Maximun Cable Length for PIC to LCD

[wildbilly]

Hello all, I have just finished a project which includes displaying on three LCD displays. Thanks to a great method by Darrel Taylor. http://www.pbpgroup.com/files/2_LCDs.gif
His idea works fantastic. What is the maximum cable length of cable that can be used between Pic and LCD? I have all the three LCD working close to the PIC but 20 feet away I get gibberish. Am I just stuck with a shorter length of cable or is there a work a around.

Thanks.

[KVLV]

i guess max around 2feet for TTL level. if you require longer length i suggest using serial lcd.

[Darrel Taylor]

wildbilly,

Great, glad that's working for you.

I can't answer the main question, as I've never tried to go more than a couple inches to an LCD.
But, I'll agree that serial might be a better way.

However, I think there's two possible sources for the garbled text. And I might be able to do something about one of them if you still want the parallel interface.

1. PBP sends everything out to the LCD at the MINIMUM pulse widths defined in the HD44780 datasheet. Which means approximately 1 Mhz squarewave signals since the enable pulse only lasts 1 uS. That's going to be tuff to get through 20 feet of ribbon cable.
   <br>
2. The other source could be inductive pick-up, like an antenna.
   Can't help much if that's the case.

If the problem is #1? It may be possible to slow down the LCD data to a point that will propagate thru the long ribbon.

Using the technique described in the LCD_Anypin thread, we could create whatever timing is required to slow it down to a reasonable speed.

If you want to try, let me know.
You'll have to do the testing though.
<br>

[wildbilly]

Thank you KVLV. I think you are correct about using serial instead. It would be nice though to be able to use the parallel interface. I do enjoy a challenge and it keeps PicBasic interesting.
Thanks Darrel, I have read the LCD_Anypin thread and that is very interesting. I am up for the leg work. I am willing to try and I will do all the testing required.

Thank you.

[Darrel Taylor]

I am up for the leg work. I am willing to try and I will do all the testing required.

Shouldn't take much, but you'll need to install the HighJack routines first.
And you'll also be using the LCD_Anypin files, so download them too.

Start with that, and you should be able to get the LCD working on a Short cable first.

After that, we'll adjust the delays and see what happens on a long one.
If you have any problems getting LCD_Anypin to work, let me know.
<br>

[w7ami]

wildbilly,

It should be possible to send data over 20 feet to your display and have it work, but not using ribbon cable. All the conductors are parallel to each other and divided by insulation. In other words you have a bunch of capacitve coupling between the lines. Also the ribbon cable is likely not shielded so any manner of garbage will be picked up from the nearby environment. I have an absolute encoder (10 lines of TTL data) working over a distance of 50 feet in a high RF environment.

First as stated earlier make sure you have cleaned up the power to the LCD with good bypass capacitors with short lead length. You will need to put the displays in some sort of metal box. Then use a shielded cable with twisted pairs (signal and ground on each pair) between the two units. Then if you are still having problems mess with the pulse widths.

Terry

[Ioannis]

....
2. PBP sends everything out to the LCD at the MINIMUM pulse widths defined in the HD44780 datasheet. Which means approximately 1 Mhz squarewave signals since the enable pulse only lasts 1 uS. That's going to be tuff to get through 20 feet of ribbon cable.

....

Well in this case, I would suggest to use double the number of flat cable wires and the middle ones connected to the ground point. It won't eliminate the problem but will help for sure.

Ioannis

[mister_e]

at very least, if you really want to waste that amount of cable, place some cap close of the LCD VDD/GND (47uF (or greater) + 10 uF tantalum and finally 0.1 uF ceramic)

I don't think there's a real problem with the length of cable so far, my feeling is that you just need to filter the Vdd rail properly at the end and probably play with COMMANDUS & DATAUS DEFINEs.