2004-09-03

Morning of the Zombies

A friend of mine asked me about old laptops the other day. It seems he had a project in mind. He wanted to create a digital picture frame. I knew I had some old ones in my junk box, but I didn't know if any of them were of any value. My friend managed to find just what he wanted on eBay, so it is good he didn't wait for me.

Well, after reading the work he has been doing on his smart-frame, and some of the pages that gave him inspiration, I was interested to see if I could follow suit. My friend had some extra ideas for his system already: he was thinking of video, and perhaps live updates from web sites, such as weather. I've had a few extra ideas myself, thinking about being able to hook up a microcontroller to feed data to the laptop. Things like temperature in the house or outside, or perhaps some sort of special input buttons, as no keyboard will be available when it is mounted.

So yesterday, my eldest was sick, and I wound up picking him up from school in the morning. So he was home recovering and I was home, undermotivated to log into work. So I unburied the junk laptops I had. I had 5 old laptops that were thrown out from a job I worked at in Denver. The same place I got the main body parts for the Frankenstein computer. 2 laptops looked older and smaller-screened. One was an AST an the other was a different name brand. The other 3 were basically the same model: ARM brand. I had previously checked the web for any documentation on ARM laptops, but came up with nothing useful about the systems. None of the laptops from Denver had a power supply. One of the ARM's even had a post-it saying dead screen. I already have another one of those: an IBM thinkpad, 486 era, that the LCD burned out.

I thought about using the Thinkpad's power supply to see if I could get any of the other computers to power up. But I was worried about jumpering the power backwards and shorting out the power system - like I had done 8 years ago with a portable VCR. But I gained two things since the last time I had played with these laptops. I had gained a Dell laptop power supply, that had been thrown out at the job I am currently at. And I thought up the idea of hooking the negative side of the power supply to the frame ground instead of guessing which point on the power plug was which input. That way I couldn't reverse the polarity.

At some point while I was setting up testing the power supply, my eldest wandered in. He seemed to be feeling a bit better. He showed a lot of interest in what I was doing. He decided he wanted to help. So I gave him the ARM laptop with the dead monitor label, that I had previously started to disassemble. He had a lot of fun opening it up and looking at parts. He kept asking about how to get off different parts. Eventually I got him one of my old Swiss army knives so he could do more work, or perhaps more damage :) . As I told him, they had already been thrown out once, so the worst that happens is they get thrown out again. And I figured if nothing else, some pieces might be useful for robotics. He had a huge amount of fun popping off all the keycaps from the keyboard. He was ready to try to turn it into a robot right away.

I took the Dell power supply and determined which pins had power. When I jumpered it to ARM laptops and the odd-brand laptop, I got no response. But when I applied the Dell power supply to the AST laptop, I got some green lights! So I rigged the wires a little more securely, and tried the power button. It beeped and whirred! I made noises like it was booting up! But the screen remained dark. I powered it off. Then I pulled over the P2-200's monitor I had lying on the floor nearby. I plugged the monitor into the VGA port of the AST laptop. I rejumpered the Dell power supply to the AST laptop. Green lights. I pressed the power button. The monitor's status light stayed orange for a moment, then flickered to yellow as I heard the monitor charge up. As the screen brightened, I saw that familiar Windows 95 boot screen. After adjusting the time for only about 5 months of lost time (it had been unpowered for over 2 years), I got to the desktop. It seemed to run fine. I rebooted, and found the status telling me it was a 486-DX50, with 16M of RAM. No very powerful, but I think it would be enough to run a lean version of Linux. But the problem was the LCD was burnt out.

Well, perhaps the LCD wasn't burnt out. Not the display itself that is. I suspect that the LCD on the old thinkpad is broken the same way as the AST. I explained to my son that there are two standard ways that LCDs are built. One type expects light to reflect off the front of the LCD screen. The other type is built for the light to come from behind the LCD. This is known as backlight, and almost every LCD for PC display uses this method. It is interesting that some gameboys are frontlit, and some are backlit. So they key factor in the laptop cases is the fact that the backlight might be burnt out, not that the LCD is non-fuctional.

