It was my impression, that the LCD panel in a laptop, didn't have plug
and play information. Information flow for the raw panel, is
unidirectional,
and consists of some high speed differential signals (like TMDS or LVDS).
But, I suppose the "thing" driving the data cable to the panel,
could sense impedance. A typical value might be in the 100 ohm
range, and between the + and - data signals in a pair.
If you look at the examples here, I think I'm seeing perhaps Power,
Ground,
plus some number of high speed data pairs. The wires arranged in twisted
pairs are differential. One signal "goes up" when the other signal "goes
down".
Where the cable hits the panel, there will be a 100 ohm resistor, that
sits
across the pair of wires.
http://i.ebayimg.com/t/7-x-Common-L...-/00/$(KGrHqMOKjME1vHc4-iJBNcfUNPnqg~~_12.JPG
+----------------------> LCD Panel
| +-------> Red input (high
speed serial)
| |
motherboard Red+ ---------+-- 100 ohms --+--+
|
Red- ---------------------------+
Green+
Green-
Blue+
Blue-
Clock+
Clock-
So when the motherboard is connected to a panel, there is a 100 ohm
load across the signal pair.
On a regular video card, the GPU checks for 100 ohm loads on the DVI
signals. On VGA, it checks for 75 ohm loads on the Red to Ground, Green to
Ground, Blue to Ground.
When I want to "fool" a video card here, I make dummy loads. For VGA, I
buy
a blank VGA connector at the electronics store, then fit three 75 ohm 1/4
watt
resistors. When that dummy connector is plugged in, the video card thinks
a monitor is present. Using two video cards and three dummy connectors,
I've
simulated a four monitor desktop. While only owning a single real monitor.
So if I was in your situation, I'd make up a special cable, with the 100
ohm
loads on it. And give it a try.
At the motherboard connector, you would:
1) Connect Red+ to Red- with a 100 ohm, 1/4 watt resistor
2) Connect Green+ to Green- with a 100 ohm, 1/4 watt resistor
3) Connect Blue+ to Blue- with a 100 ohm, 1/4 watt resistor
4) Connect Clock+ to Clock- with a 100 ohm, 1/4 watt resistor
5) Make no connection to the potentially thicker power and ground wires.
You can see why I'm trying a 100 ohm resistor, in this Wikipedia picture
from the LVDS article.
http://upload.wikimedia.org/wikipedia/commons/b/bb/Basic_LVDS_circuit_operation.png
The wires in the LCD panel data cable, will be arranged for this (+) and
(-)
orientation. So even if you don't have a schematic with signal names on
it,
you identify things that look like "data pairs", then arrange your 100 ohm
resistor on each data pair. If a ribbon cable is used, a ground wire could
be
arranged between each data pair, for crosstalk spacing (to reduce
crosstalk
from one pair into another). The impedance of various cable assembly
methods won't be exactly the same, and the objective is to match the
impedance
of the cable. So if the ribbon cable had "110 ohm" stamped on it, and you
needed to terminate it, you could stick 110 ohms on each pair.
The hardest part, will be making the electrical connection. But I think
you probably guessed that, before even getting to this paragraph.
For $1.19, you can get enough resistors to try your hack. The lead length
should be kept short for best results. How I connected these, would depend
on whether I expected to ever connect another panel to the computer or
not.
If that was the case, I'd simply start shopping for the new LCD panel,
right now. And if the user data is what you're after, just pull the drive
and connect it to a desktop computer, using the appropriate adapter cable.
That's better than trying to make the laptop work again.
http://www.radioshack.com/product/index.jsp?productId=2062339
This is a *guess* on my part, that your problem is based on a chip
clever enough to do impedance sensing. I don't know for a fact yours works
that way. I've worked with LVDS devices in the past (dumb ones), that
don't care what you connect. But LSI (large scale integration) devices
with panel output drivers, would be good candidates for including sensing.
You'll notice that all the connectors on the faceplate of a video card,
they attempt to sense those. So it would be pretty easy, to include a
sensor
for an LCD panel in a laptop.
Years ago, some panels were driver by lower speed, massively parallel
TTL. Those would not be equipped for sensing.
And the older, standalone cheap LVDS driver chips, they wouldn't have
sense capability either.
But the GPU chips used in computers now, they have sensing.
I don't know if any other, smaller chips, have had that added or not.
But the fact you can't use the function key, to get the LCD panel to work,
that's all I can think of. Where the LCD panel cable goes, you somehow
need to terminate the high speed signal pairs. Then, retest.
*Note* - we could use the panel itself as a dummy load, but I'd want
to disconnect the "power" wire (the more positive potential, fatter
wire) so that any short circuits in the panel don't damage something.
With no power or ground connection, it may be sufficient to just
connect the data pairs of the panel to the motherboard, to get the
desired termination effect. So if I was on a desert island, I didn't
care if I ruined the flat panel cable, I'd try cutting the power wire
and leave the other connections intact. If you don't think there is
a danger of shorting, then just connect the panel back up again, using
the flat cable. Using separate resistors, and avoiding the panel
altogether,
is my way of arranging a "clean start" for the thing. So no additional
damage might happen. But if you're not equipped for electrical hijinks,
and are desperate enough, just connecting the panel back up,
might make your F5 key or whatever, work again. The external monitor
should be connected, before you press F5. The external monitor should
also be turned on.
Paul
