Digital outputs = just connect them up (eg 573 latches)
Digital inputs = use a 74HC244 and the HiZ control goes to /board_select
Bidirectional bus when you can control the direction = use a 74HC245
Bidirectional bus when you don't know the direction = use a 74HC4316 (eg for keyboard/mouse)
...
I hope there are no bugs I haven't thought of. 244 I/O pins would be cool!
That's still a lot of hard-wiring configuration ? - & PCBs cost by area...
I see another candidate for low cost smarter I/O is a Z51F3220/3221 - comes in at 4.9c or 5.6c per IO (100+)
Nice wide Vcc, <2.0-5V, and it has 12b ADCs, and quite a few serial options 2 SPI and 2 UARTS/SPI/i2c, and easier to code than the Mini51, I mentioned above, and has much more code space @ 32K, tho you'd struggle to use even 10% of that for IO expander work.
Can also support a RTC, etc - imagine the COGS you can free up for more serious work
Thanks for the clarification, and referring to your comment:
Oh, and by the way idbruce, I have some good news for you
That is good news, but it is also old news, she called last week looking for a shoulder to cry on. I didn't have time for all her sob.. sniffle... sob.. sniffle... mumbo jumbo, so I gave her your number and told her to give you a call
That is good news, but it is also old news, she called last week looking for a shoulder to cry on. I didn't have time for all her sob.. sniffle... sob.. sniffle... mumbo jumbo, so I gave her your number and told her to give you a call
Ha ha. Yes - cats are much lower maintenance than supermodels!
Of course my line of thinking is much different then yours, but I am still leaning towards the MAX7301AAX or MAX7301ATL IC. Not that I am any electrical genius, but here are my thoughts.
Let's suppose a master board, having a primary purpose of supplying power, holding a propeller chip with prototype area, as well as providing connections to numerous slots (something akin to PCI). These slots would be intended to hold daughter boards, with each board having a MAX7301AAX or MAX7301ATL IC, and also having a prototype area. These "slots" are all inter-connected to enable MAX7301 daisy chaining. I personally think this would be a slick setup for some complex development.
That would be 28 pins per daughterboard, plus the leftover prop pins.
That would work and it could be interesting to consider that in conjunction with a parallel solution. You could also extend the 2 I2C pins onto all such daughterboards.
I suppose I am thinking more in terms of parallel rather than serial as the speeds are higher. eg the touchscreens have SPI interfaces for the display but you can't get the insane 64 megabit speeds to update a screen like you can with parallel.
Or maybe it is that I am turning into an old fuddy-duddy who thinks any chip newer than the 74xx series is just too new fangled to deal with!
So it is absolute speed you seek in addition to more pins
Well both actually. I guess I tend to start thinking in generalities (wouldn't it be nice to have lots of pins and lots of cogs and lots of ram) and then consider what the practicalities dictate.
For instance, I see just now that someone has posted on the Prop II blog a long comment about the fact the prop II would be surface mount. But if you want lots of pins DIP is harder so it has to be surface mount. Is it a case of being careful what you wish for?!
For simple text displays speed does not matter. 80x40 is only 3200 bytes. Once you go to graphics, speed becomes important.
And for large programs (LMM and XMM) the faster you can get code out of external memory, the faster it runs.
But there are other instances where speed does not matter so much. Analog I/O for instance. So maybe one could consider a common SPI bus with 3 common pins, and then alongside that, another bus to decode which SPI device to address (maybe even an I2C bus with I2C expander chips). Your max chips could be part of that.
Having said that, thanks to all the great comments on this thread, I think I have enough now to think about a design for a generic motherboard. I am thinking of something like the Gadget Gangster format, but bring out another 8 common pins driven by a 137 to select which board in the stack is active. For prototypes I will probably go for a larger board, maybe the eurocard size of 100x60mm, but if it works I am sure there will be clever boffins who will be able to shrink it.
For daughter boards, one could be the dracblade, one might be a TV/VGA/Keyboard/mouse, one might have some propeller co-processors on it, one could be an I/O board with your max chips, spi etc, one would be the touchscreen and its two 512k memory chips. One could also replicate the I/O of the demoboard and the C3. What I am thinking of is a stack where you can switch between multiple hardware formats with some simple software (about 7 lines of spin) so that you can develop languages like C in multiple hardware environments at the same time. Test that "printf()" works on TV, VGA and a touchscreen.
I'm also trying to keep things backwards-compatible with all the clever things we already have working.
I'll see if I can draw something up tonight after work that might explain this a bit better.
Of course my line of thinking is much different then yours, but I am still leaning towards the MAX7301AAX or MAX7301ATL IC. Not that I am any electrical genius, but here are my thoughts.
Let's suppose a master board, having a primary purpose of supplying power, holding a propeller chip with prototype area, as well as providing connections to numerous slots (something akin to PCI). These slots would be intended to hold daughter boards, with each board having a MAX7301AAX or MAX7301ATL IC, and also having a prototype area. These "slots" are all inter-connected to enable MAX7301 daisy chaining. I personally think this would be a slick setup for some complex development.
