Last week I walked in my room smelling burning electrical. My power supply in my emachine was fried. I got a new power supply off of ebay, installed it, and crap! the computer turns on for a few seconds and shuts itself off. Any ideas on what could be wrong? Thanks for the help!

well that eMeat could simply have just given up on life, and died.
thats what happened to my old eMeat.
plug it in, turn it on, and it turns on for a split second, then turns off.

Hmm. It seems like maybe the mobo got fried? :confused:
Not sure.

The last time I worked on a fried PSU, the CPU was bad also. But this only led to the power supply powering up but just sitting there with no video etc until I'd power it back down.

Yours:
If it powers up for a few seconds, my thoughts are either the 'power good' sensor (mobo related) is not sending back a signal to the PSU, OR the processor may be overheating.

My power supply in my emachine was fried.Eeeee...

Was your eMachine system manufactured between 2001 and 2004? Did it have a 250W or 300W Bestec PSU? If so, from what I've seen, the catastrophic failure rate on those PSU's is around 50%!! :eek: And I do mean catastrophic! We're talking like a 50% chance that the PSU dies and takes something with it! Usually the mobo.

If you're lucky, either by the 50% mark or the year of manufacture, you should first try resetting the CMOS and removing the battery. See if that helps.

... my thoughts are either the 'power good' sensor (mobo related) is not sending back a signal to the PSU...
"Power good" is not a sensor on the mobo, it is an output from the PSU.
http://www.pcguide.com/ref/power/sup/funcPowerGood-c.html

"Power good" is not a sensor on the mobo, it is an output from the PSU....which must also be reciprocated by the mobo in order to complete the handshake.

It takes both to make the system run. Just because the PSU reports itself as being good doesn't mean the mobo is ready as well. If you and I were getting ready to go for a walk, and you came and told me you were ready to go, what would happen if I just stared at you blankly? :p

What on earth are you talking about?
There is no feedback from the mobo to the PSU. No handshake.
The PSU turns on when the PS-ON signal is asserted, and thats it. It does not need any handshake.
The mobo just waits for the 'power good' before the CPU reset is released.

Checked all the drives are plugged into the mobo properly - especially the floppy drive cable.

There is no feedback from the mobo to the PSU. No handshake.Good point. I used the wrong word there. It's not a handshake.

However, even if there is no "input" or "signal" going from the mobo to the PSU, FTT had a good point as well about whether or not the mobo's sensor was still working properly. When things get fried, power-plane functions are likely to die first. So even if a new PSU is sending a proper power-good signal, that doesn't mean the mobo is interpreting it correctly.

Maybe I should rephrase?

When I replaced the PSU, I still had problems. No POST, nuthin. It would just sit there with the power supply running . It was not until I replaced the CPU (after replacing the PSU) did the computer return to functioning. But the PSU needed replacing most definitely because there was absolutely NOTHING up to that point.

Maybe the power good has to do with the PSU and not the mobo but it still depends on a return signal (path) from the mobo to have it decide whether or not to continue or provide full power. A matter of semantics? Doesn't matter. I'm trying to help someone fix their computer, not keeping track of my grammar etc... It is in this way I thought that the mobo used a sensor and not the other way around.

A bunch of people here go directly to the mobo as a scapegoat for a problem and I just felt the need to mention that this is not always so. In my case, the mobo was just caught in the 'middle'. ;)

saphalline - you "corrected" my post with complete nonsense. You can brush it off with "wrong word", but it seems clear that you have no idea how this works. It looked very convenient of you to 'forget' this thread.

FTT - you are still wrong about this "return signal". There is only one signal from the mobo to the PSU, and that is telling the PSU to turn on or off.

This is not about grammar. I am sure both of you are good about helping people, but when you think you know more than you actually do you can also confuse people when you provide totally wrong info.

Hagar - It's unclear where you're going with this. You have provided no help at all in this particular thread - all you've done is tell those of us who are trying to help that we are wrong. What's more, you're doing it in a very confrontational way. We don't really appreciate that around here, especially where your implications are concerned.

I'm willing to admit that my first post in this thread was incorrect thinking on my part. It was late and I applied concepts that were untrue. I was trying to get at the point that the PSU's power-good signal is useless if the mobo is broken in some way. Perhaps I did not make that clear enough for your tastes, but you could have simply asked for clarification or called me on an error, instead of saying "What on earth are you talking about?" and "it seems clear that you have no idea how this works". Or perhaps you are too new to this forums to know that we value kind and helpful people who offer constructive advice.

Think about that next time you repeatedly tell people they are wrong while offering nothing helpful yourself.

He's from norway, a cold place...

The PSU 'power goods' the load that exists on it at power on, and then it's turn for the mobo to respond, if the mobo can actually respond. of course, cpu must be alive for this. well.. this is my opinion, since this thread is now about opinions...

oh, no hagar hasn't been back since after this post.

i forgot to mention that norway's a cold place filled with people who have warm hearts!