Handling Those LCDs
We find LCDs showing up in all sorts of applications, in
addition to the ubiquitous laptop computers and now the desktop
monitors and TVs. These LCDs, and smaller versions, have become
the de facto human interface for electronics, often replacing
indicator lamps and (with the inclusion of touch panels) switches.
The LCDs are seeing application in a range of industrial controls,
medical monitors and other instrumentation.
The big problem with LCDs is that controlling EMI is difficult
effective shielding tends to block the light as well.
So your choice is to pick shielding that passes some light
or try to contain EMI in spite of the big hole in the shield.
But any kind of opening makes for an antenna and also provides
a good discharge point for ESD.
There are no real good answers for this kind of problem, but
there are things that can be done. Lets take a look
at the situation and the possible alternatives.
Figure 1 illustrates the problem. The LCD panel is driven
by common mode currents on the display cable, by field coupling
internal to the enclosure, or by bit switching activity within
the display itself. This situation would be easily handled
by a shield, except that the options for shielding are not
very palatable, so we need to look for alternatives. After
all, the laptops dont use shields. How do they get by?
The answer is that the shield goes behind the display, not
in front. Then the emissions can only come up the cable or
originate in the LCD itself. If the common mode current on
the attached cable can be effectively blocked, either by cable
ferrites or effective circuit board grounding, the remaining
emissions must come from the LCD itself. This is then a matter
of minimizing loop areas by mounting the LCD as close to the
enclosure shield as possible. This assumes that the LCD comes
complete with a full metal back plate (most LCDs do), otherwise
you will need to provide a supplemental plate to close the
In practice, the easiest way to produce the described shield
is to ground the display frame to the enclosure around the
entire perimeter. Figure 2 shows this being done with an EMI
gasket, which is the most effective approach. If you are trying
to get by without gaskets, you might try grounding at various
points around the perimeter the better the ground continuity,
the better the performance. But you need a short fat ground
in each corner as a minimum, even with small displays. Larger
displays will need more contacts. Fortunately, most LCDs now
come with a full metal black frame which facilitates grounding.
The bad news is that the LCD may have the metal backplate
grounded to circuit ground. This works well for minimizing
common mode currents generated within the display, but can
make overall grounding difficult.
If you are dealing with a laptop or portable device, it may
be perfectly acceptable or even desirable to have circuit
board ground common with the enclosure. If you are dealing
with industrial controls, this may not be palatable. As a
minimum you will have to work to avoid ground loops. Grounding
the backplate through capacitors at intervals is a possibility
but is cumbersome to effect.
The above approach is usually enough to reduce the emissions
to an acceptable level for commercial and residential emission
requirements. More stringent requirements, such as military
or avionics applications, may well require a shield over the
LCD. This is usually accomplished with the addition of an
ITO (Indium-Tin-Oxide) screen bonded to the enclosure all
around the perimeter, as shown in Figure 3. ITO thin enough
to see through is not a particularly good conductor (about
50 ohms/square). We advise retaining the shield in back, blocking
everything but the emissions directly from the LCD itself.
If you need still more shielding, you will need to go to a
better screen. Silver coating is significantly better than
ITO, and much more expensive. The last resort is a conductive
screen it makes a good shield, but does degrade the
visual image, especially noticeable for high resolution images,
such as would be found in a computer screen.
Whichever shield you use, you need to make sure the screen
shield is grounded to the enclosure around the entire perimeter.
Anything less is almost sure to fail.
Unfortunately, the LCD is quite exposed to electrostatic
discharge, especially if it is equipped with a touch screen.
The discharge goes not to the screen but typically to the
enclosure at the edge of the screen, finding a small nick
to arc to. In most cases the effect is not destructive
the discharge usually results in nonsense symbols on the screen,
which remain there until refreshed by a data update.
If you are figuring on using an ITO shield or similar, grounded
as described above, you should be in good shape for ESD as
well. But if you dont want to shield, you must not compromise
on grounding the LCD backplate. This solution is far from
being bulletproof, however.
Of course, if your data is continually being refreshed, such
as with a TV or computer screen, there is no problem with
nonsense characters, as they will be there no longer than
one refresh cycle.
You can try to solve the problem in software. The idea is
to refresh the screen on a periodic basis once a second
is adequate. If the person discharges to the screen and the
contents is scrambled for a second, it is not a problem. LCDs
typically require a significant initialization upload and,
unfortunately, this will likely be needed each time you refresh
the screen. It is not a big issue you have already
written the initialization routine. You just need to have
a routine to call on it periodically to reset the display,
thus keeping the screen contents the same.
If you are working with an LCD, you might as well save yourself
some grief and figure out how you are going to handle the
big leak. It will be there, make no mistake. You can follow
some basic rules and minimize the risk.
Regardless of the extent of the emissions problem, start by
grounding the LCD to the enclosure around the entire perimeter,
making sure that the LCD has a metal backplate. Gasketing
is preferred, but multiple ground points may be adequate.
Suppressing common mode currents from the cable is important.
If your measures are inadequate, you may have to go to an
ITO window, or even a screen.
ESD problems may respond well to software suppression techniques.