That would be 28 pins per daughterboard, plus the leftover prop pins.
Bruce
I know this is an old thread, BUT.. did you ever do anything else with the Max7300/7301? I have an application where I need a lot of pins for inputs and I like this chip and don't care about high speed.
I too have a project needing quite a few I/O, and they need to be able to flip back and forth between input and output. So, the thing I dis a while ago was buy a 100 or so of the small package TI430 chips, Each has 10 port pins, and I use 8 for I/O and 2 for serial comms to the propeller.
Hook up as many as you like...... about 40 cents each as I recall. VERY flexible chip.
Comments
That's still a lot of hard-wiring configuration ? - & PCBs cost by area...
I see another candidate for low cost smarter I/O is a Z51F3220/3221 - comes in at 4.9c or 5.6c per IO (100+)
Nice wide Vcc, <2.0-5V, and it has 12b ADCs, and quite a few serial options 2 SPI and 2 UARTS/SPI/i2c, and easier to code than the Mini51, I mentioned above, and has much more code space @ 32K, tho you'd struggle to use even 10% of that for IO expander work.
Can also support a RTC, etc - imagine the COGS you can free up for more serious work
Thanks for the clarification, and referring to your comment:
That is good news, but it is also old news, she called last week looking for a shoulder to cry on. I didn't have time for all her sob.. sniffle... sob.. sniffle... mumbo jumbo, so I gave her your number and told her to give you a call
Bruce
@idbruce
Ha ha. Yes - cats are much lower maintenance than supermodels!
Of course my line of thinking is much different then yours, but I am still leaning towards the MAX7301AAX or MAX7301ATL IC. Not that I am any electrical genius, but here are my thoughts.
Let's suppose a master board, having a primary purpose of supplying power, holding a propeller chip with prototype area, as well as providing connections to numerous slots (something akin to PCI). These slots would be intended to hold daughter boards, with each board having a MAX7301AAX or MAX7301ATL IC, and also having a prototype area. These "slots" are all inter-connected to enable MAX7301 daisy chaining. I personally think this would be a slick setup for some complex development.
That would be 28 pins per daughterboard, plus the leftover prop pins.
Bruce
I suppose I am thinking more in terms of parallel rather than serial as the speeds are higher. eg the touchscreens have SPI interfaces for the display but you can't get the insane 64 megabit speeds to update a screen like you can with parallel.
Or maybe it is that I am turning into an old fuddy-duddy who thinks any chip newer than the 74xx series is just too new fangled to deal with!
All computers should be powered by steam, I say.
So it is absolute speed you seek in addition to more pins
Well both actually. I guess I tend to start thinking in generalities (wouldn't it be nice to have lots of pins and lots of cogs and lots of ram) and then consider what the practicalities dictate.
For instance, I see just now that someone has posted on the Prop II blog a long comment about the fact the prop II would be surface mount. But if you want lots of pins DIP is harder so it has to be surface mount. Is it a case of being careful what you wish for?!
For simple text displays speed does not matter. 80x40 is only 3200 bytes. Once you go to graphics, speed becomes important.
And for large programs (LMM and XMM) the faster you can get code out of external memory, the faster it runs.
But there are other instances where speed does not matter so much. Analog I/O for instance. So maybe one could consider a common SPI bus with 3 common pins, and then alongside that, another bus to decode which SPI device to address (maybe even an I2C bus with I2C expander chips). Your max chips could be part of that.
Having said that, thanks to all the great comments on this thread, I think I have enough now to think about a design for a generic motherboard. I am thinking of something like the Gadget Gangster format, but bring out another 8 common pins driven by a 137 to select which board in the stack is active. For prototypes I will probably go for a larger board, maybe the eurocard size of 100x60mm, but if it works I am sure there will be clever boffins who will be able to shrink it.
For daughter boards, one could be the dracblade, one might be a TV/VGA/Keyboard/mouse, one might have some propeller co-processors on it, one could be an I/O board with your max chips, spi etc, one would be the touchscreen and its two 512k memory chips. One could also replicate the I/O of the demoboard and the C3. What I am thinking of is a stack where you can switch between multiple hardware formats with some simple software (about 7 lines of spin) so that you can develop languages like C in multiple hardware environments at the same time. Test that "printf()" works on TV, VGA and a touchscreen.
I'm also trying to keep things backwards-compatible with all the clever things we already have working.
I'll see if I can draw something up tonight after work that might explain this a bit better.
I know this is an old thread, BUT.. did you ever do anything else with the Max7300/7301? I have an application where I need a lot of pins for inputs and I like this chip and don't care about high speed.
I too have a project needing quite a few I/O, and they need to be able to flip back and forth between input and output. So, the thing I dis a while ago was buy a 100 or so of the small package TI430 chips, Each has 10 port pins, and I use 8 for I/O and 2 for serial comms to the propeller.
Hook up as many as you like...... about 40 cents each as I recall. VERY flexible chip.
Cheers,
Peter (pjv)