>         I've been having problems reading the busy flag on an LCD. The LCD
> is using
> a 4-bit interface. I have finally figured it out (I think). It appears that
> the order you send
> the various control signals matters (as the datasheet indicates, but several
> examples
> I found on the web don't follow that order).
>
>         It is also important that the data lines *NOT* change. I've tried
> setting the
> pins to HiZ before, after, and at the same time as setting 'E' high. The LCD
> gets
> messed up every time. The only way it works is to leave the I/O pins
> "as-is".
>
>         The code is for an ADuC812, which has an 8051 core. I would like to
> port this code to a PIC.        Now, on the 8051, I can simply let the
> external signal
> drive the I/O pins, even though they may be outputs.  But on the PIC, which
> has
> output drivers, this won't work. I've though of using the weak pull-ups (on
> an 'F84),
> but aren't these disabled when the port is configured as an output? Do I
> have
> to add hardware to support this?
>
>         Does anyone have any ideas of how to accomplish this on a PIC? Has
> anyone successfully read the BUSY flag while using a 4-bit interface? I'd
> greatly
> appreciate some code snippets if you're willing to share.
>
> Phil Eisermann
> H:(440) 284-3787 (mazerspam_OUTix.netcom.com)
> O:(440) 329-4680 (@spam@peisermaKILLspamridgid.com)
>

> Actually, it's bit 7 of the 8-bit word read that includes the BUSY flag
>  and the DDRAM address or some such -- whether you use the 8-bit or 4-bit
>  interface.  And we're trying to read the BUSY flag, so the PIC pins gotta
>  be inputs.  No?
>
>  I've tried reading the pins once and twice, since the data sheets I have
>  are maddeningly sparse in precise detail and riddled with bits that don't
>  make sense, explain things halfway, or conflict with other sections. In
>  8-bit mode things work fine, but not in 4-bit mode.
>
>  Dale

>
> On Thu, 20 Apr 2000, Tim Hamel wrote:
>
> > Dale,
> >
> > Ok...I think I understand what you're saying...you can't read the busy flag
> > (bit 7) unless you're using the 8-bit interface? I've found something that
> > might contradict that:
> >
> > home.iae.nl/users/pouweha/lcd0.htm#4bit-transfer
> >
> > Apparently, you CAN read bit-7 regardless of the i/f type.
>
> In *theory*, yes.  In *practice*, we're having trouble doing so.  Every
> time I try it, the LCD goes south for the winter.  That's why we're
> working on this.  It's more than a little frustrating -- I now have a
> routine that uses 1ms delays, which works perfectly.  But now it's getting
> personal, and I wanna squash a bug.
>
> Dale
> ---
> The most exciting phrase to hear in science, the one that heralds new
> discoveries, is not "Eureka!" (I found it!) but "That's funny ..."
>                 -- Isaac Asimov

> rather simple. The way I see it is:
>
>  RS=Low
>  R/W=High
>  E=Low -> High
>  Read Bit-7
>
>  No offense, but is that how you're doing it? If it is, and the LCD still
> goes schitzo, then I'll work on this as a weekend project (teenagers have
> more free-time than working adults <g>). If I figure it out, I'll be more
> than happy to post the code!
>
>  Regards,
>
>  Tim Hamel

>
> > Dale,
> >
> > While I haven't actually tried a BUSY bit read (I use a delay too), it
> > seems
> > rather simple. The way I see it is:
> >
> > RS=Low
> > R/W=High
> > E=Low -> High
> > Read Bit-7
> >
> > No offense, but is that how you're doing it? If it is, and the LCD still
> > goes
> > schitzo, then I'll work on this as a weekend project (teenagers have more
> > free-time than working adults <g>). If I figure it out, I'll be more than
> > happy to post the code!
> >
> >
> well, i'm not Dale, but:
>
> yes, that's how I'm doing it. except E toggles low->high->low.
> and with a 4-bit interface, you're supposed to do it twice.
>
> That's the way it is supposed to work in theory. In
> practice, you need to change the data lines from
> input to output in order to read the data. And as
> soon as you do, the LCD (at least mine, and
> apparantly Dale's as well) goes south.
>
> As stated previously, it works fine on an 8051,
> where you don't have to change the pins from
> input to output. The LCD easily overpowers
> the weak pull-ups. But now i'm trying to get
> it to work on a PIC.
>
> Perhaps my LCD is half-dead/ESD impaired?
>
> If you figure it out and post the code, that
> would be greatly appreciated.

>
> Dwayne Reid wrote:
> > 1) It should NOT matter what happens to the data lines so long as the E
> > line is LO.  Make sure that E is LO before changing anything.
> >
> > I *know* that the LCD ignores data while E is LO because I have 68hc11
> > boards that put the LCD on the data buss - and they all worked just fine!
> >
> >
> well, I tried that this morning. All i did was change the data lines while
> E was low (confirmed on scope). The order of operation was:
>         1) change data lines to input
>         2) bring RW high
>         3) bring E high
>         4) read data DB7-DB4
>         5) bring E low
>         6) repeat 3-5 for reading DB3-DB0
> Note that E, RS, RW are all low at the start of the operation.
>
> The LCD remained blank, but data was sent continually back and forth.
>
> The version that I was able to get to work omitted step 1.
>
> I have another of the same display. I think I will wire that one up
> and see if it behaves similarly.

>
> I apologize...E does NOT go from L->H, it's negative-edge (High->Low).
>
> Sorry,
>
> Tim Hamel
>
> In a message dated 4/21/00 9:52:58 AM Pacific Daylight Time, TekPhobia writes:
>
> > Dale,
> >
> >  While I haven't actually tried a BUSY bit read (I use a delay too), it
> seems
> > rather simple. The way I see it is:
> >
> >  RS=Low
> >  R/W=High
> >  E=Low -> High
> >  Read Bit-7
> >
> >  No offense, but is that how you're doing it? If it is, and the LCD still
> > goes schitzo, then I'll work on this as a weekend project (teenagers have
> > more free-time than working adults <g>). If I figure it out, I'll be more
> > than happy to post the code!
> >
> >  Regards,
> >
> >  Tim Hamel

> Anyway, here's the code as I'm currently using it.  The LCD uses PORTB0:3
> for the data lines, PORTA.0 for RW, PORTA.1 for RW, and PORTA.2 for EN.
>
> void waitlcd(void) {
>         char temp;
>         TRISB = 0b.0000.1111;   // Input from LCD
>         RS = 0;                 // RS low to get BF
>         RW = 1;                 // Read operation
>         do {
>                 EN = 1;         // EN up to read data
>                 temp = (PORTB & 0x08);  // Mask off all but BF
>                 EN = 0;         // Drop EN
>                 EN = 1;         // Finish up last 4 bits
>                 EN = 0;         // and finish the read
>         } while(temp != 0x08);  // (we ignore the low bits)
>         TRISB = 0;              // All pins output again
>         RS = 1;                 // and set up to send to LCD
>         RW = 0;
> }
>