bought a MAX232 and 5v reg stuff, bidding on some pic16f877's on ebay! just gotta start on a schematic, any things i should know?
ok, have got a programmer (on its way) some PIC 16F77 (40DIP jobs) and a MAX232... now then, schematic time.... any theries on pin layouts... i will copy bits from other stuff i have.. Will let you know in a few days..
Just a little update and a few questions i hope some nice person can answer: I have been looking at the datasheets of how the display portion operates and have worked it all out. It uses 4 BCD signals (4 lines) and a clock signal input from the controller as well as a ground connection. 6 cores total. (its gets a separate 5v line from the PSU) the BCD signal works the rows, the clock works the columns) Now, what i want to do is test the display part to make sure it has not been damaged by static or anything and figure out which LEDs need replacing. I have been looking through and think i have an ideal solution. The input signals are buffered by a 74ls244 IC, the 4 BCD lines going to a 7445 BCD - decimal decoder. the clock line goes to the column driver "area" (its all way to complicated). Now, can i ask this question: If i plug power onto the board and supply 5v (what the display works on - regulated 5v) to 2 of the bcd lines (so all outputs are high)(or other BCD lines - since it might need them low to light the LEDs) and give it a manuel clock via a button to 5v and get columns lighting up, moving along one place with every clock. My main worry is of doing something stupidly obvious to someone more talented than me. i wont kill anything by doing this will i? will this work?
Is there any indication of the resistors used for the LED's? By running the clock at such a slow speed (i.e. manually) it's possible that you could burn out all the LED's if they've selected tiny resistor values to account for the low duty cycle of the multiplexed display.
There is resistors on the LED's, both on the columns and rows, two 1k's (in series) On the bases of the row transistors, 39 ohm on the columns. my calculator shows that a standard LED @ 5v should be 150 ohm, so i guess this would be ok. maby i should connect a 555 timer to the clock and give it some serious clocking... # edit: BTW the leds are the see through type, ie withought knowing you wouldent know what colour they are.. so maby they are higher voltage type..
The base resistor isn't going to limit current enough, and 39 ohms is a bit low, so drive the clock at a decent frequency to prevent LED blowage.
ok, i have been working on this thing and have not really got anywere special.. i connected the clock line to the output of a 555 chip giving about a 3hz output. i powered the 555 circuit from a different PSU, like it was origionally, i got weird results when running it from the same source as the sign board. (the status led on the 555 circuit dimed very slowly when turned off..) anyway, at one freak moment, one row did a lovely chaser pattern to about the middle of the board and then went out and i spent an hour trying to get it to do it again. i now have only 6 wires to worry about that enter the sign board (apart from power) these are: Clock for column select 4 BCD lines for row select (rows pulled low to activate) ground line for 74ls244 chip. uhh, pray for me...
Some small developments in this field. I have discovered that the sign works very similar to this one here: HERE with the exception of 2 things... 1) the rows use 164n's instead of the 154n's as used there, but are otherwise identical, there is also no data line, the first 164 has its A and B line shorted and going nowhere.. 2) instead of having the 7 base lines for the column transistors, it is converted into BCD, now this is where it gets uber weird.. Looking at the truth table for the 7445 BCD - Decimal decoder, it looks like you can only light one column at a time, so the pic would have to clock a row, give a BCD code to light one LED, give another BCD code to light the next one and then clock the next digit... am i right? i plan on using PIC 16f77 for this, i am trying to decypher he code on that site, is it worth trying to modify that for this application?