But how to verify that the problem was the backlight and that the LCD display itself was fine? One possible way was to open up the LCD and see if the screen showed something when a different light source was placed behind the LCD screen. This was something I wasn't brave enough to try previously with the Thinkpad. But after reading the related websites from the other people making smart-frames, I was feeling braver. Also the fact that the laptops I was working on were already considered garbage gave me a little more bravery. But before I tackled the AST, I decided to try my hand on the ARM laptop that was already in pieces, and marked as a dead LCD.

I pried apart the screen panel. I unplugged the LCD and backlight connections. I carefully worked the back off the LCD itself. This mostly required straightening about a dozen little metal flanges that were angled about 30 degrees to hold on the back, along 3 edges. After the back came off, I could see all the different layers. The back was an electronic insulator, with some extra light blocking insulation. Then there was a layer of plastic, probably to transmit the backlight from the long, thin fluorescent bulb at the top, to light up evenly along the whole back of the LCD. There were polarizing sheets and the LCD glass itself. I didn't see anything that I didn't think I could handle.

So I slipped the ARM LCD back together, at least holding together, if not quite putting all the parts back. Then I more carefully approached the AST. It was about this point that my son announced he wanted one of the laptops for his own computer. While I might consider letting him use it a little bit, I think he will be very disappointed. It cannot play most of the games he likes, and it doesn't have a working battery. The AST doesn't even have a CD drive. So we will have to see what happens. I didn't even know if the LCD was any good.

Well, I slowly and carefully pulled the LCD out. There were more screws holding the LCD together than the ARM. I was surprised to see the same pink and white wires going to the backlight. And the LCD was also held by a bunch of familiar looking 30 degree flanges. It makes me think that both LCDs were from the same manufacturer. Well, I finally got the LCD apart. It had a similar set of layers of insulators and plastic. I jumpered the power back on again. I pressed the on button. Good lights and booting noises. I held up the LCD with the back off.

I didn't see anything. Maybe I was wrong about the backlight being the problem. But then I noticed the edge of something. It was text! It was a little hard to see, but it was there. Then I tried to show my son. At first he couldn't see it. But then he saw the Window 95 boot screen! I proved the LCD display was fine, it was just the backlight. I powered it off, and carefully reassembled the LCD panel.

So the question becomes, what is wrong with the back light, and how to fix it. I noticed that on both laptops there was this board along side the screen, with the pink and white wires plugged in at the top. I assume the backlight is some sort of fluorescent bulb, and the boards are the high voltage systems for the bulbs.

So then I wondered if the problem was the high voltage boards or the fluorescent lamp. After powering the AST back on, with the LCD ajar, I probed the high voltage board with my multi-meter. I found some interesting voltages on the lower half of the board. But the upper half of the board, near the pink and white wires showed no voltages. That suggests to me that the problem is in the high voltage boards. And I bet that is the same problem in all the laptops I have.

I was hoping to try a little swapping of the power boards and LCDs, but they all have different plugs. I might be able to jumper the power to the bottom of the boards, but it will be hard, tricky work.

I wish I could have gotten one of the ARM laptops working. Not only do the have much larger LCDs, they have CD drives and Pentium CPUs. (Yes, Pentium - not Pentium MMX not Pentium 2, just Pentium).

I think there may be some options on replacing a high voltage board, what with fluorescent tubes for case mods. But if nothing else, I might have the ability to use some other light source to backlight the AST's LCD if I decided to build it into a digital picture frame. I think I'd want to do something along those lines, at least temporarily, to test the display resolution on the LCD. If I can't get at least 640x480 with hopefully above 256 colors, it may not be worth making into a picture frame. But I suppose it might make a neat base of a Linux-based, wall mounted home information system. Especially if I hook it up to sensors and controls.

More hacking to come!

No comments: