Slashdot Mirror
Recycling Parts From Dead Motherboards
An anonymous reader writes "I had this dead motherboard on my hands and I wanted to see what would happen if I cut out the clock generator and used it stand-alone. So I removed the Winbond chip from the motherboard (I cut out the section of PCB with a hacksaw), powered it up and it was still working. Add a display, a microcontroller and two switches, and I got a cheap frequency generator. Here's my progress so far. Be kind to my Web skills, I'm really just a hardware monkey.
It's not completed yet, but I just wanted to get the idea out there."
Dear Michael's a Jerk,
I for one (and i don't think i'm alone) do not think this is "ool, but a Waste of Time".
repurposing parts from old motherboards to make new test equipment IS cool and IS NOT a waste of time. Just because you can't produce a thousand, or even two, doesn't make it not worthwhile; it's silly to think that Tesla or Turing or whoever should never have made anything, because they could only make one.
and just because something *might* be damaged DOES NOT mean it isn't worth a little hacking!
I'm certain that many slashdotters have gotten tons of use out of "broken" and "useless" throwaway parts from old machines. I know I have.
It's tiring to see every cool hack posted on slashdot be berated by people who don't think it's worthwhile. That attitude has nothing to do with the experimental mindset of hacking, and does nothing to construct anything new. This person did something new *and* shared the knowledge with us! Many, many inventions have come from tireless "frustration".
As for the old motherboard for a source of parts, I keep a couple of big boxes full of motherboards and adapters for salvaging parts. Even though I'm at a point where I can get free samples of nearly anything I want, there's nothing like having the part you need when you need it.
"Eve of Destruction", it's not just for old hippies anymore...
Yes this is really neat. It's a great project. You will rarely save money with this approach, but it's no reason not to try it.
The most important reason is that you are learning to use the parts by example which is really cool. You get the benefit of the hard work of the designers and testers. When you start from scratch with a new part, even with all the specs and theory it sometimes takes a few tries to get it right.
I spend as much time as I can building stuff out of junk because it is what I love. Over the years I've figured out that some cool stuff isn't worth the salvage labor. You can get it another way and it will work better, especially when it's a newer surface-mount, multi-layer board. You really have to weigh the alternatives carefully.
However, you definitely do well when you find boards with parts in sockets and things like that. Old ISA cards and very old motherboards are a great source of unpluggable parts. Most of them have serial eeproms like 9346's, you can get 8051 and 6811 microcontrollers off old modems just by popping them out, UV eproms and eeproms to make your NIC bootable, and if you're lucky you can find an ANCIENT card covered in sockets full of 74xx logic chips of all kinds.
Sadly, the newer things are the less you can do with them. Newer toys, electronics, and computers are becoming so cheap and highly integrated that it's getting really hard to do anything interesting with them. The speak'n'spell was completely hackable. Today's toys just have a transistor and a tiny chip under a drop of epoxy. No label or anything.
It's good to see people are keeping it alive, and not letting the multilayer surface mount stuff slow them down!
Just a couple of points.
By building his own high frequency oscilator, he has a better understanding of just what it is capable of. It's one thing to have a table of possible outputs for your high frequency oscilator. It's something different to know why those outputs are what they are.
Buying a 33.3, 100, 133 mhz oscilator should not be particularly difficult. I am reasonably sure that you could pick up some on e-bay and have them delivered next week. At the same time you will probably not get the experience you may some day need to replace the component should it fail on you. You will probably have to go out and pick up another one. By building your own, out of cast away parts, you will know what to look for to repair or upgrade the one you build. With 400mhz FSB systems out there today, (and higher) when one of these motherboards fails, you may find that it is exceedingly simple to determine what component failed, and possibly upgrade your variable frequency oscilator.
In a high proportion of the motherboards that I have seen fail, the primary culprit is the hig curremt transistors that support the CPU. When these go, it is very often visable as they leave a smoke patern on the heat sink they are mounted to. you may even see the resin housing for the transistor shattered or cracked.
If this is what has failed, then the CPU will not get power, and the board is functionally dead. It is very unusual for a failure of this type to have harmed the clock chip on the motherboard. I will grant that this is not always the case. It is possible to blow the clock chip, at which point the MB won't be able to start the CPU, or any of a dozen other chips and asics that will cause different failurs.
If you have a PCI modem, that takes a lightning strike, the most likely candidate for failure is the PCI bus controller. This does have a lead that goes to the clock chip, so you may loose the clock as well, but as he pointed out in the article, you can apply power to the clock chip and see if it generates a square wave on the outputs you are expecting, and if not, you haven't invested more than a little bit of time and thought to the project.
Then again, that's just my opinion. I don't claim to speak for the author of the article.
-Rusty
You never know...