Help - How to debug a hardware failure

If you are using more than one stick of ram try taking out a stick at a time and see if one stick is faulty. I'm not exactly sure how to trouble shoot your PSU with out swapping it for another and seeing if it solves your problems. Let me know how thing go for you.

The biggest two things you want to narrow out before anything else are power/heat

1) Power - double check the main 24 pin connection and 8 pin cpu connection (also cpu fan connection for next step)
2) Heat - specifically cpu. Make sure the cpu heatsink is seated

I had this issue back during the holidays with a 775 build where the stock heatsink was refusing to stay connected to the board (pushing down the fourth locking mechanism pulled out a different one)

-After those are out of the way it's time to swap psus for a "known" working psu
-same with memory

Eliminating the fault from those three things in that order narrows it down to your motherboard

I assume you have access to a computer or another device you are posting with on AVS. You did not say which operating system you are using. I would recommend booting in safe mode and reading everything that comes up on the screen to try to pin down where the computer is hanging up. Once you find where the computer is hanging up you can search the internet for a solution.

Another method is to strip the computer down to the motherboard and CPU and SSD, plug the monitor into the motherboard and start in safe mode. If that works begin adding other components one at a time and booting until you find the culprit. If it doesn't work replace the SSD with another storage device with basic software and try to boot.

Frequently lie down and put a warm rag on your head. Heavy alcohol consumption will not help.

Swapping out RAM and the PSU is the easy option. If you can't get past the BIOS then the issue is not with Windows. All you need to run the BIOS is a motherboard, CPU, RAM and of course power.

David

The last problem I had was with the windows boot loader. It was not the PSU or the hard drive both of which I reused in the rebuild. I eventually decided it was the motherboard because that was the last thing that posted on the screen was the windows boot loader before boot hung up. But I do not know what your problem is. But my system was about four years old and I just rebuilt it as an I5. It would be worthwhile to blow the case out with some canned air and reseat your boards. That would be cheap and easy.

Boot your machine up with a single stick of RAM and keep swapping the RAM sticks until you find the defective one. I just fixed a friends PC who was having an almost identical problem last month.

Are you reaching the same point every single time?

Did you make sure that your processor heat sink was completely seated? Is there anything white/crusty around the heatsink/cpu connection? How long had it been running before issues

I take it that you mean the 3 strips of thermal compound that comes with the cooler as "pad" like in the picture at this link http://www.hardwaresecrets.com/imageview.php?image=35163

The stock i3 cooler is sufficient with Intel's provided thermal compound (though most of us don't use it). I've seen crusty areas form when a loose corner keeps that portion of the heatsink from making contact with the motherboards cpu socket heat-spreader. It sounds like you're planning to order memory anyway, so add a $4 tube of thermal paste. No need to get the extreme high end. Use some alcohol on the existing processor and cooler to completely clean it, then reapply and make sure it's got good contact

I had the problem you're experiencing happen to me, and re-seating the cpu heatsink fixed it for me. I've had the same tube of thermal paste sitting around for a long time. It doesn't take a lot, so I cleaned the old off and applied new.

It doesn't make a difference to me when I'm selecting parts. I usually go for the cheapest. Last two times I bought memory were both frontpage slickdeals. 2x4GB Ripjaws for $30 was as cheap as I could find for anything, and about 6 months later 2x4GB Patriot Viper was $24 so I bought it as well. I actually split that one into two different builds. I've always gone cheap, never had any problems

To reiterate

1. Power
2. Heat

It doesn't "seem" likely to be power if you are sure that all the PSU cables are staying connected (sounds like a loose cable). If it's not a loose cable, then testing will only be an exercise in futility. Here's a good reference for testing with a case fan, but if it's powering on at all the odds are extremely low that a rail in your psu has crapped out and stopped supplying sufficient power in the past three months. Very extremely rare, but just to check use this guide and let it run a case fan LINK

If you reseated the CPU heat-sink with new thermal compound after cleaning the old compound off, you should be okay for cpu. If not, you should try re-seating the CPU heat-sink and make painstakingly sure the thing is seated. The next area to overheat and cause such issues would be the passive cooled PCH. In this image, LINK, it is the blue square near the bottom left. Be extremely careful and just visually inspect if any of it's edges wiggle when you touch it. Similar to a jeweler inspecting a diamond ring in its setting. Look very closely and apply light pressure at the corners. It its loose it should be evident

Reseating the PCH would be a next step, but I would first contact Intel support and explain the situation and troubleshooting steps you've taken.

Since it sounds like the problem is actually getting worse (you've gone from bluescreens to no post), I'm betting on a dying PSU instead of heat issues. I've POSTed 35W i3s without a heatsink.

After memory, PSU, and heat, some other less-likely candidates are:
--Electrical short: pull everything out of the case and wire it up on a tabletop.
--It's a video problem: plug in a keyboard and press Numlock. If the indicator light changes, try testing with a known-good video card.

If you can, test each component with other known-good components separately.

To each their own I suppose. We only know i3, and with an mATX rather than ITX it's probably one of the 65W. I had a 65W core2 shutoff mid-post with a "loose" heatsink. I had a lot of similar problems in the past with a EVGA 790i motherboard that I finally fixed by replacing the south bridge heatsink
Since OP is using integrated, I'd just say I've not yet heard of an i3 working for months and then crapping out. However, this is a great troubleshooting tip for that

Have you reinstalled on the new board? What is the error it crashes with?

Yes but what is the exact message on the bugcheck? Or alternatively load up the dump in c:\windows\memory.dmp and poke around

Also why do you think this is a HW issue?

Sorry - I thought the machine was blue-screening. What do you mean by "bios flash screen"?

When you first turn on the PC, the default behavior of most BiOS is to show you a splash screen instead of the detailed BiOS screen. The detailed screen scares most consumers. The splash screen would be something like a mostly black screen with a huge INTEL INSIDE logo on it - displayed during the brief time the BiOS is figuring out the hardware in use. In the old days you could see it for many seconds, now it is there for under a second.

I'm familiar with the BIOS. When the OP said "flash screen" I thought he meant flash as in flash update. Now that you mention it, I suppose he could have meant splash. OP - please clarify what you see.

Intel boards sometimes have a jumper for something like factory mode vs normal mode - make sure it's in the right position