Quote:
Originally Posted by Jinn
It's one thing to be wrong, it's another thing entirely to try correcting someone else when you're wrong.
|
That kind of personal attack isn't necessary, especially when you can't seem to even settle on what you're talking about from one post to the next.
Quote:
|
This is patently false. We're not talking about digital signal processing; this is analog power and it subject to peaks and valleys, even with a power conditioner. As long as the CPU and mainboard have adequate power, the computer will not just 'shut off'. I've doing computer repair on the side for about 10 years and I've seen faulty PSUs cause all sorts of intermittent peripheral failures, including intermittent black screens and blue screens.
|
WHAT is patently false? That the PSU isn't going to pick favorites and that it just supplies power on a first come first serve basis? And what on earth are you talking about with "digital signal processing"? I didn't mention that until I responded to your contradictory and backwards claims about DVI vs Dsub power draw. When you say ridiculous things that have no connection to reality it makes me wonder if you're trying to deliberately obfuscate the issue and distract people from actually looking at the facts of your claims.
You just proved my point, the power supply is going to give power to whoever needs it first and if the CPU and mobo lose power the computer shuts down. If his mouse, which is incapable of drawing more than 5v@500ma unless it breaks the USB specs, is pushing his PSU past its limit of being able to power everything else then his CPU and GPU would do the same thing when they start drawing more power.
I don't discount entirely that it may be a faulty PSU contributing to the problem, but the fault is most certainly not that the two and a half or so watt power draw of his mouse is overdrawing his PSU but somehow magically his CPU and GPU can draw dozens of watts more. Whatever the PSU could be doing to cause this, it's not from the mouse overdrawing it.
Your argument just does not make any kind of coherent sense in terms of a working computer, hell it's not even
internally consistent with itself like I'll point out further down...
Quote:
|
I think you're essentially proving my point here. Any spike in draw can cause a PSU to be unable to provide appropriate power to all of the devices. I'm not sure why your monitors staying on is important at all, seeing as how they're self-powered. If you mean to use it to say that the CPU is more likely to fail and cause a hard shut off than the GPU, then your anecdotal experience is great, but *ANY* part can fail depending on the immediate voltage conditions to them all. I've suggested the mouse because it is linked to the actual failure, and a ground condition on the USB jack could easily cause a momentary power draw. It just as easily be ANY of the peripherals, but the solution would be in replacing the power supply, not the peripheral.
|
Again, no, you're proving mine. If any spike can cause the PSU to be unable to power everything how are the CPU and GPU's many times greater power draw when he suddenly starts doing something that requires them to work hard going to be exempt from this while his mouse, which is not capable of drawing more than about two and a half watts (since its unlikely they're totally throwing USB specs to the wind for a mere mouse), is somehow able to bring his power supply to its knees?
If you want to start bringing in special conditions like the USB mouse interfering with the monitor somehow still being related to the mouse overdrawing the power supply then good luck with that but don't expect me to wait for you to find some way for that to work without also messing with the rest of the motherboard, which is where that USB jack attaches to and would be solved just by switching to a different set of ports (or even a different one in the same set)
Quote:
|
That's great when you're talking about sustained, but this appears to be a spike. Even temperature-ramped fans draw a stable current and aren't subject to spikes like DRAM, CPU and GPU are.
|
Sustained what? Spike of WHAT? Are you talking about a spike in voltage or amperes? A fan suddenly jumping to max speed and the CPU suddenly getting loaded are both sudden increases in the amount of power being drawn but if you're trying to say voltage or amperes there's a really big difference here and it's really freaking important. If your CPU or GPU draw more amps under load that's normal, but if your CPU suddenly starts drawing more volts that's a really big problem since your vcore is supposed to stay with what it's set at in the bios. Same with your GPU, you don't want that thing basically overvolting itself and causing problems (visual artifacts being one of those).
