Searching \ for '[OT] LCDs and ESD' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/lcds.htm?key=lcd
Search entire site for: 'LCDs and ESD'.

Exact match. Not showing close matches.
PICList Thread
'[OT] LCDs and ESD'
2000\01\10@155346 by Eric C. Thompson

picon face
I'm hoping that someone out there has dealt with this problem and can help.
(Please)

Background info:
I am currently working on a project that includes an LCD, it is a 2X24 LCD
made by Seiko with LED backlighting.  The LCD is connected to a
microcontroller (not a PIC?, oops) and the micro is also connected to an
RS-485 line.  The entire project is housed in a metal case with a clear
piece of plastic in front of the LCD.  This project also includes 5 push
buttons that are directly under the LCD and come out through the front of
the case.  The LCD and our PCB are mounting to the case with 4 screws, the
holes in our PCB are insulated from any signals on the circuit.

The problem:
There seems to be a major problem with the LCD and ESD.  Every so often when
one of use goes to hit a push button we shock the box and cause the LCD to
scramble all it's characters.  I have concluded that the microcontroller is
running OK (although it may have reset?) but the only way to get the LCD to
work again is to cycle the power.

Things we have tried: (and did not make a difference)
+ Grounding our metal case
+ Grounding the metal case that is part of the LCD board

So are these LCD that sensitive to ESD?  It seems like everyone uses them
for projects, I can't image everyone has this many problems with them?   Is
the only answer to switch our case to plastic?  We have other projects
around here that use all plastic cases and LCDs and they do not seem to have
this problem.

Thanks,
Eric Thompson

2000\01\10@161113 by Peter Keller

flavicon
face
That's a known problem. I had it also.
Connect the metallic frame which holds the glas of the lcd to ground.
I used coppertape with adhesive (3M-Company). Any other connections may help
also.
Also connect the case to ground.
Peter Keller

"Eric C. Thompson" schrieb:

{Quote hidden}

2000\01\10@161942 by Wagner Lipnharski

flavicon
face
Question:
What do you mean by "...push button we >> shock << the box... "?
I've being working with LCDs for years without any ESD problems (at the
units), not even had a situation where the LCD got locked up needing a
power cycle to return operational...  I would suggest to replace the LCD
by another manufacturer (Try Optrex for example), and check all your
pulses on the LCD interface, mainly related to ground and voltage, check
decoupling capacitors, pulses timmings and wave form integrity.
Good Luck.
Wagner Lipnharski




"Eric C. Thompson" wrote:
{Quote hidden}

2000\01\10@174349 by James Paul

flavicon
face
 Eric,

 Check to make sure you're not getting some pin bounce from
 the 'Shock' to the system.  Use pullups on the data and control
 pins if necessary to keep them in a known state.   I don't know
 if this will help, but it is something to think about.  I did
 have some pin bounce on a project I had and this is what I
 did to fix it.  Worked fine after the pullups were added.
 Give it a shot.  Let me know if it helps you too.

                                       Regards,

                                         Jim




On Mon, 10 January 2000, "Eric C. Thompson" wrote:

{Quote hidden}

spam_OUTjimTakeThisOuTspamjpes.com

2000\01\10@175519 by John De Villiers

flavicon
face
Ever consider that youre the one thats staticly charged?

Try touching something thats grounded before you press a button. Something
close by, so you dont have to walk to get to the LCD.

I've found then when the air is dry enough and you have the right (wrong)
kinda carpets you become nicely charged. I've seen plenty switchboard
operators blowing the console up when coming back from a tea break.  Most of
them wear static discharge wristbands now.

> I'm hoping that someone out there has dealt with this problem and
> can help.
> (Please)
>

2000\01\10@195219 by Mark Walsh

flavicon
face
{Quote hidden}

We have had the same problem with LCD's from a variety of mfg's.  Once the
display memory gets screwed up (usually the displayed characters are in the
wrong position) you have to reset it.  We can't cycle power in our
application, so I just go through a software reset (initialization) and it
seems to clean up the mess.

Mark Walsh

2000\01\10@211020 by Brent Brown

picon face
Lot's of good suggestions for you already, but I'll add my 2 cents
as well.

The exact same problem you describe showed up in one of my PIC
designs when it went through ESD tests for CE compliance
testing. Fortunately for me the software both re-initialised and
power cycled the LCD routinely. The question is not so much "will
it fail?" but more like "how much will it take before it falls over?",
and "will it get back up again by itself?"

If by chance you're using a '51 micro then adding strong external
pullup resistors (eg 2k2) to the LCD control lines may make it less
sensitive.

