TR Forums

So I had this really old Dell computer that is at least 10-12 years old that has been sitting in my basement for I don’t know how long.

I decided to have some fun with it and see how much abuse the mo/bo could take before it gave out.

I have always handled computer hardware with extreme care, making sure nothing is inadvertently damaged while putting a build together.

My first test was to build up a static charge by walking around on the carpet and then discharge into the mo/bo while the computer was up and running.

One charge caused the computer to restart, otherwise it still ran just fine.

The second charge caused to computer to completely shut down, while it would not restart, once I unplugged it for a minute and plugged it back in, it fired up and operated perfectly normal.
Interesting so far, but let’s take it a whole new level.

While the computer was booted up and sitting at the desktop, I took a small screw driver and lightly, touched various parts of the mo/bo and did not get anything to happen, so then I dragged the tip of the screwdriver over the surface and finally the computer instantly powered off.

Once again it would not start back up until I unplugged it for a minute and plugged it back in and again it booted up and operated normally.

Next, I decided to see what happened if I picked a Cap randomly and twisted it until one of the leads broke, I thought for sure this would kill it, but to no avail , this computer still booted up and operated normally.

Ok, time to get some water involved, I got a glass of water and while the computer running and sitting at the desktop, I took my fingers and dipped them into the water and flicked droplets of water onto the mo/bo and everything else for that matter, it took far,far more water than I would have thought until the machine suddenly shut off.

No start up, no start up after it was unplugged for a while, so I thought I finally did this computer in, but the next day after it had dried out, the freaking thing fired up and booted to the desktop and operated normally.

I am somewhat at a loss at how this computer is not dead yet, but now I find myself rooting for it to keep taking the punishment as I think of some news ways to test its seemingly amazing levels of endurance.

I should have taken some pics or video, but I did not expect this computer to survive the screwdriver, let alone disabling a capacitor.

I may try spraying some whip cream into it or something like that next, any ideas?

Tannerite + a rifle from a safe distance.

I would make a video and see just how many capacitors it takes for the computer to finally give up.

Throw in some witty commentary, followed by some painful screwing and unscrewing of capacitors until the thing blows up!

Well tearing a single cap off is probably not going to do much. Your motherboard is a digital device, after all, and those capacitors are really there to keep the signals looking "digital" more than they are actually performing a constant function as they would in analog devices. (Removing more of them would allow more transient behaviors to pop up, eventually resulting in errors).

The screwdriver shutdown is probably just a short between two circuits. No big deal, and it sounds like the motherboard did what it was supposed to to - shut down to avoid damage. The water probably did the same thing.

If you want to do permanent damage to this thing, you need to...well, do something that is more permanent. You'll need to damage the traces such that shorts occur all the time. It could get somewhat spectacular if you remove heatsinks, though. Without cooling, your silicon should essentially melt itself (assuming that your motherboard can't shut down again when the a temp spikes).

Now if you want something interesting (read: "partially dangerous"), you could reverse the voltage to the capacitors and force the PSU to stay on somehow. This is partially dangerous because electroytic caps will heat up and explode if hooked up backwards, and the gas that is expelled is toxic. You might also kill the silicon chips by forcing current in the wrong direction.

You're US based right?

Take a very long stick and switch the PSU from 115V to 240V while it's loaded. Make sure it's not a metal stick while you're holding it...

Once upon a time I had a drive power connector attach itself to a mobo USB header...... The system did not shut down. The power wire melted, glowed and that mobo was dead. The PSU was an old PC Power & Cooling 300W and apparently didn't detect this as an overload situation.

superjawes wrote:

It could get somewhat spectacular if you remove heatsinks, though. Without cooling, your silicon should essentially melt itself (assuming that your motherboard can't shut down again when the a temp spikes).

The old videos of Athlon and Duron chips exploding when heatsinks were removed under load come to mind!

My brother works at a concrete place and they have several loaders.

Could put it on a concrete slab and lower the bucket.

http://www.im-mining.com/wp-content/upl ... -wa500.jpg

Well you have the usual: iron filings, steelwool, saltwater.

If you want to get creative, put it in the oven for for 30 minutes. It will probably come out and still work.

Then there is the nuclear option: put it in the microwave.

swaaye wrote:

Once upon a time I had a drive power connector attach itself to a mobo USB header...... The system did not shut down. The power wire melted, glowed and that mobo was dead. The PSU was an old PC Power & Cooling 300W and apparently didn't detect this as an overload situation.

If the motherboard wasn't drawing more than the PSU's rated amperage output it would think it's normal operation. Those tiny ~24 awg wires in the usb headers don't tend to like double digit amps, which the PSU would gladly feed it.

I did something similar a while back. I tried to kill the comp via burning it up on its own power. Even though it had two bulged capacitors in the VRM, I couldn't get it to fry itself...even with the cpu heatsink off. Damn safeguards.
(It had been running for years with the budged capacitors, so I wasn't willing to spend the amount of time necessary running Prime95 OC'd to fry those)

This thread is fantastic. I like your attitude.