Words have meaning Jenn, in 10 years of repairing computers and all that experience with power supplies you should have learned that, along with learning the most basic and fundamental facts about the components you're talking about. Especially when making statements like this where you're dealing with things where the distinctions are important and you get those most basic and fundamental of facts utterly wrong.
And once again, because this really does need to be continually pointed out since you apparently are trying to obfuscate the issue to hide this on purpose, a "spike" of about two and a half watts from a USB powered mouse is going to be a damn sight less than the "spike" of any of those others except for his RAM.
Quote:
You mean the adapter on the *outside* of the video card? Of course it doesn't 'care' at the termination point, because it's already done with signal processing.
You just made my point, so I'm not sure what the rambling bit above was about.
|
See now I think you're just playing games, you're basically contradicting yourself here. Lets go back to your original post:
Quote:
|
The intermittent nature of it makes me think it's even more likely, and the fact that you've got dual output is an even bigger draw on power. VGA (D-SUB) takes less GPU cycles than does DVI, and would require less V, strictly speaking.
|
to which I responded:
Quote:
|
Again: No. Ever see those dvi->dsub adapters that you just screw onto the end of the monitor cord? Your videocard does not know or care the difference when you use those, the only time it notices a difference is when using a DVI-D connection and says "Yo, I'm all-digital, stop running the signal through the DAC and save yourself some work" and then it's DVI that's easier since it doesn't need to convert it to an analog signal.
|
Now how the hell did you do this? First you claim that the connector that
doesn't need to run the DAC uses more power, then when I point out both possible situations of why that claim is ridiculous you respond first by acknowledging that you do know about signal processing and then by claiming you have no clue what I'm "rambling about" despite
acknowledging signal processing a few lines up.
Quote:
|
I could have left this alone, but really? You do realize that W = VA, where Watts = Voltage * Amperage? It's the fundamental measure of current. If you're trying to 'correct' me to 'ampere', you'd do well to recognize that amperage is the measure of current, measured *IN* amperes.
|
It's an electrical joke from same style as this (abridged) chemistry joke:
"... and it's so stable you could put another few nitrogens on it."
"No you can't."
Or alternatively flatly denying that the Star Wars prequels exist because George Lucas died tragically of a stroke in the mid 90's.
Quote:
|
I'm quite familiar with what insulation is for, thank you. And as you so dutifully noted above, the actual current carried on the line is not enough to create any meaningful EM field to nearby cables, even with damaged shielding.
|
You realise that by admitting that you contradict your entire argument that the mouse is pushing his PSU over the edge, right? If you admit that the power draw of the mouse is that negligible then you're admitting that if the margins here were so close that said negligible power draw pushed his PSU over the edge then any number of much more power hungry components would have done worse already.
Quote:
|
I'd wager Real Money that the OP's problem will go away if you test with a higher rated PSU, but feel free to blame.. what was it you think is the failure, again?
|
Any number of things starting with plain bad cables and possibly some interference since he's also got an issue with the screen being wavy, which suggests to me he isn't getting a digital signal since that sort of wavy issue tends to be more what you get with a bad analog signal.
Whatever it is, and it may be PSU related I never said it wasn't, it is not specifically that his mouse is somehow pushing his computer all two watts or so past it's limit and yet somehow the CPU and GPU magically aren't.
But since you're probably going to take the wall-of-text war even further and continue to try and find ways to twist things to mean what you want them to or dance out from this how about a simple test that will absolutely prove whether you're right or not: run prime95 and ATItool's "scan for artifacts" to stress both the CPU and GPU to the limit. If the monitor doesn't shut off then it's not an issue of how much power is able to be supplied because running both of those at max will absolutely outweigh the power draw of moving the mouse while loading a page or playing a video. After that just troubleshoot the mouse in different ports, or if you've got one try a non-USB mouse (assuming you have a USB mouse to begin with).
Those actions will test every factor of your theory that somehow the mouse is causing problems with the power supply: It tests the mouse, it tests the ports, and it tests whether the power supply is actually near its limit or not.