Ask yourself if the ESD is getting in anywhere in particular, like
straight through your switch and into port pins or power/GND rails.
In this case series resistors on switch lines may reduce peak
currents from ESD zaps enough to make a diference.

In another example I discovered a problem with a membrane
keypad and LCD mounted in the dashboard of a coach. The panel
it was stuck to was earthed well, every possible shielding trick was
tried, pullup resistors on scan lines, series resistors in scan lines,
capacitors everywhere, keypads and cables replaced, but erratic
key presses persisted giving garbage on the display.

As a last resort I dived into some old "don't touch it - it's working"
software and found that the key debounce routine was to blame.
Although true key "bounce" was ignored it assumed that the very
first switch closure was a valid press without checking it for a
period of time. When this was changed all the problems went away.

Brent Brown
Electronic Design Solutions
16 English Street
Hamilton, New Zealand
Ph/fax: +64 7 849 0069
Mobile: 025 334 069
eMail:  .....brent.brownKILLspamspam@spam@clear.net.nz

2000\01\10@235559 by quozl

flavicon
face
Some LCD panels require so little power that you could drive them from
the PIC pins directly, so that you could force a power-up reset on them
under program control.

--
James Cameron   quozlspamKILLspamus.netrek.org   http://quozl.us.netrek.org/

2000\01\11@021803 by Graeme Zimmer

flavicon
face
> That's a known problem. I had it also.
> Connect the metallic frame which holds the glas of the lcd to ground.

Also, make sure that the contrast line is bypassed with a small electro.....

Any noise on that line (even LCD induced noise!) will scramble the LCD
ocasionally.

It took me ages to find that one !

regards ............... Zim

2000\01\12@101535 by M. Adam Davis

flavicon
face
If you can spare the processing time and ram, it would not be too
difficult to emulate the LCD memory on the PIC, and redraw the lcd a few
times a second.  While this does not fix the source of the problem, you
know that the display will always be self-correcting if any other
problems crop up.

-Adam

"Eric C. Thompson" wrote:
{Quote hidden}

2000\01\12@112158 by Peter Keller

flavicon
face
Probably you have to re-initialize the lcd-module after the text is corrupted !
So rewritung does not help.
Peter

"M. Adam Davis" schrieb:

{Quote hidden}

2000\01\12@115307 by Douglas Fast

flavicon
face
Hello,

Make sure you check the LCD power up requirements carefully as well as the
LCD initialization sequence as given in the datasheet.

I have used an Optrex 16x2 LCD display (Hitachi controller) extensively
(connected to a PIC) and I have seen what you are describing numerous times.
The solution in our case was to be very careful with the power on sequencing
and rise times, as well as the LCD initialization code.  For example,  the
datasheet for the part we were using indicated that the power on reset
circuit (in the Hitachi LCD controller) would not function properly if the
supply rise time was not 0.1ms < rise time < 10 ms. To meet that spec we
added a 10 ohm resistor and a 10 uF cap to the power supply line.  In
addition, the software initialization code rigorously followed the timing
sequence given in the data sheet.

Our problem was solved ONLY when we implement both of those items.  Now we
can reset a running display without difficulty.

Regards,

Doug



{Original Message removed}

2000\01\12@115931 by Wagner Lipnharski

flavicon
face
Some more details:

Most LCD's power consume is very low, what means high impedance
circuits, thus more ESD sensitive, also Power Supply noise sensitive.
They are not *that* sensitive for Power Supply voltage deviation, but
for fast voltage change.  I have battery operated LCD's (designed to
work around 5Vdc) working from 3Vdc, display contrast went blank but the
LCD still working.

Doing some experiments, I was able to generate garbage at the display if
changing power supply fast from 5 to 4Vdc at the same time commands were
being sent to the unit, but nothing happened at other times. So, appears
that during a command sequence it is more sensitive to voltage
variation.

Some nasty display scramble was observed if the Busy bit was not
checked, and commands were sent before the right time, during the time
the unit is busy.  In those situations where the busy bit is not
checked, increasing the loop time after Clear command or Address Set
solved the problem.  If you send commands while the unit is still busy,
it will ignore the commands, but if in the middle of it the unit turns
"not busy", it will decode a complete different command, and it can
result in crazy things.

In your application, the LCD is probably updated right after the user
press the buttons at the panel, so to check it out, change the software
to wait 5 seconds after a button is pressed to update the LCD, so you
will be able to understand if the display scramble is caused by ESD
(user pressing a button), or by software (or other).

Wagner.

More... (looser matching)
- Last day of these posts
- In 2000 , 2001 only
- Today
- New search...