Showing posts with label Experimentation. Show all posts
Showing posts with label Experimentation. Show all posts

Saturday, November 14, 2009

Neck Ache: Part 4: Ugh

Right-o lets get this over with. 

Took this super tiny piece of wire and soldered the two broken ends together. 

I know it's not the prettiest thing in the world, but the three bits are securely attached to each other. A little electrical tape and I should be in business... Should....

So at this point, I plug everything in to test it. It doesn't work of course. Here's what happened: The backlight turns on for a second and then turns off again; the status light is amber. I wiggle the video wires a bit and get the backlights to stay on and on the screen I get black static. Wiggle the wires a little more and the black static turns a little bit into the image of what it should be showing but with lots of static and the status light turns green. A wiggle later, the backlights are off again and there's not even a black staticy image left on the screen. However, the status light is still green. 

If you click on the image, you'll get a bigger version. Notice the LED is still green and there's nothing on the screen...
Now in the process of doing all this, I removed the neck from the LCD back so I could get to things easier while I was troubleshooting. Unfortunately, I couldn't think of anything else to do. I went back and multi-metered all the video wires and everything connected right were it was supposed to be and there were no exposed wires touching themselves or anything like that. 

I'm really out of ideas. It doesn't seem to be a problem with the backlights because they were functioning perfectly when they were on. The LCD itself seems to be doing well: when it did show a staticy image of my computer, besides being staticy and stretched, the colors were all accurate and things seemed to be in the right place... I don't know, but that's enough for tonight anyway... 









Friday, November 13, 2009

Neck Ache: Part 3

With the bottom cable all soldered up It's time to move on to the top half. Instead of soldering the top half, I gambled and decided to just replace the pins. The gamble was that even though they're the same size, the pins on the iMac's cable might pull out easily if the little plastic keepers don't go all the way back down.

So I just grabbed my handy x-acto knife and carefully pried the little keeper bits up the smallest amount possible to release the wires. Once I had them all out, I consulted my pinout (see bottom) and started plugging the iMac's wires in. Because of that pesky extra green wire, I stole the green wire from above and soldered it to the green cat 5 wire I ran so it would fit into the adapter.

It actually took me a really long time to get all those tiny little wires pushed into the plastic thingy. But I eventually got it.

Like i feared, many of these wires aren't really in there all that well and tend to want to slip out with relative ease. I'm not 100% sure how to remedy this issue. I thought about hot-gluing where the wires go in but I'm paranoid about melting any insolation or screwing anything up like that. For the time being, I just flattened out the wires and wrapped them in electrical tape.

Once the video cable was all finished, I moved on to the backlight wires. For these, I'm using the cables Apple allocated for the Inverter to run the two pairs of backlight wires. I started by cutting back the heat-shrink on the backlight power cables.

Incase you haven't noticed, I have a lot of... lets say, "special" moments. I get it into my head that I know what I'm doing and got it all figured out, and then, do something really stupid. Like cut off the the end of that video cable! grrr... Anyway, in this installment of "Dave's Super Special Moments", (most likely brought to you by leaded solder), instead of cutting off end so I have the connector part to solder onto the bottom, I just pulled off the plastic part as I've been doing and soldered the the plug ends right to the inverter cable, or rather, the cable that plugs into the inverter cable:

This is the cable that connected the original inverter to the inverter cable right behind the display.
Anyway, so that was dumb. Luckily, since I just happened to have some spare monitor parts lying around (weird, right?) so I was able to solder on the end from that:

So, that's it. Everything was ready for the big test. As of that moment, I was 50/50 with this monitor stuff working. After I got everything hooked up and turned the monitor on... it didn't work. Now at this point a familiar feeling of dread passed over me, but I was able to shake it off and go right into troubleshooting mode.

Symptoms: The backlights turned on for a second, then turned back off. The picture on the screen was present even after the backlights turned off and was very fuzzy.

Testing: I grabbed my multi-meter and began testing to make sure all the wires were still connected and no exposed wire bits were hitting any other wires. I did the continuity test on the inverter wires first, and they all checked out. I then moved onto the video cable.

