Help! Two PCs not working.

[ViperVin2]

I really need help right now since two of my computers are not turning on. Rig 1: ECS K7S5A w/ Athlon XP 2000+, Rig 2: Asus A7N266-VM w/ Athlon Tbird 1.2 .

On Rig 1, I was replacing the stock AMD HSF with my Dr. Thermal but the temps on the CPU were rising really fast, going from 35c to 54c in less in 30sec. I noticed that the CPU core wasn't centered with the copper core of the Dr. Thermal HSF. I took that out and put the AMD Stock HSF back in, only to find that my PC wouldn't POST. The fans and harddrive spins, but I don't get a beep and BIOS doesn't load. I cleared CMOS but that didn't help. All it does is a 3sec loop trying to POST, resets, and tries to POST again, but it never loads BIOS.

On Rig 2, I was replacing my duron with the Tbird 1.2 and when I got into BIOS, it read as 900MHz instead of 1200MHz. I turned it off and checked to see if the jumpers were set at the right bus frequency. They did check out at 133MHz, so tried to power up the pc, but nothing came on. Using the power supply from Rig 1, I tried to power up again, but nothing. I am thinking something is either wrong with the mobo or power switch.

So I am left with two dead PCs, and am posting this message from my sister's PC
Help...

-Viper

 

[mechBgon]

Ok, for starters, are you aware of the asymetrical nature of the heatsinks? They only go one way, and if you put them on 180° backwards you're in for big trouble.

 

[mechBgon]

Maybe I should add more detail. The base of most AMD heatsinks, including your retail one and your Dr. Thermal, have a step cut into the base at one end. This step gives clearance over the solid end of the CPU socket. The clip is also asymetrical, with the pressure point closer to one end than the other.

If the heatsink is 180° backwards, two bad things occur:

• I love bullets, by the way
• The solid end of the CPU socket jacks up the unstepped end of the heatsink, preventing it from making square contact with the CPU core, preventing effective thermal transfer and placing lots of pressure on one edge of the CPU core, risking core breakage
• The pressure point of the clip is nowhere near the core

The Dr. Thermal is a bad design. Even when installed correctly, the copper core is not on center with the CPU's core, so if the copper core is slightly recessed due to manufacturing error, you get rim contact from the surrounding aluminum (intense pressure plus lack of adequate contact with the core).

On the A7N266-VM, if your CPU has a thermal diode (which the AthlonXP does but the Thunderbird doesn't) and the board detects a thermal spike from the CPU's thermal diode, it immediately cuts off electrical power to the CPU. You would see it spin the fans for a moment and then shut down again. With the A7N266-VM/Thunderbird combo I think the Thunderbird could easily fry itself in 10 seconds if you had the heatsink on backwards, because there's no thermal diode on the CPU to warn of overheat.

On the K7S5A, I don't know if it has any thermal-shutdown circuitry like the A7N266-VM does. As a result, if you had either CPU in the K7S5A and put the heatsink on backwards, you could probably fry the CPU in about 10 seconds too due to the lack of overheat-protection circuitry on the mobo. Here's hoping that isn't the case.

I sure wish AMD would have designed SocketA so it had two lugs on one side of the socket and one lug on the other, preventing reversed heatsinks :|

Solution for testing: get out the Duron (I assume it's healthy) and test the boards with the Duron and the retail AMD heatsink (using thermal grease after scraping off the stock gummy-pad). Make sure the heatsink's oriented the right way. Once you know a board works, then you can drop the other CPUs in there and see if they're ok.

Addendum: what brand and model of power supply are you using?

 

[ViperVin2]

For my rig with the A7N266-VM w/ Tbird 1.2, I am using the Alpha PAL6035 HSF. I am pretty sure I installed the HSF correctly for the tbird, but even if I did fry it, shouldn't the mobo/fans/harddrive still power up? I also forgot to mention, when I noticed that my CPU was being read has 900MHz instead of 1200, I turned it off. When trying to look for the jumper, I heard whisper quiet noises, that sounded like frying. I don't see how the CPU could have fried because the temps were around 35c. So could the mobo have fried somehow?

For my rig with the K7S5A, will it POST even with a fried CPU?

The PSUs are:
Rig 1: In Antec Case - Antec PP303X 300W
Rig 2: In some case I bought from NewEgg - Enhance ATX 1125BTA 300W

And I will try to test out the duron tomorrow morning, its getting late now.
Thanks for the help.

-Viper

 

[mechBgon]

The whisper-quiet noises could be the PSU doing normal stuff.

As for the temp reading, nope, it could still fry with a 35C reading at the socket thermistor. If the heatsink's not on right, the CPU can go from room temperature to 100C+ in seconds, fry and die. The external thermistor on the motherboard's surface is only going to notice a little bump, because it's isolated from the CPU itself by a big air gap plus the CPU's base package. Sorry that sounds so pessimistic...

The Antec PP303X is the oldest of Antec's 300W units and has a skimpy 150W available on the 3.3V + 5V lines. It's concievable that it might not be able to get the 1.2GHz or 2000+ rolling... you could try booting with no drives plugged in (neither data nor power cables, sometimes a non-powered drive on a data cable will confound the mobo). The Enhance is beefier and was on AMD's recommended list for up to a 2100+ IIRC. Today's Antec 300W unit has about 50% more power on the 3.3V + 5V lines, by comparison, so you may want to pick up a new PSU for one of those rigs and use the Enhance for the other

No board is going to POST with a dead CPU, nope. Incidentally, the A7N266-VM's 100/133 jumpers are depicted upside down in the manual. If your Tbird 1.2 is designed for the 100MHz bus instead of 133 then that could cause it to not POST because it's overclocked at the 133 setting, which the manual would say is the 100 setting.

Good luck! Mobos are hard to kill, so it could be the CPU, but hopefully it will all work out.

 

[ViperVin2]

Well, I got a few updates: Tried out the Duron in Rig 1, and it did work and POST, so I did fry my XP 2000+ (I blame on Dr.Thermal heh) For Rig 2, I tried out the Duron but it still won't power up. Nothing turns on, no fans, no harddrive, nothing. All I see is the green light on the A7N266-VM mobo. So what options do I have left now? RMA?