Maker Pro
Maker Pro

Troubleshooting LCD (alphanum) display problem

D

Don Y

Jan 1, 1970
0
Hi,

I have a NAVITEK LAN/cable tester (for want of a better name).
<http://www.tequipment.net/Ideal33-846.asp>

It includes a two line alphanumeric (plus icons) display to
present prompts and results to the user.

The display has developed faults -- similar to "dropped pixels".
I'm trying to figure out the best way to sort out where the
problem might lie -- in the display connection *or* in the
(serial) interface to the display PCB.

And, of course, ideas as to how it might be repaired. While
the device wasn't inexpensive ($500), it's long obsolete so
probably not worth sending it in for service (I will contact
the manufacturer just to see how outrageous their service
prices are :> ).

OTOH, it's *really* handy when I need it.

I.e., fixed it would be wonderful. Even "as is", it has some
value (though it requires patience to determine what the display
is saying as the prompts cycle by).

I suspect the first step would be to see if the individual
"displays" (prompts?) are consistent each time they appear
or if they exhibit some variation.

Then, determine if there are any patterns in the faults.

Finally, mechanically stress the device to see if these patterns
change.

Any other things I might want to try or observe?

Thx,
--don
 
D

Don Y

Jan 1, 1970
0
Yes. Make sure it has fresh batteries.
I know.... I'm sure you checked. But how many times.....

Opened a brand new lithium 9V just before my post
(after seeing the problem with the *installed* battery,
previously)
Could it be the LCD display itself?

That is my current thinking. But, I will have to drag out
a movie camera to capture the various displays (it cycles
through various displays every second or two as it runs
its tests -- no way to "pause" it anywhere but at the start
of the test sequence). Then, I can look to see if certain
pixels are missing in *all* characters in a particular
display position -- or, just a particular character
(i.e., the font may be defective, assuming pixels are
being transferred to the display instead of characters)

There doesn't *seem* to be any change in any *particular*
"display". I just haven't been able to correlate those
deficiencies with *other* "displays"
If it's just a character LCD, you may find one off-the-shelf to
replace it, and should be relatively inexpensive, especially if it is
the 14 or 16-pin parallel variety, and not the more expensive serial
types.

The "display board" interfaces to the "processor board" via an
8-pin connector fabricated similar to the way LCD glass mates to
an elastomeric connector (i.e., just "touching" surfaces, not
"engaging" pins)

[this complicates troubleshooting as opening the package
causes the display board to lose its connection to the
main board -- I'd have to find some suitable vias and
attach pigtails]

So, I suspect the display interface is serial -- it doesn't make
sense to interpose a serializer to a parallel display.

The display has icons as well as the graphic area for the 2 lines
of text so I imagine this might complicate locating a COTS replacement.
Do a little research first to make sure the line addressing (cursor
placement) is the same between the broken and replacement LCD
display. I've seen LCD's that assign weird starting addresses for the
cursor positions, which vary by line, and sometimes, in the middle of
a line.

If the LCD is a graphic type, my "guess" is the display is bad, since
a few bad pixels here and there should cause "that much" readability
issues for you.

If this is indeed the case, then being able to objectively
review images of the various "screens"/displays should allow
me to verify those pixels are, indeed, dead. Given that
the missing pixels don't seem to change (e.g., when flexing
the device), any other result would suggest a defect in the
data being sent *to* the display (i.e., font defined in rom)
 
J

Joerg

Jan 1, 1970
0
Don said:
Hi,

I have a NAVITEK LAN/cable tester (for want of a better name).
<http://www.tequipment.net/Ideal33-846.asp>

It includes a two line alphanumeric (plus icons) display to
present prompts and results to the user.

The display has developed faults -- similar to "dropped pixels".
I'm trying to figure out the best way to sort out where the
problem might lie -- in the display connection *or* in the
(serial) interface to the display PCB.

And, of course, ideas as to how it might be repaired. While
the device wasn't inexpensive ($500), it's long obsolete so
probably not worth sending it in for service (I will contact
the manufacturer just to see how outrageous their service
prices are :> ).

OTOH, it's *really* handy when I need it.

I.e., fixed it would be wonderful. Even "as is", it has some
value (though it requires patience to determine what the display
is saying as the prompts cycle by).

I suspect the first step would be to see if the individual
"displays" (prompts?) are consistent each time they appear
or if they exhibit some variation.

Then, determine if there are any patterns in the faults.

Finally, mechanically stress the device to see if these patterns
change.

Any other things I might want to try or observe?

Yes: Open it up and see if the LCD contacts the circuit board or a
sub-module via the usual two rubber strips. They look like thin
mechnical buffers but in reality they contain contact fibers. Over time
this can dry out a bit and then some segments lose contact between LCD
glass and the board below. This is also true in most stand-alone LCD
modules except that you often can't see the rubbers because they are
behind a thin aluminum frame. Such conducting rubber strips can be bought.

A typical sign is that when you have the unit apart and then gently
press on the glass sides where the strips are the fault pattern changes.
But not always.

Of course there's always a chance that something else happened, like on
one of our large atomic wall clocks: One fine days the leading "1" had
vanished. So it would show 1:00 instead of 11:00. Opened it up but
turned out the driver was shot. Since that was under a blob of tar that
was the end of the clock.
 
Top