[Lcdproc] Re: hd44780 goes 4-bit

[David W Studeman]

Pierre Ossman wrote:

> Hi again,
> 
> I've scoped the stuff now and the waveform looks nice and clean. The
> delays are also well within the specified limits. The card also already
> has a pull-up, so now point in adding another one.
> 
> The module is a Samsung 16T202DA1 with a HD44780 compatible NEC
> D16314AGJ-011 chip on it. I don't suppose it has any know issues? I
> couldn't find a data sheet for the chip to check for any special
> requirements.
> 
> I'll see if I can get the module replaced, in case it's simply broken in
> some well hidden place.
> 
> Rgds


I doubt it's broken. My Cobalt Raq 4's and Symantec Velociraptor do the same
thing and I even went so far as to try a new Sure Electronics DE-LM003
which is yet another HD44780 compatible display and it behaves the same
way. With these units, if you were to be able to convert to 4bit, you would
lose the rom functionality at power on which allows you to change where it
looks for the boot kernel and so forth. One of the IPCop developers told me
that the two bytes that compose a command are sent too quickly by LCDproc
and he sent me a perl script that will drive the display by itself with no
external driver. It puts out clean text but eats more cpu cycles than
LCDproc does. I compile my kernel without the built in Cobalt LCD driver to
be able to do any of this. On a side note, the old Cobalt LCD and button
utilities are an endless maze of perl scripts that would need a lot of work
to work with IPCop so I keep trying to find an alternative.

 Dave