I made sure each of the wires was where is was supposed to be based on my pinout. All good. Then I did a continuity test from the black connector that plugs into the video card to the plastic thingy (above) that plugs into the back of the LCD, making sure that there were no broken solder points. I also wanted to make sure there were no exposed wires touching other wires, so while I had one end of the multimeter on a wire in the black connector, I slid the other lead of the multimeter across all of the gold ends on the plastic thingy and made sure it only beeped when touching the right pin. A lengthy explanation for a lengthy process.

Anyway, after doing all this, I found that a single wire had managed to go and get broken.

That's right, that little bastard, RIGHT THERE, ruined my entire day! To add insult to injury, it was one of those hair-thin stupid wires that was a pain the butt to strip and solder the first time when it wasn't surrounded by a heaping glom of un-shrumk heat-shrink tubing. Yuck.

At this point I grabbed some cookie dough out of the refrigerator and called it a day. I REALY hope that is the culprit and I'll have the monitor fully functional by the end of the weekend. We'll see.

Oh, here's that pinout I promised:

I know it's messy, but it I understand it and that's all that really counts.

Monday, November 9, 2009

Backlight Extension.

These backlights are weird. There's just not much else I have to say about them. But let's get to that in a minute. So as I've discussed before, there seems to be some question as to whether or not lengthening the cables from the inverters to the backlights will cause any dimming of the lights. Well here's what I've managed to come up with:


Can you tell which is brighter? They look the same to me. One of these has a 3 foot cable from the lamp to the inverter, the other's is about 4 inches. And I gotta say, They look pretty identical to me. So, it would appear that lengthening the cable from the backlight to the inverter (at least a short distance) should not affect the brightness of the lamps. Now to the weird bits.

When both lamps were attached via the 3 foot cable, and that cable was coiled up, weird things stated happening. When I touched the aluminum housing on one of the lights, (grounding them I suspect) it got a lot brighter and started humming at some high pitched frequency.

When I touched the aluminum housing of the other one, half of it went out. (It didn't matter where I touched it, it went out in the same place.)

Very interesting. The cable's definitely didn't like to be coiled. When I stretched them all the way out, they both lit up just fine and didn't appear to be affected by being touched.

Note: All 4 of these pictures were taken with exactly the same exposure settings. Their relative brightnesses should be quite accurate. 

Oh, and I really didn't feel up to soldering today. Probably tomorrow though. 

Saturday, October 31, 2009

When Life Gives You Lemons: LCD Edition

The only benefit I can see that came from my little upset yesterday is that I am now able to do a little experimenting. I have first decided to test the backlight dimming theory. The ideas is, as stated a few days ago, that if you lengthen the wires from the power inverted to the LCD backlights, the extra resistance will cause the light to be dimmer. 

So using some extra cat5 cable I had, I soldered in an extension just short of 3 feet. My results: Hard to tell. In the process of doing something, I managed to kill one of the backlights. However, the remaining backlight was still producing a decent quantity of light. It's difficult to judge if it was dimmer or by how much, but it definitely seemed close. So that remains a viable option anyway.

There has also been another development. For the iMac, Apple designed a special, standalone inverter which sits right next to the display. It was suggested that if I could figure out how to use that one instead of the built-in one on the new LCD's power supply, I wouldn't have to worry about cable length. It turns out that might just work.

As seen in the picture below. The iMac's special inverter has a 6 wire "power cable" coming out of it. On the new LCD's power supply, where all the rest of the inverter circuitry seems to be located, there's that big black square thingy which has, at the front of it, 6 beefy solder connections to the board. My theory is, if I pulled that piece out and soldered in six wires in its place, I could run that from the power supply to the iMac's special inverter be home free.



The only problem is, I have no idea which way to orient the wires. Even if you make the giant assumption the wires are in the same order, if I hook it up backwards (the red wire at the top when it's suppose to be at the bottom), I'd probably fry the inverter if not the entire power supply. I'll soon have a backup power supply, but only one inverter...

By the way, here's the code printed on the top of the black square thingy. I googled in but didn't find anything. It's unclear whether the 0 is a zero or the letter. I tried both but neither came up with anything.

SPW-095
LSE r15Ch

If only I had won the lottery, I could just buy up a pickup load of LCDs and iMacs and not have to worry too much about the whole screw up thing. if only...