Thread: How to find out LCD inverter pinout?

Firstly try to find the GND. It is usually the largest metal part on the PCB, e.g. that metal rim around the attaching hole is GND.
Next task is to find the power. Quite easy, becuse it is protected by a fuse on the PCB (small, smd fuse) marked as F1 on the PCB.
It is NOT neccessary to have only one gnd and power. Determine the twin wires by a multimeter with shortcircuit beeper function.
On newer backlight the power is usually works by 12V.

There is an enable line (ttl, so usually needs 5V) and a light control (give it 2.5V first, then you can increase or decrease). You can guess. Give the voltage to these remaining pins via a 100 Ohms resistor.

When there are no twin lines, try to determine, whether they are connected anywhere at all?

well, i tried to apply your suggestions, and i found GND quite easily , and power pin, which is number 1, connected via your said fuse. Also, it seems that two pins are additional ground, although there is an area on the pcb, that is for connection with notebook's metal case (also ground). so, i have 3 pins left.
You said about enable and light control lines. but do i really need to control these ones? let's say i put these 5V on enable pin, but do i need to regulate light control?

The control needs sometimes 0-5V, more rare 0-2.5V and 2.5-0 it decreases the brightness again. I have no idea about the third pin, perhaps 5V for logic.
Provide 4-4.5V for all, without resistors. Measure the current. Enable/control pins consume less than approx. 1-2mA, the logic - if any - over 5mA.
Try to search the central IC type with yahoo or google and check their voltage - could help...

You do not need to CONTROL them, but you have to SET them to an appropriate voltage.

damn, it seems it was too high voltage to supply to the inverter (12v), because now i have only half the screen shining . althouh i successfuly connected it to an old atx psu. gave 12v to the 1st pin and gnd and 3.3v to the enable and light control lines.
So now i have half-working lcd...

damn, it seems it was too high voltage to supply to the inverter (12v), because now i have only half the screen shining . althouh i successfuly connected it to an old atx psu. gave 12v to the 1st pin and gnd and 3.3v to the enable and light control lines.
So now i have half-working lcd...
and one more thing, although i have no good multimeter for measuring such high voltages like comming out from inverter, mine multimeter shows only 600v, but as i tried to measure it, numbers that i saw were much more than 1000v (of course, i can't asume these numbers were correct).

You can not measure it because the quite high (approx. 40kHz) frequency.
Your multimeter metal became a part of a capacitor and ...
When just a half of the lcd is enlighted it can be
1. there are two tubes and you drove just one (another pink/white silastic cord pair unconnected)
2. the voltage is too low (wrong type of backlight unit connected into the display)
3. try to apply higher voltage (5V) on the control lines, guess, which one is control and which one is enable. Use the third one, I describe below:

Now I am remember, what usually the third wire is: the half-brightness, used at powersaving modes in certain notebooks. Try what happens, when you apply GND instead of 5V into certain pins (there is only 8 possibilities in worst case...).

I think it is not a sharp line below the upper half that is not lit. It must be a gradually fading-type light, right?

yeah, i think i was wrong saying that something broke down. I found which is for light control and which is for enable by putting a potentiometer on the light control pin. So i will try later putting more voltage on control line.
As and you were saying, it must be insufficient voltage from inverter, because there is no other two wire pair to connect.
Yes, there is some fading down area, where it seems that the bulb is maybe of two segments? so now i need to try higher voltage to control line.
ah yes, one more interest: does lcd panel has to be white when no signal is given to data connector, or gray (say, these pixels are closed). or this depends on the model of lcd?
i found in lcd datasheet this: Display Operating Mode: Transmissive mode, normally white. so this should mean, kat its normal that display shines now not quite white, but rather light gray?

and what would i achieve by by connecting that third unknown pin to GND?

p.s. i saw your pics, and in one there is laptop keyboard, connected to ps/2 cable. can you describe shortly, how did you do that? used some controller?
sorry for off topic.

I am not sure, but it worth to try. It can be a type of input that is internally pulled up with a resistor and you need to ground for function or for switching between half or full brightness - I suppose.
Be careful, not to set the voltage higher than 5V for control lines!

It can be the power voltage higher than 12V eg. 18V. It depends on the laptop and unfortunately without the original notebook you can not guess. Generally it is the same voltage as the akkupack.

Responding previous post: normally white means WITHOUT driving (applying drive) just the backlight is visible. The LCD needs voltage for blackness! This fact came from the principle of working: polarized light goes trough the panel. The liquid crystals are random without electric field. That random arrangement does not polarise light again (perpendicular to the original direction of the polarisation). In presence of electric field, the liquid crystals are all set to a certain direction and that will polarise the already polarised light again, but because the two axes are in right angle, it will mean no light can enter (come out from that pixel, in other words that pixel is shaded).


That notebook keyboard comes out of a VERY old notebook (I think 486) and hopefully it had a separated controller.
I am also using a touchpad out of another notebook, that is PS2 compatible, if you are interested in.

But I am planning to link an old wireless keyboard electronics to a new HP Pavillion keyboard matrix. Ordinary notebook matrixes are not usable, because those flatribbon cables are unsolderable (no copper, just some disgousting aluminium-powder-like substance that refused any attempt of sodering...) That scrap HP matrix has nice REAL wires - a dream came true...