Another thought occurred to me, I could have it running as normal and then with the side panel off, blast it with a hair dryer and see how long it runs while be blasted with heat.

Jam some expansion slot covers into the expansion slots and boot.

I also thought about pushing one of the locking pegs for DIMMS and pop a memory stick out of its slot while running.

Scorpiuscat wrote:

pop a memory stick out of its slot while running.

I haven't tried that personally, but I'd guess it would only result in a one time system crash. As soon as you boot up again everything would be fine (1/2 the RAM available if you don't put the DIMM back in, or full RAM if you do)

How about hot-plugging/unplugging hardware? I've see machines where PCI cards have worked themselves out, yet everything continued to work after reinsertion. I've also seen people only partially insert RAM modules. Try hot-plugging cards, RAM, and power connections like the CPU 12V. You could also overload one string of PSU power connectors, or try overdrawing power various types of USB ports (2.0 vs 3.0, front ports vs rear header, etc).

If it's 12 years old it might have one of those big cartridge processors installed. You could easily pull that out while running.

Pez wrote:

You're US based right?

Take a very long stick and switch the PSU from 115V to 240V while it's loaded. Make sure it's not a metal stick while you're holding it...

I've done that many times. No damage. It's when you're EU or UK, and you flip *from* 240V to 120V, with 240V coming in, THAT will blow stuff up.

jackbomb wrote:

If it's 12 years old it might have one of those big cartridge processors installed. You could easily pull that out while running.

I once had a really sweet P2B-D setup with dual Coppermine P3s on it. The Athlon came out and I was hot to get on board, so this nice guy Twofer from these forums bought my old rig off of me.

Some months later, I was chatting with him. He'd been using the PC and it seemed sluggish. He checked task manager and only one CPU was showing. The other had come unsnapped at the top and partially lifted out of the slot.

He shut down, snapped it back in, fired up, and all was well.

Killing this machine may be more work than you'd think. They were built much more robustly, with wider tolerances, back in the 90s and early 2000s.

Personally, I'd take it somewhere well-ventilated and fire-resistant, outside, and run an extension cord to it. Fire it up, make sure it's running well, then splash some lighter fluid around inside and torch it up. I'd take bets if the thermal protection in the CPU would turn the machine off before something cooked to death or popped. Could be any of those.

Hook it up outdoors somewhere .Pour liberal amounts of gasoline on it, light a match. Record the results on video for science.

edit: whoops, Forge already said that. From my significant experience lighting things on fire, you may want to try brake fluid. It burns nice and slow, but smokes a lot (I don't live in a place where air pollution is much of a thing)

Come on guys, think bigger. Since it still runs.... Create a list of outcomes (see below). Write a program which spawn a thread for each possible outcome. Each thread loops and generates a random number. The first thread to generate the magic number wins (oh, the suspense). The threads spawn a physical activity when they "win", which produces one of the outcomes in the list. For example:

start the floppy drive, which powers an igniter in a bottle of alcohol inside the computer
Start the CD-ROM, which powers an igniter in a batch of thermite sitting on top of the case
Activate the hard disk, which powers a solenoid that drops a 100 lb mass onto the computer.
Output a tone on the speaker, which powers a solenoid that drops an ax into the mother board.
Outputs the color "blue" from the graphics card, which ignites a black powder charge taped to the processor.

You get the idea... leave the method of destruction up to random fate and let the computer "decide". start the program (I'd suggest a remote login from a good distance away), sit back with a beer, wait and watch.

How dangerous are you willing to get?
Take off the CPU heatsink, and buy some black powder/ gun powder. Then ask the question: will it get hot enough to ignite the powder?

OsakaJ wrote:

How dangerous are you willing to get?
Take off the CPU heatsink, and buy some black powder/ gun powder. Then ask the question: will it get hot enough to ignite the powder?

The thermal limiters will prevent it from getting hot enough... unfortunately.

Actually, that depends on how old the old PC is. If it is Athlon pre-XP or pre-P3, it'll get pretty hot. Probably not hot enough for black powder, though.

The etherkiller would definitely be unsafe.
http://www.fiftythree.org/etherkiller/i ... killer.jpg

Hmm, assuming that it's a beige mid-tower, that would make for a sporting 400yd target and a right challenging 1000yd target. Bring your high-velocity flat-shooting rounds (130gr JHP in .270 Win for me). Toss some Tannerite in the case to signify a hit. Closest to the CPU socket wins.

EDIT: JAE; JBI agrees.

Install Windows XP with no virus / malware protection, no service packs, no firewall, and surf the web with IE6.

Should be far more entertaining than shooting it.

JustAnEngineer wrote:

The etherkiller would definitely be unsafe.
http://www.fiftythree.org/etherkiller/i ... killer.jpg

One altsysadminrec graphic:

http://upload.wikimedia.org/wikipedia/c ... coa-na.svg

You remember the motto right? "Down not across"