BSOD on cold boot

The following issues are signs your power supply might not be up to the task:

1. During boot, your machine takes a very long time to get past the POST screen

2. During heavy graphics use, you start getting hard drive errors. The hard drives and the video card both live on the 12V rail and thus if the video card starts pulling too much power it can cause hard drive errors.

3. Memory errors in programs, but when you test the memory it tests clean. This is a similar issue to the second one. The electricity in a computer has to maintain a certain flow and that flow is used to run the memory. If that flow starts spiking around, it will cause memory errors.

4. Random shutdowns and reboots during CPU intensive activity. Like the memory, the CPU has to maintain a certain frequency across its calculations. That frequency is directly related to the voltage the CPU receives, and if this voltage fluctuates too much, the CPU will fail just like memory will. When memory fails, a program will crash, but when the CPU fails, the whole machine crashes.

5. Hangs during boot where POST never even comes up that can then be fixed by resetting the computer, sometimes multiple times. This is the same as the POST problem, where the hardware isn't getting checked properly and hangs while waiting.

6. Your computer starts to boot, but then just shuts down, or gets locked in a loop of constantly trying to boot.

There are some basic issues to consider. Wattage must be high enough to run everything, and for a modern machine that often means 500 watts or higher. However, people are buying 500-650 watt power supplies and still having problems, and the reason is that not all wattage is created equal. Amperage is just as important, and arguable more important on modern machines. Your video card especially will have an amp requirement. It should be right on the hardware specs where it will say "Requires a minimum of 28A on the 12V rail" or something similar.

Now, here is the tricky part: many modern power supplies share their amperage across multiple 12V rails. This was originally meant to distribute the current equally, making sure nothing failed. However, no one checked with the video card companies, who had figured out a way to dynamically alter the voltage required by the video card based on use. This gave us all those neat features like cards that can overclock when you start a game or where the fan changes speed based on the heat of the card. However, it also means those shared rails get all confused and try to balance the load, often causing graphical slowdown and hard drive errors.

This means a power supply may say "30A" and then have 4 12V rails sharing that 30A and balancing the load, which will not be enough for a video card that says "minimum 30A on the 12V rail". You will get tons of errors. This gives you two choices: shoot for about 5 amps higher than your card says, which may be pretty hard to find, or find a power supply that offers the amperage you need in a single rail.

I'm glad you agree with my assessment of the problem ComputersPlus.

This problem can also happen when the power supply has plenty of power. The problem mostly occurs on a cold boot because that is when power draw is highest. As soon as the board senses the power dropping below a certain level it will automatically shut down to save itself. It's a well documented problem with certain power supplies but total output sometimes isn't a factor.

I would still go with a higher output model for that system. I'd say 750 watts with plenty of amps on the 12 volts rails will do the job. I'm using a 1000watt modular unit myself for a single GTX280, which leaves plenty of power for my substantial overclock.

Thanks for your answers.

I couldn't image, that my PSU is responsible for these errors, because I disagree with most of the points computersplus posted. If my system is once running, I could play crysis at ultra details 2 weeks long without errors or bluescreens. No errors, drive errors and random shutdowns at all while system is running. I'll upload the dump file tomorrow and we'll know a bit more than.

Good night from Germany :)

I just wanted to post the info for you to digest this is something that gets overlooked all the time people focus on the other high power hardware PSU's are boring to most people but it is truly the heart of any good system and I have seen post after post here with people trying to run today's most demanding video cards with stock OEM power supply's or power supply's that promise lot's of watts but don't really deliver the goods these power supply's for sale are like the cheap stereo speakers and amps that say 1000 watts but only deliver 100 the way they rate the power supply's are trickery they list peak watt's which don't mean a thing you want constant power or they will rate it at 40 degrees celcius and not 50 because the lower the temp the more power it can deliver but it's not a relative factor we need to know what it can do at a high temp real world hot case not on a bench with no restrictive air flow issues so all is not what it seems.

Yes there are a lot of other things that could be causing the problem but until you resolve the power issue you will be chasing your tail in my opinion.

chev65 said:

I'm glad you agree with my assessment of the problem ComputersPlus.

This problem can also happen when the power supply has plenty of power. The problem mostly occurs on a cold boot because that is when power draw is highest. As soon as the board senses the power dropping below a certain level it will automatically shut down to save itself. It's a well documented problem with certain power supplies but total output sometimes isn't a factor.

I would still go with a higher output model for that system. I'd say 750 watts with plenty of amps on the 12 volts rails will do the job. I'm using a 1000watt modular unit myself for a single GTX280, which leaves plenty of power for my substantial overclock.

Loss of or unstable power good signal

dele2k said:

Thanks for your answers.

I couldn't image, that my PSU is responsible for these errors, because I disagree with most of the points computersplus posted. If my system is once running, I could play crysis at ultra details 2 weeks long without errors or bluescreens. No errors, drive errors and random shutdowns at all while system is running. I'll upload the dump file tomorrow and we'll know a bit more than.

Good night from Germany :)

Like I said earlier power output may or may not be a factor. The board is causing the shutdowns due to low voltage at start up, it's a built in safty feature. Easy way to tell is to try out an another power supply and see if the situation changes.

Here I come with the dump file. Bluescreen again today :/ But I found out, that I don't have to turn off the PSU, I only have to turn off my pc and turn it on again to get it working. Attachements are not working for me, so I uploaded it to a file hoster.
Btw: Updating to the latest bios version doesn't help.

Hotfile.com: One click file hosting

what are you using for security software? is this a brand new install? rootkit infections are everywhere and love to cause BSOD's

The installation is about 1 month old. I currently don't use any antivir software.