LED moving message display - interfacing problems! RS232

Nobody here has played with RS232 or 8031's?????

Anyone using a LED message sign at work? What brand is it? How is it connected?

I'm sure at least some of us have played with RS232 and/or 8031's, just not within the message board you outlined. And even if one hasn't, it really doesn't matter anyway.

The ones that I played with (a million years ago) were RS232-interfaced, (back when RS232 was trendy), but you used a dumb terminal to talk to it. Certain characters were used for control purposes, everything else (alphanumerics) were used for display.

This doesn't help you of course, not only because can I not remember the brand of the displays, no other brand might help either - simply because there are no communications standards for these things.

I've heard of AT style keyboard connected interfaces, USB, parallel printer port interfaced, you name it, I've "heard" of them all. All of them different with the way one programmed the text.

Regardless, unless someone doesn't have that model you have and knows how to drive it, nothing else will help.

I would be cynical of the usefulness of "looking under the bonnet" because if you want to disassemble the code, by the time you go through it, it may very well be cheaper to buy something off the shelf that has software and instructions all nicely laid out for you.

When purchasing things like these, you really have to do your homework first, otherwise you might be buying a vintage black box. Very nice, but black boxes don't really do anything exiting by themselves.

Keep looking, you might get lucky and hear from someone that has used them.

Otherwise, it might be useful to stick a protocol analyser in there and see if the boxes bounce any bytes back in response to your probing. It'll give you an idea of the speed at least. Failing all the above, some trial and error may be called for.

Good luck.

As John mentioned, it's a lot of effort to try and reverse engineer something like that, and I wouldn't bother. Sounds like you've tried the basics, so I'd give at that point. What I *would* do however is to simply write new software for it from scratch, that should be quicker than trying to decode what the old one does. Either use the existing 8031 with a new ROM, or replace it with the micro of your choosing. Should use all discrete hardware, so that should be fairly easy to figure trace out and figure out how to drive the LED's.

With your own software you get the benefit of making it do exactly what you want, likely with more features than the original software contained.

Dave.

I have a suspicion that LUA morphed into Axent Global

not so sure about that Dave.

I did a moving message board (see Silicon Chip 1989, over 4 issues)

To assemble the text, make it move, calculate the next frame ahead, and have it move up, down, left or right, then add all the scroll left, right, up down, etc, etc.,

Then if you want animated 3 colour crocodiles, space vehicles, telephones, flowers, etc, etc, as I did, it may take about 18 months of your spare time. A part time labor of love.

If you just want to write "Happy Birthday" sure. That's how I started.

BTW Princo Coasters, I used an XT keyboard for data input. That's what was popular and convenient then. :-)

Sure, yours is RS-232, if you can't reverse engineer the comms code, scrap it, and buy a very cheap TW import.

Don...

With a high level language compiler and possibly some pre-written code for the high level stuff (text-to-bitmap etc), it need only take a fraction of that time. If you offload the high level stuff to the PC then the actual LED message driving part of it becomes relatively easy.

The problem PrincoCoasters might have though is that he might have to do all that anyway. Even if he decodes the software and gets it to talk he might find that the message board is nothing more than a low level device which holds a bitmap image and scrolls it. i.e. it might not have any ASCII to bitmap decoding built in, so he won't be able to simply send it ASCII text and have it displayed. In that case a PC application would have to be written anyway.

Dave.

all z80 assembly Dave, and part time as I say. Mind you, I just kept writing it until I was happy and exhausted all possibilities. :-)

started off as a single colour display, (Silicon Chip articles) but I added alternate rows of red and green LEDs to produce orange as well, so it became a 3 colour device.

Don...

You might be interested (or relieved) to know you don't hold the record for development time for a published project. I think that honor goes to Peter Baxter's 100MHz Logic Analyser in EA which took in the order of 5 years to develop, only to be narrowly beaten to publication by my 40MHz Logic Analyser project by a couple of months! (my one only took a few months to develop)

Dave.

Be aware that Z80 code differs widely from 8051 (8751, 8031) code. I used both, ever wrote even disassemblers.

petrus bitbyter

I've never programmed a 8031 or any of its relatives, but after a quick consultation with Mr G, I would guess it is at location 003B (reference to SBUF, aka SFR 99 = Serial Data Buffer).

Could that be some 'command' decoding just beyond that?

Andy Wood snipped-for-privacy@trap.ozemail.com.au

"princo coasters" schreef in bericht news:46b2b9cc$0$18984$ snipped-for-privacy@news.optusnet.com.au...

As I said, you really need the datasheet of the 8051 (8031, 8751).

The first instruction jumps to address 0x800 where a lot of registers are initialised. Amongst them SCON (Serial Control) which is set to 0x70 so an

I checked neither the Baudrate nor interrupt setting but I found the serial interrupt service routine at address 0x23. It looks like the routine handles some other interrupt(s) as well but at address 0x2D it checks the receive interrupt flag (RI) explicitely. If set it jumps to 0x3B and picks the received byte from SBUF. From that line on you can find out what actions are taken depending on the contents of SBUF. A 0x1B (ESCAPE) for instance results in placing a 0 in r2. The routine goes up to 0xD2 but always exits via 0x3A (the reti instruction). No doubt the actions depend also on the contents of the registers which in turn may contain values set by earlier received bytes. I saw at least r0 through r5 used. But you can find out what characters have some meaning and which ones, if any, are discarded.

petrus bitbyter