Slashdot Mirror
Extending LCD Display Life?
polymath69 asks: "I use a laptop as my primary home machine, and wish the display to last as long as possible. There are two main camps of opinion on how one ought to ensure this, each grounded on a seemingly logical point. Opinion One goes like this: An LCD's backlight is only going to emit photons for some number of hours, therefore shut it off when not in use to maximize life. Opinion Two counters: A backlight can only be turned on some number of times, therefore leave it on to maximize life. The conclusions in each camp are diametrically opposed. So what is the truth? And how do you make a choice when two seemingly irrefutable arguments are in disagreement? Both these truths start from the same given: that eventually, the LCD will fail. But looking at that given from two points of view leads to contradictory answers. Now, one of these arguments has got to prevail. But which is it, and why?"
If this is a real concern to you, you might want to look into a few other options.
One, is that many LCD brands, if they use flourescent Backlighting (like most Laptops do), often times the flourescent tube can be replaced -- sometimes with ease, sometime with difficulty. Examples of ease are the Mac PowerBook 1xx series, where a new tube is only a few bucks and can be replace in a a a few minutes. Examples of difficultly are my Fujitsu Lifebook, where the entire LCD panel itself had to be taken apart.
The other option you have is, depending on how old you machine is and how much time and money you want to put into it, is to just replace the screen when it becomes a problem. If it's an 'older' (i.e. 1 year old) machine, you can pick up a new-from-factory LCD for 300 USD, or your can watch eBay for a model like yours with a dead logic board or something.
Anyway, if it's your main machine, then I have a feeling that other parts of the box will go before the backlight does -- especially the battery and the hinge.
Hilary Rosen's speech was about her love of money and her desire to roll around naked in a pile of money.
a. n. A person of much or varied learning; one acquainted with various subjects of study.
b. adj. Very learned.
(OED)
11.0010010000111111011010101000100010000101101000
IBM THinkpad 600X, three mouse buttons, works great with SuSE 7.0, sound supported via ALSA or kernel 2.2.18's OSS/Lite. Lucent driver from http://walbran.org/sean/linux/stodolsk/ works with 2.2.x and 2.4.x kernels as long as you read the 1ST.README file and follow the directions. Main problem with the 600X is that the total RAM is not recognized on boot and it's not quite the N meg you think you have, but (N-0.5) M because of the "EZ-BIOS" or whatever.
Give a monkey a brain and he'll swear he's the center of the universe.
Well, really, I did try and simplify the solution, since, as posed, the solution would be underdetermined. I need more data to determine the optimal solution: the cost of the system, with a depreciation schedule, the specific MTTF and MTBF of the system, backlight replacement costs, and time, with projections into the future, including contingency plans if any of the constituent parts reaches end-of-life in the service interval. Also, I think you would have to factor in the cost of the downtime of the system with a respect to lost productivity. You could recoup some of the cost by dedicating the laptop with a dead backlight to a home server, or possibly as a tax deduction, but that is a question better posed to your accountant.
I think that this quickly changes from a failure analysis problem to an accounting problem, which takes it far out of my range of experience, which is primarily engineering.
With one or two of the lifetime lengthening measures, and without expending the effort on the PDE solution to the problem, you can probably easily extend the lifetime way past the expected lifetime of the laptop, or at least past the "crap, I spilled Coke all over my keyboard" event.
I think your laptop's MB or something else will fail long before your LCD. The thing you would probably have to worry about the most is physical damage. I, personally have never seen a LCD fail, but that doesn't mean they don't. If your that worried about it, either get a LCD flat panel to plug into when your at home, or a 15 in monitor to plug into. For that matter, you can extend the keyboard and mousing device (trackpoint, touchpad or whatever) by throwing in a cheapy mouse too. Oh, and Cmdrtaco will wonder this too, but how come there are NO laptops with three mouse buttons???? Anyone who spends a lot of time under Linux or UNIX (FreeBSD or anything running X windows) will come to appreciate the extra button (no Apple, it's not that hard to figure out a lousy three buttons).
Gorkman
Ok I just checked it and I see it has been updated....finally! Looks like some resourceful hackers hacked on the original to make it work similarly to the way the aureal hackers did. Actually a coworker of mine has this same laptop (I think) and he's running lilo dual boot with slack ware on the other partition. He needs to get a new HD cuz he has barely 1.2 gig of space and that's not hardly enough to seriously play.
You are definitely right about the hardware though. I see most laptops getting replaced before the LCD burns out.
Gorkman
Judging by your nickname, polymath69, you are probably mathematical in nature. The problems that you describe, brightness dimming over time Vs. power cycling breaks up very nicely into a linear separable partial differential equation. I know this because I've solved a similar problem for a harddrive (spin up/down Vs. spinning time). I'm fairly certain that the solution to the PDE is analytic and you will be able to find an exact solution, yielding an optimum amount of time to leave it on as opposed to turning it off. If it doesn't have an analytic solution, then you can use the linear simplex algorithm or find a solution using Galerkin's method. Each of these numerical techniques are fun and will converge to an optimal solution relatively quickly.
For further info, I recommend MacCluer's industrial mathematics book.
Keeping
I've been using laptops and other portable computing devices for years.
Generally there are a few ways that an LCD will die. None of them have anything to do with the backlight. The backlights on my LCD displays have never failed. My earliest palmtop's case broke so it wouldn't stay open. Not an LCD failure. My next palmtop got smooshed and cracked the screen. My latest palmtop has a nick-mark on the screen because I accidentally dropped something on it. My first laptop had a depression in the screen becaue some @$^$ kicked it, but it was still usable. It finally died when the motherboard stopped working. Also not a LCD failure.
Don't worry about your backlight. Wory about everything else breaking.
Gentoo Sucks
You have another concern, as detailed in this PDF. Backlight lifetime is going to go down as you increase the brightness. From a brief google search, backlight lifetimes run from 20,000 to 50,000 hours, with intensity decreasing with time, and end-of-life determined as 50% brightness of new. You may declare it too dim far before that time, but even then 20,000 hours is 2500 8 hour days, or almost seven years at 8 hours a day.
.5 to 1 hour of non-use of the system.
That gives you some options:
1. Don't run the backlight at 100% intensity- try to reduce the lighting in the areas where you are using the laptop. This saves batteries too.
2. Turn off the backlight after something like
3. If you must leave the monitor on to be able to check things at a glance, reduce the intensity as much as possible.
Also, just for the general lifetime of your laptop, use the power-down/suspend power saving features- laptops are dense little packages of electronics- which generate a lot of heat, but don't dissipate it well. Heat kills electronics and batteries, try to keep the heat down, and you will greatly improve the lifetime of the whole laptop, not just the screen. If you're not using it, and the laptop is warm, you're wasting power and laptop lifetime.