lcdproc/docs/lcdproc-user/drivers/ppttrouble.docbook

<sect1 id="ppttrouble">
<title>Parallel Port Troubleshooting</title>

<para>
Unfortunately attaching an LCD module to a parallel port is not
trivial.
</para>
<para>
In most cases it requires soldering abilities and basic knowledge
of electronics.
</para>
<para>
The following hints might be helpful:
</para>
<sect2 id="ppt-check-wiring">
<title>Check The Wiring</title>
<para>
Wiring errors can easily be made. If you are unexperienced
with the soldering iron better have someone solder it for
you. Display modules are sensitive to electro static
discharges, so touch an earthed surface (computer case,
water pipes...) before you handle these.
</para>
</sect2>
<sect2 id="ppt-power">
<title>Power Source Unregulated / Noisy</title>
<para>
Make sure your power supply delivers steady 5 Volts with-
out noise or interruptions. The bare wall plug in trans-
former is not suitable, though you can make it stabilized
by adding an 7805 and a few capacitors.
Some noise induced in the supply lines my be tricky to
track, even if you have an oscilloscope.
</para>
</sect2>
<sect2 id="ppt-GND-lift">
<title>Ground Lift</title>
<para>
The power supply wires and especially the GND wires should
be a little thicker than the other wires. If GND is not
thick enough (or not existent, see 1) the resistance of the
wire may cause differing GND potentials in the circuit.
This may lead to strange display behaviour. It may also be
wise to solder a 100nF capacitor directly to the GND and
VDD pins of the display.
</para>
</sect2>
<sect2 id="ppt-latchup">
<title>Latchup</title>
<para>
Never let the supply voltage get much below the io signal
voltage. It may lead to a latchup condition which will
destroy the controller chip on the display.
</para>
</sect2>
<sect2 id="ppt-contrast">
<title>Contrast</title>
<para>
If you don't see anything on your display it may be that
your contrast voltage is set wrong. Turn your contrast
potentiometer all the way to the end connected to GND.
Contrast is highest then.
</para>
<note>
<title>Beware</title>
<para>
The module you got so ultra cheap may be an
enhanced temperature model which needs a negative
contrast voltage for sufficient contrast - see chapter
99 on how to make negative voltage.
</para>
</note>
</sect2>
<sect2 id="ppt-voltage">
<title>Parallel Port Voltage</title>
<para>
Many modern mainboards and especially notebooks will not
nearly output 5V for a logic H as the older parallel ports
did, because the operating voltage of computers is lower than
5V these days. I have measured voltages between 2.5V and 4V
for logic H, which is barely within specification of the
HD44780. If you account RCL of your cable, this may not be
enough and can cause unreliable operation.
</para>
</sect2>
<sect2 id="ppt-signal-rise-timing">
<title>Enable Signal Rise Time</title>
<para>
If you ever read the HD44780 datasheet you will notice that
somewhere in the 'signal timing' table is written: 'Enable
Signal Rise Time  max. 20nS'. That means the Voltage on the
HD44780 pin called 'Enable' has to rise from 0 Volts to 5
Volts within 20 Nanoseconds and the other way round. They
should better print that in big fat red letters, because
most HD44780s are really picky about the enable signal rise
time.
</para>
<para>
That is a Problem: If you count together the bad driving
characteristics of the parallel port combined with the
capacitance of flat ribbon cable you may easily get an
order of magnitude slower rise time.
Therefore you should only use really short cable ( shorter
than 50cm) for connecting the display to the parallel port.
It may also be useful to use pull-up resistors on the display
module or a schmitt-trigger.
</para>
<note>
<para>
The rise time of a digital output can (usually) not be
altered by Software.
</para>
</note>
</sect2>
<sect2 id="ppt-emi">
<title>EMI</title>
<para>
The cable from the parallel port to the display may
be sensible to electromagnetic interference and may emit
electromagnetic radiation. If you place your cellphone near
the cable, you may get unexpected display readings, on
the other hand your house neighbour may not be able to listen
to his/her favourite radio station any more - so better use
shielded cable and put the display in a metallic case, perhaps
a computer case.
</para>
</sect2>
<sect2 id="ppt-black-lines">
<title>One or Two Black Lines</title>
<para>
If you see one or two black lines on the display it means
nothing more than that the display is powered and contrast
voltage is present. If one or two black lines appear the
controller has not been reset properly by the on chip power
on reset generator. No need to worry - it will be reset by
the <application>LCDd</application> software. But if the
black line will not disappear
although the wiring is working, the controller on the display
may be defective.
</para>
</sect2>
<sect2 id="ppt-too-fast">
<title>Software Too Fast</title>
<para>
If you have a super GHz computer it may happen that the signal
timing generated by <application>LCDd</application> is too fast.
Adjust DELAYMULT in the
source file to a bigger value. Parallel port wirings usually
don't permit to read back the busy flag of the controller chip,
so timing must be adjust so that the controller never is busy.
</para>
</sect2>
<sect2 id="ppt-backlight">
<title>LED Backlight</title>
<para>
Check whether you need a resistor for your LED Backlight and
which value it should have. If you forget the required resistor
the backlighting LEDs might become hot and draw excessive
current.
</para>
</sect2>
<sect2 id="ppt-hd44780-compatible">
<title>HD44780 Compatible</title>
<para>
The original HD44780 controller that we advertise to support
has become the industry standard for alphanumeric character
displays. The original HD44780 is out of production. It has
many successors from many manufactures, which sometimes won't
tell you that their chips are 'compatible'.
</para>
<para>
To name a few: KS 0066, KS 0070, KS 0076, LC 7985, NT 3881,
SED 1278, ST 7066 ...
</para>
</sect2>
<sect2 id="ppt-misc">
<title>Miscellania</title>
<para>
This text has originally been taken from a <ulink url="http://lists.omnipotent.net/pipermail/lcdproc/2002-May/005832.html">message</ulink> by
Robin Adams <email>robin@adams-online.de</email>
</para>
<para>
Converted to docbook and slightly modified May 2002, Rene Wagner <email>reenoo@gmx.de</email>
</para>
</sect2>

</sect1>