QOTD: Did You Ever Fry Your PC by Overclocking?

Page 5 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
Status
Not open for further replies.
yeah! my gigabyte p35 mobo, I tried to install the new big heatsink for Q6600, with those stupid pushpins of intel's! I didn't notice that I cracked one of the components of the mobo, i think it was a capacitor! long talk short, after installing the new HS, started the system, after a minute, BOOM! heh! even now I miss my P35! good for me CPU was not damaged!
 
[citation][nom]jsc[/nom]First OC: 1978 - TRS80 Model I Z80 from 1.77 MHz to 2.01 MHz. Did it by piggy-backing chips, cutting PCB traces, and running jumpers.Had a Pentium 233MMX in a Acorp motherboard. Eventually found a set of undocumented jumper settings that let me run it at 333 MHz. Fitted a S370 cooler when I did. Threw it out last summer.The famous 300A Celeron at 450 MHz.E6600 to 3.6 GHz. Q6600 at 3.6 GHz. Q9550 at 3.6 GHz. It'll boot at 3.8 GHz. but it's not stable.[/citation]
man you got history!
 
I find it appropriate that the picture at the top of this page appears to be one of AMD Socket 462 chips. I work on computers and I tend to see these fried more often than any other CPU type. The thermal sensor/controls on these must have been really poor. If the fan goes out or gets filled with dust, these just get hotter and hotter until they fry. Intel seems to have this figured out for a while and the newer AMDs (socket 754, 939, AM2, etc.) seem to have gotten this solved.

There was a video of these 462's frying here a while back when the heatsink was removed. They would actually smoke and the temps reached were like 700 F.

Also, the motherboards used back around the times of the socket 462 were equally bad. The electrolyte in the capacitors seemed to be the weak link. I have seen who knows how many with bluging and blown capacitors. These were often the budget option at this time and it seems that cheaper components were used. I personally owned several 462's and loved them but wouldn't want to count on them to live a good long life. These just tend to crap out over time.
 
I have fried THREE AMD chips!!! The first one only quit working at higher speeds, but continued to work at stock speeds. It was an AMD K6-2 @550 rather than 500 lol. The second was an Athlon 1900 Palomino, I never cleaned the heatsink thoroughly, and it got all gummed up and overheated the chip, it died. The last one was an Athlon XP 2400+ that I dearly loved, but she just quit working on me one day. I did replace the XP 2400+, but I haven't looked back since I switched to a Core2Duo 2160, overclocks very well.
 
I had a 1.4 Athalon chip and Mobo that I was messing with after I had built a new gaming rig. I wanted to try O.C.ing it to improve performance and use it for LAN parties. I was drunk one night and was messing with the FSB trying to get another bump and I had it set way too high. Somehow I remember knowing it was wrong but forgot to fix it before I rebooted.
 
[citation][nom]cwatkin[/nom]I find it appropriate that the picture at the top of this page appears to be one of AMD Socket 462 chips. I work on computers and I tend to see these fried more often than any other CPU type. The thermal sensor/controls on these must have been really poor. If the fan goes out or gets filled with dust, these just get hotter and hotter until they fry. [/citation]
Back before the athlon 64 people were told to go for intel chips (p3/p4) because these had throttling/shutdown features in case of overheating. Amd processors at that time didn't. So if an amd cooler would fail, you were more likely to fry the cpu. An equivalent p4 would just crash and wait for you to turn off the system. Also the very poor attachment type on socket a meant that heat sinks would more frequently be improperly mounted. And the clips tended to break if you were to remove it too often. So all in all, amd was just more prone to failing.
I still had 2 thunderbirds and 3 bartons processors before I switched to a northwood though (replaced by a brisbane before the opty and the current c2d). Socket a processors cost 30-35% less than intel, so it was worth the risk, but the risk was there.
 
My favorite overclock was back on my 366 Celeron. Got my Voodoo 3000 running a good 30 percent faster and cooled it with a shoestring mounted across the inside of the case to hold an extra case fan that pointed straight across the card and blew it out of an empty but open PCI slot (right by the AGP). It was cooler then at default clock settings! It was so fast that replacing it with a Voodoo 4500 made no difference in performance except with applications that needed the extra memory (the voodoo 3000 only had 16MB).

No, ever fried anything.

My current system has an overclocked FSB, Kingston Hyper X memory from 400 to 466, Geforce GS Overclock (came stock overclocked) up 50 mhz on the core and a whole 12 mhz on the memory. That may not sound like alot but I think it actually increased performance a good 10-20 percent. Yes, an old system, but all the overclocking and Windows tweaking has so far saved me from an upgrade.
 
A few typos. It was a Geforce 7600 GS.

No, never fried anything! And I've overclocked every system I've ever owned at some point or another. I think one of the keeps is overclocking as a last resort rather then the life of the machine. I never increase voltages from defaults. After an overclock I always check overall case temperature and try to find ways of better cooling. Can be as simple as moving a cable out of the way. If you can't fix it or add another fan, don't overclock.

Also take it in small steps if you don't know what your doing. Better to burn a whole day of constant testing with 1 mhz increases then burning the card or chip. Also backup any files you accessing during the overclock in case of data corruption (especially from overclocking CPU's).
 
I remember my first overclock was with a XT. the standard ones ran at 4.77MHz, and if you changed the clock generator crystal you could bump up the speed, got it to run at 8MHz like that.

The first time I noticed that one needs to be careful was with a Pentium 60 which I overclocked to 66MHz, it had a TINY cooler on it, and after it had run like that for quite a few months, i had the cooler off one day and noticed that the sink had burned lines into the bottom casing of the fan.

As far as toasting something due to overclocking or attempts thereof, killed 2 items so far.

1st was a K6-2 450, it was totally unintentional though. I had a Creative AWE64Gold soundcard, and the drivers could not detect that the K6 was indeed faster then the minimum spec of 133MHz, so I put a Pentium MMX 233 in the board just to load the drivers/software for the card. At that stage you set everything using dip switches and jumpers. Once I had finished, I put the K6 back, changed the multiplier and the bus speed, but forgot the core voltage, POW!.

2nd was with a barton/thuroughbred or something thereabouts, it was a 1600+ socket462 chip - that I can remember. I saw on the 'net that one could re-do the "jumpers" on the chip (THG had a video showing this too) and i think i cut a bit deep on the removal of the one jumper and killed the chip.

My coolest overclock with the least amount of work applied to it must be my 2500+ Barton, dial up the bus speed from 333 to 400 - voila, a 3200+ :)

Now, I am running my Opteron175 @ 2536 (up from 2200) and my 8800GT at 647/966 from 600/900 all on water... nothing extreme to be honest, but it works :)
 
fried a athlon 3200+ newcastle, dfi lanparty 250gb, and 1 gb patriot ddr600 ram in one go. got all RMA'd for free. next time i didnt go as far. its been OC'd since.
 
[citation][nom]akoegle[/nom]The only thing in this message that's useless trash is your typing.[/citation]
You're being too nice, akoegle. That's not the only useless thing. His imbecilic message, his knowledge of the English language, his knowledge of the immediate subject matter, and not to mention his attitude are all equally infuriating. :fou:
 
Status
Not open for further replies.