Help! New rigs does not boot, fans spin for about 15 seconds then shutdown

[Caz5512]

Hi!

I just bought an upgrade for my aging/dead system.

This is what now have (UPS delivered today)

(new items)
Barton 2500+
512 MB Samsung RAM (1 stick)
Gigabyte GA-7N400-L
Sapphire Radeon 9100 128MB
360W tiger pro Power Supply

(I know, should have gotten better vid card and 2x 512MB, but I ran out of $$)

Old items:

Case (standard mid-tower)
DVD-ROM (Hitachi 4x)
7200rpm 80 GB Western
Symptoms:

Computer seems to boot, but shuts iteself off after 15-20 seconds. No POST beeps at all. Only the fans start spinning and MOBO leds come on.

If left untouched, the computer seems to try powering iteself on again 10-15 seconds later, but shuts down immediatly. It keeps at this until I kill the power.
If I try to start it up again immediatly, it does not even run the 15 seconds. But if I wait a couple of minutes, it runs the 15 seconds.

I have removed the extra USB connector and the Videocard and I do not get any different results.
I was suprised by this, as I thought that without the vid card it would at least give me POST beeps?

Steps I have taken

1- Unplugged DVD
2- Reseated Vid card and Ram
3- Checked power and cabling 2x


Any suggestions?
I do not believe I have a DOA, this seems to me like a power/heat issue or some type of RAM problem...

Any help would be appreciated as I want to play Enemy territory with friends later tonight!

Caz5512

 

[mechBgon]

Welcome to the Forums Caz5512 🙂

• Can you have a close look and confirm that your heatsink isn't on backwards? One end of its base has a step cut into it, like shown here: photo. The step is there to provide clearance over the raised, solid-plastic end of the CPU socket where the lockdown lever swivels: photo 2 In this second photo you see the heatsink's base avoids touching the plastic end of the CPU socket, which is good.
  
• Also, if you didn't remember to peel off the protective slip on the patch of thermal compound, that would cause problems too. If your heatsink didn't come with any patch, then hopefully you're using high-quality thermal grease instead?
  
• Since I've never heard of a Tiger Pro power supply, I'm wondering if it's another of these generic ones that cause problems despite their seemingly-ample wattage rating. :disgust: Start by checking the switch on the rear of it, to ensure that it's set for the right voltage for your locale (110V or 220V). Double-check that you got the secondary power plug hooked up, the four-pin one with two yellow and two black wires, too... the receptacle is at the CPU socket's 11 o'clock.
  
• Check the jumpering on your Western Digital hard drive. If it's solo on its own IDE cable, jumper it for the "Single Drive" setting, because this has been a known cure for many people with a Western Digital/nForce2 combo.
  
• Put your memory module in one of the three memory slots closest to the CPU, not the one at the far right.

Hope one of those suggestions helps 🙂 If it won't run, an Antec, Enermax or Sparkle Power 350W+ power supply might be a good idea, or there's a good dual-fan 360W Enlight for $39 shipped here if you're in the United States.

 

[hagrin]

Someone actually posted something like this the other day.

Make sure the heatsink is touching the CPU (I never would have thought about this until teh other day). Since it is just turning itself off, it very possibly could be in fact that.

Worth a shot I guess.

Otherwise, try swapping out the memory sticks and the video card. That would be the two normal things I would try.

 

[Caz5512]

Thanks for your advice

I removed the CMOS battery for about 1 hour, came back and nothing works now, don't even get the fans to spin up...
DAMN...


1) I will double check the heatsink, but I know what you are talking about, I remember asking myself why it was not equal underneath, until I read the AMD instruction poster that had very good details about the install (my Barton was a retail 2500+, it comes with a poster thingy to install the processor. It has what looks like a thermal pad underneath and there was no tape over it. This is consitent with the AMD instructions that mentions that it either has a peel off tape or some type of plastic cover for the whole heatsink (this is what I had). I had a little bit of trouble clipping the heatsink unto the processor (I did not want to use to much force, but it seems to require a moderate amount of force to clip it in...) so I will only unclip it as a last resort.

2) The PSU is a cheap powersupply, bought from a local store. But I have had this brand before without any issues. All of my power connectors are in (including the secondary 12v, as stated in the mobo manual). The Volgage is set to 115, the other option is 230...(I though that was supposed to be 110/220?)

I have 4 memory slots, I have tried the first two...closest to CPU.


The WD's jumper was set to Cable Select, I set it to single drive. That may have even been my orginal problem. But like I said, now nothing works...

Could I have fried something by removing the CMOS battery?


Next steps I will try:

1) I have other PowerSupplies, but they are only 250 and 300W atx. Will these work as a good test?
What is the minimum power needed?
2) I will check the processor and make sure the heatsink is correctly seated and the processor is correctly mounted.
3) If this fails, I will try and find another stick of DDR and another motherboard for the Barton...

I would not even know what to RMA now...

Thanks for everybody's help. Any more ideas?

Caz5512

 

[mechBgon]

Taking out the CMOS battery shouldn't hurt anything as long as the system's unplugged.

Your keyboard isn't plugged into the wrong PS/2 port, is it? The keyboard port is the one closer to the surface of the motherboard, color-coded green.

Otherwise... if you haven't gotten to this stage yet, pull the motherboard out of the case and lay it on cardboard. Give it just the power supply, keyboard, memory, CPU/heatsink/fan, and video card. Ream the video card into the slot very firmly, since an AGP slot has two rows of contacts.

Now you have the board laying on cardboard... use a metal object such as a screwdriver, and make momentary electrical contact between the motherboard's Power Button pins to signal it to start up. If it runs and you get a picture on your monitor, you're on the road to recovery.

If you get no picture, and the board does start but starts doing its wierd loop again, then cut the power and start taking stuff away. Take out the memory and see if it has some beep codes (and if it stays running now). Take out the video card. As a last resort, you can even take out the CPU. If you power up a completely bare board, your power supply should start running and stay that way.

 

[Caz5512]

I just another 300 PS, and it at least reverts back to the previous "broken, but semi powered" state. I think my 360W PS died. I will be testing with the 300W until I can buy another one, or confirm the state of the 360W in another machine.

I took out the videocard and it still does the same thing, the fans and leds come on for 15 seconds, and then nothing.
I even tried without the keyboard, to see if I works better. No go.

I will will try removing the motherboard from the case as a last resort, but since it will not have a speaker, how will i know if it beeps? or do I have to plug that cable into the case? which one of those small cables are REALLY necessary to boot, so I can try unplugging some of them. I think I will try only with the POWER SW and PC SPEAKER, just to see.

I double checked the board and it seems seater correctly on the mounting screws, and does not seem to be touching the case on the bottom. What about the back of the case, where the ports and cards touch the case. Is is possible to cause a short of the case is too close to the back?

I think I will need to remove the heatsink and check the CPU again, this is so similar to the problem that another forum member has a few days ago...

Shit! I thought I could install this myself, but I didn't figure on not having availaible parts for testing like with my Celerys and PIIIs...

Caz5512