I've aquired a suposedly non-functional led matrix.
It's about a 1.5m long and in a nice metal case.
I've ,amaged to power it up, and it goes into a demo mode, all LED are working. It have a 15pin d-sub for a keyboard, i have a keyboard but it's in bits. It also has a 25pin rs232, which looks intresting.
Yes, of course I've opened up the case and had a look inside.
There are four large PCBs.
One has the power supply.
One has various processors and eeproms.
the other two have the LEDs.
The LED PCBs were easy to disconnect and examine, both appear to be identical.
Each has :
12 x 5x7 led matrix arranged in a row. CSM-57221E [1]
8 x 8bit serial-in parallel-out shift registers. SN74LS164N [2]
8 x Darlington transtitor arrays. ULN2003A [3]
7 x PNP darlington transistors. TIP127 [4]
a few 104 caps and a bunchof resistors.
Each has a power connector and a 14 way ribbon cabel for data (presumably). I'm not sure how much if the ribbon cable is used the connector is a 2x7 DIP.
I don't quite understand why you would want to use an arduino without all of the boards. Presumably with the keyboard input it is designed to have some kind of temporary memory to store messages...
If you could test it with a keyboard and confirm that it is like that then you could use the arduino to emulate the keyboard so you can then send it messages from anything then connected to the arduino...
This should be much simpler than trying to code all of those matrixes (matrices?).
Do you have an idea of what the dispaly is and who it is made by? A quick google search would then probably find you the data about input.
I don't quite understand why you would want to use an arduino without all of the boards. Presumably with the keyboard input it is designed to have some kind of temporary memory to store messages...
If the keyboard connector is non-standard and the RS-232 undocumented, then you may not have a choice but to replace the processor module. Or it could be that the processor module doesn't have any functionality for graphics, or something, that you want it to do.
From the pic and your description, it seems the discrete transistors control power to the rows, and the shift registers feed data to the transistor arrays to control power to the columns. The registers are probably daisy chained so that one feeds into the next as data is shifted in, likely in pairs with a single data line to each pair, and a single clock line to all 8. The caps look to be just decoupling caps, and the resistors limit the current to each column.
So the controller pumps the data for the first row into the shift registers then turns on the first row transistor for a period of time, turns the row off, pumps in new data for the second row, and then turns on the second row transistor, etc.
As far as the cable, you probably have the 7 control lines for the row transistors, maybe a signal ground if it's not just using the PS connector ground, then for the shift registers a clock line, a clear line, and at least one data line. There might be more then one data line so that the controller can fill the shift registers faster. IE, data1 to SR1, chained to SR2, data2 to SR3 chained to SR4, etc. That means sending 4 bits 8 times rather than 1 bit 64 times, so much faster.
Other than that you'll have to start examining traces and poking it with a meter, but I hope this helps.
I managed to get a keybaord working, though not all keys work.
The device has an extensive menu and capabilities.
Multiple messages, graphics, transitions, can be linked to other devices, passwords, even supports a modem.
Entering a message via the keybaord is quite laborious and the whole process is played out on the matix; that's not desireable.
I did spend a few hours trying to reverse engineer the data it sends via rs232. i could read my messages in plain text, but there were so much other data i could not fathom.
I did managed to capture the data, change the message and transmit it back. Although the message changed, there was some wierd flashing and changing the message length cause an errror. also the new message did not run automatically.
I'll put the rs232 and keyboard options aside for a while, it not a dead end but I'd take up the challenge of figuring out the hardware.
As suggested, the 7 rows are controlled through the data cable.
I found ground, clock and two data pins. The data pins are joined and it's a chain of 8x8bit shifters.
I cannot find Clear (yet), which is odd as there are spare channels in the data cable and the ls164 support a master reset.
does this mean I have to write a bunch of 0s to clear a row?
I have a freeduino hooked up to one LED board, and can squirt data in.
My next task is fiddle about with timings, loops and such to get a good non flickering display without too much brightness.
Thinking ahead, does anyone know of a source of a 5x7 font?
From the description you're giving about the matrix board. I think I had the same one... a Litek 8015. The Z80 control board is partially fried, but the demo sequence still runs.
I've come just as far as you, two data lines identified and the clock line. The RS232 has died with me.
Maybe we can join forces and figure out how to control the board.
I'm getting ready to hook it up to a Arduino Mega.
