I started building the motherboard for the matrix 1000 editor today. I had the basic schematics done since a while back and have also been fiddling with the stripboard layout using LochMaster.
As I won't be building the full editor using strip boards there is no need for a full motherboard able to handle all the different editor modules.
I figure two modules will do for testing purposes so that is what I am going to support on this board.
Here's a photo of what it looks like so far:
Not much to hang in the Christmas-tree as we say, but it's a start.
The two black box headers are the connectors for the two modules. The two blue headers are power supply connectors carrying +5V, +3.3V and GND for the two modules.
The +5V will come from the green connector in the bottom right (with a reservoir capacitor next to it).
The problem is the +3.3V voltage. I am going to use the A/D converter integrated in the microcontroller on the LPC-P1203 board. The reference voltage for the A/D controller is hardwired to the +3.3V supply on the LPC development board.
So, no separate digital and analog supply for me, at least not in this prototype stage. If you look at the board you see I have drawn some lines on it. That is where the connector that will connect the LPC board will be. The pin to the far left carries +3.3V from the on-board regulator of the LPC board and is supposed to be able to handle 800mA. I don't know how much of that is already being used but I will need less than 20mA during this prototype stage so I guess it will be OK.
The sockets are for the IC's that will go on this board. They will be a buffer for the address lines, two bidirectional buffers near the black headers for the data lines, an OR chip for some R/W-control logic and a 4-16 decoder for selecting which module should be active (by enabling/disabling the bidirectional buffers).
The BIGGEST problem is:
I have not investigated how LCD displays should be wired on a module. I have a small nagging concern that I might need to send the module adress signal (like a chip select) to prevent the LCD from getting latch pulses that are not meant for it.
I have no "module select" line. I just use the module adressing for controlling the R/W line, so any module not adressed will be in read mode (presenting data to be read) so no change should be possible on a module. But as I said, I am not familiar with the LCD displays so I don't know if I can put one in a weird state by giving it latch pulses when it should do nothing. I'll have to look at that a bit closer and maybe even add an LCD to the module so I can try it out.
Also, I didn't put pull-ups on the data lines after the data bus transceiver... Duh? I hope I won't have to scrap this board but there's not really a lot of space left on it for alterations. Adding an AND chip for combining the module adress and latch should be possible but the other module isn't prepared for it.
Adding 8 pullup resistors between the data transceiver and the connector will be messy but doable...
To be continued...