Most people know one parallel port can drive 12 channels (the trick being the 4 control lines have to use a buffer chip like uln2803 or LS244 and can't go straight into SSRs as with the 8 data lines. And sometimes even the 8 data lines need driver/buffer chips).
But now that I know how to make the control lines work I want to tackle the parallel64 with one parallel port running 64 channels (so if you have 3 parallel ports you can run 192 channels)
My first thought was to go first with parallel24 that uses the PPi chip that runs 24 lines from one parallel port but that would only give a total of 72 channels with 3 parallel cards so I decided the parallel64 with the possibility of 192 channels was more dramatic (LOL)
Now my decision is about with the parallel64. Should I use a 3 to 8 decoder on the control lines (3 lines, 3 bits, counts out to 8 in binary). So with 8 data lines and 8 select lines I get 64 channels. This seems the simplest. So I'd have a 192 channel system with 3 ports.
However if instead of 3 to 8 decoders on 3 of the control lines if I use 2 of the 4 to 16 decoder chips on the 8 data lines they give me 32 channels and then with control lines I can pick 2 multiples of this to get the parallel64.
But which way would be best? Has anyone tried parallel64? Any suggests about which is the easiest way to do the parallel64 on one parallel port. (Already figured out the way to modify the parallel12 driver to do parallel64 so don't need help on that) Just need some ideas from somebody who's done the electronics side of a parallel64.
This sounds much like the Grinch 64-channel controller - you may want to look at that as an example how to control 64 channels with a parallel port interface. If not that then the older 595 controllers might get you there also. Here's the Grinch: http://www.doityourselfchristmas.com/wiki/index.php?title=The_GRINCH_Controller .
Ray,
I applaud your continued efforts to create you own solutions. I understood the desire for the Parallel12, simple and easy to put to into action. But your new idea is either just a strong desire to do something unique or the lack of understanding of what is currently available.
The Olsen 595 and Simple 595 plug-ins will give you over 1000 channels (based on how many pulses could be sent out during a single VIXEN event) with just using 3 lines on the parallel port and the 8-line 595 plug-in would be over 8000 using 10 lines.
So if your goal is more channels, then it seems that existing methods would be the way to go.
If your goal is just to do something unique then by all means continue on your path.
Don't want to discourage but wanted to make sure you knew what was available.
Don't agree with you that it's anything unique or original at all.
I'm almost positive that people have done parallel64 many, many, many times before since parallel was the number 1 experimenter/ engineer's choice up till 2004 (well after the arrival of USB by the way). And there are 1000s of projects from robotics to temperature sensors and more , all using parallel port.
Believe also that many people have even done parallel64 with christmaslights (just got a link from ErnieHorning that will pursue at computerchristmas.com)
Yes, the Olsen595/grinch is my next project after the parallel64. (Doing this for fun and knowledge). My understanding of it now (will check later when I get into that project) is that it is basically using shift registers to turn the parallel port into a type of pseudo-serial port. But as I said I'll get to that later when I start that project.
Back on topic of the thread: My concern at the moment as I originally stated was between a 3 to 8 decoder and a 4 to 16 decoder. The concern is if the 4 to 16 decoder choice would limit me to only 32 lines by using too many control lines for chip select etc. and not allow the full 64 lines.
Followed ErnieHorning's suggestion and looked up the article by Hill Robertson at computerchristmas.com
Not just parallel64 but actually parallel320.
That is Hill Robertson gets 320 channels out of one parallel port.
But it takes him 240 writes to the parallel port to get all the data out.
That is because he uses the 8 data lines in 2 ways,
first to set the 8 data bits for a latch (set of 8 channels)
and then to choose the steering to the right flip flop out of 40 (40 by 8 for the 320).
And he steers to the right latch of the 40 by dividing the 8 parallel lines into,
Five data lines to chose one of 5 groups,and the remaining 3 data lines to feed the 3to8 decoder to select one of 8 latches in the group (5 by 8 to select one of the 40)
So his 6 write procedure (data, control, control, steering, control,control) to select one latch has to be looped through 40 times. Wow.
But he's done it and says the 240 writes produce no noticeable effect for the human eye even on a slow old computer.
Still it's an accomplishment. As he says "I wanted to prove to myself that I could do it....So I did"
Bravo. It's all for fun and to prove you can do it.
With 3 parallel ports that would allow one to control 960 light channels. Not that I'd ever want to run that many light channels.
But when I look at what he has to do it may be similar to the Olsen595 in some ways. But I'll keep that for later.
Still not answered by questions about using 4 to 16 decoders instead of 3 to 8 decoders.So maybe I'll just try with 4to16 decoders first and see what comes out.
