is SO much easier now... (ergo my handle) <<<)))