Another T22 Backlight Problem

[ezeb1473]

On my T22 the screen is dark. But on a CRT works fine. In a bright light I can see the screen is working.

However after reading what I could find on this problem I came across someone saying that the button that is pressed upon closing the screen might be stuck and not a problem with the invertor or lamps. I have pulled the motherboard out and upon visual inspection the button seems to move fine.

How do you rule this out as a problem and move on to the invertor/screen. Or is that simply it and I should now move on?
Tnx
Ed

[rkawakami]

If you've manage to pull the motherboard out and inspect the switch, then the only way I know of positively ruling it out is to test it with a multimeter. You will have to remove the I/O card that the switch is mounted on (I think the T22 is very similar to a T23 on this), flip the board over and locate the terminals for the switch. Depending upon the type of switch it is, you will have continuity (i.e., 0 ohms) across the switch terminals when the button is not pressed down OR when it IS pressed down. I don't have an I/O board handy to tell you which one it is. There could also be more than two terminals. If you are stumped on how to test it, send me a PM.

However, I would say the switch would only be the problem IF you have also turned off any of Windows' power saving modes that are related to "closing the lid". As it appears from your descriptions that all you are seeing is a lack of a backlight, then I would say that the most likely problem is a bad inverter board. Fairly easy and cheap to obtain on eBay, it can be accessed by simply removing the LCD bezel from around the display.

There is still a chance that the issue is with the backlight (CCFL) inside the LCD panel or with the ribbon cable that goes from the motherboard into the lid. If you replace the inverter with a known good unit and still don't get the CCFL to fire up, then my next step would be to replace the ribbon cable (only because it's cheaper than the LCD). If replacing the cable does not solve the problem, then there's a 95% chance you simply have a CCFL that doesn't want to light up.

Was the loss of light very sudden? In other words, it was working fine for quite a long time then all of a sudden it just went off? That usually is an inverter problem. Does moving the lid around sometimes "cure" the backlight problem for a short time? That's typical of a ribbon cable problem. Did your backlight have any pink or red tint to it when first powered on? That suggests an old CCFL that needs to be replaced.

(edit: Deleted section where I suggest that it may be possible to measure the output voltage of the inverter with a voltmeter. DON'T try it!)

[ezeb1473]

Actually it was working fine. There was some dust on the pastic that is located above the keyboard. So I took my finger and wiped the area going across the button and all at once the screen went dark. Exactly at the same time I brushed my finger over the button. There was no pink/ reddish and moving the lid does not cure the problem.
Thanks

[rkawakami]

Then all you may have is a loose connection to the inverter card. Remove the bezel, re-seat the inverter card and the connector and fire it up.

(edit: Re-read your last post... did you wipe along the top of the keyboard or the bottom of the LCD panel?)

[ezeb1473]

I swiped along the top of the keyboard from about the power button over to the right passing over the button .

[rkawakami]

While it's possible that you damaged the lid switch, I wouldn't think that it is that easy to break just by running your finger over it. I'm assuming you've checked the button sticking up out of the hinge cover and that it's not stuck in a down position and that it's free to move up and down.

The T23 I have with me here at work has a bad inverter card (my guess), but it does power up and down without any problems even when I hold the lid button down. The screen image DOES appear on the LCD (can see it with a bright light) so if the switch is indeed broken, you are seeing the same thing I am (no light on LCD, external video working).

If the lid switch is on the similar card to my T23s, the good news is that the I/O card is only about $20-$30 off of eBay. Search for 26P8023, 08K3373 or T20/T21/T22 I/O Card. If you go this route, be careful that you get one for the T20/21/22 and not the one for the T23.

[rkawakami]

Well, maybe the inverter in my T23 wasn't bad after all... I hooked up a Fluke 73 voltmeter to the output terminals of the inverter card, powered up the laptop and was greeted with a loud buzzing noise, followed by some "magic smoke" . I'm pretty sure that I had the connections secured enough and that nothing was shorting out, but during the buzz/smoke incident, the voltage meter was indicating an overrange condition (>600VAC). The plastic end of the clip I was using to grab onto the terminal inside the connector on the inverter card melted, possibly accounting for the smoke. I guess I don't know enough about the loading conditions allowed for the output of the inverter card, but this was an interesting experiment.

So for now, do NOT attempt to measure the output voltage on the inverter card by plugging a voltmeter across the pins. It's possible that the only way to correctly measure the voltage would be to have a lamp connected at the same time.

Time to call up my newest Thinkpad parts supplier and buy a couple of inverter cards...

edit: Should have searched for and read this article before my failed attempt at measuring the inverter:

http://www.ecnmag.com/article/CA602416.html

Application of a high starting, or strike, voltage —500 to 3000 Vrms — across the CCFL creates a momentary arc that turns the lamp on.

Do not use a handheld multimeter, though, because most such instruments lack the capability to make RMS measurements...

Quickly scanning the article seems to say that the best (and probably the only) way to test/measure the inverter is with an oscilloscope.