Amperage pull of a computer?

Forgive the top posting. Seems to me you and I have been on the same page but NG communications being what they are, misinterpretation runs rampant so some of us are anal in our explanations to avoid confusion, myself included.

About the amp clamp. This one is very accurate down to about 2A so I believe the readings I got were pretty darn close.

In order to draw the maximum current I could, I had every peripheral available running at full tilt...CD drives (all of them), sound system on max, shredder running....everything. Obviousely this is not normal and I'd hazard a guess that at standby, my entire office doesn't draw but 4-5 amps with the lights on.

On Mon, 26 Jan 2004 09:29:39 -0500, Keith R. Williams Gave us:

Eat shit and die, you troll f*ck retard!

YADTAAUW

Yet Another Dipshit That Assumes That All Use WindBlows.

Get a clue.

You couldn't be more wrong.

You are an idiot. The halt instruction is per process. That means that THAT process passes its time slice back to the CPU. There are several processes that always rune regardless of what your retarded ass thinks.

SAID CPU IS ALWAYS processing instructions, and said CPU is also always using the same current in operation. Average of 50% high bits,

MOBILE CPUs have slow down modes that reduce power consumption.

That would be Apples and Oranges, dipshit.

No. The FPU clock multiplier is stopped. The clock at baseline still runs.

No. It changes when it is turned off by the system as in a power saving mode, a totally different circumstance.

Desktop PC need no such facility, and rarely use it if available. Again, Apples and Oranges.

Bwuahahahaha!

Neither are GUI OSes. The wattage consumed on a PC varies little to none between idle and fully loaded work (CPU not peripherals). The ONLY time it is less, is \when it is put to sleep, which is not often on a PC that actually IS being used.

If my PC sits for an hour, the cpu is the same temp as it is when two seti units are being crunched down, and I am playing a full Open GL game app, and DLing a new Knoppix ISO in the BG.

Temperature is the perfect indicator of activity. No change means that the cpu uses the same whether doing work, or not.

Got Clue?

On Mon, 26 Jan 2004 14:51:59 GMT, "Anthony Fremont" Gave us:

It does no such thing.

It (they) run at 143F from 6 minutes after boot till it gets shut down, even if idle.

On Mon, 26 Jan 2004 14:51:59 GMT, "Anthony Fremont" Gave us:

Such calls are per process.

I am from Missouri.

On Mon, 26 Jan 2004 10:10:03 -0500, Jaxon Bridge Gave us:

Wherewithal? Do you have the wherewithal to learn about Usenet, and refrain from being a top posting Usenet retard?

LOOK UP the stats for YOUR hard drive. Write it down.

Then, your CD ROM drive.

Then, your MOBO, Vid Card, net card, etc. for all connected peripherals.

Add it all up. DOH! Add a bit for PS inefficiency.

| Why you hang non-essential appliances like a shredder and "halo" | lamps on a UPS is beyond the grasp of ny logical man. In a power | outage, they could easily be added, if one really needed to operate | them in such circumstances.

Actually, I put a low wattage lamp on my UPS so I would not be in the dark. I used to not even put my monitor on the UPS, but then I had an outage and couldn't get my machine shutdown because I couldn't see.

| Did you break the line cord open, and place the meter on one line | only?

Don't do that. Buy a separate very short extension cord for the purpose. Break it open and then tape the neutral and ground back together, leaving the hot wire open. Of course you might also want to measure ground current as a safety check, too (if not using a GFI circuit/outlet).

| You are an idiot. The halt instruction is per process. That means | that THAT process passes its time slice back to the CPU. There are | several processes that always rune regardless of what your retarded | ass thinks.

The halt instruction is privileged. If a process could halt the CPU, that would be incredibly insecure. I once found a bug in IBM's big mainframe OS called OS/VS1 that let me do that from a batch job. But that was the college days. Processes that have nothing to do, or are quitting, have system calls to do that (different between Windows and Unix, but same concept).

| SAID CPU IS ALWAYS processing instructions, and said CPU is also | always using the same current in operation. Average of 50% high bits, | 50% low bits. Get a clue.

Gates! ... no not Bill. The number of gates changing state per second matters (and not all are equal). Floating point ops (and SETI does a lot of those) are the most intensive. It doesn't matter if the bits are 0 or 1; it's the changing the matters. Try increasing or decreasing the clock rate to see.

Running SETI really does use more electricity, but not a lot more in comparison to the whole machine. The figure I heard is about 1 to 2 watts over a totally idle machine in the 1 GHz Pentium III class.

The SETI project recommended people NOT run SETI work in California during the power crisis there a few years ago. But the issue was not how much SETI uses, but rather, the fact that it encouraged people to leave their computers on.

| MOBILE CPUs have slow down modes that reduce power consumption.

Like I say, gate state changes per second, approximately.

| No. The FPU clock multiplier is stopped. The clock at baseline | still runs.

But not all the gates are working every cycle. It depends on what operations are taking place.

| No. It changes when it is turned off by the system as in a power | saving mode, a totally different circumstance.

Power saving mode helps, but it's not zero.

| Neither are GUI OSes. The wattage consumed on a PC varies little to | none between idle and fully loaded work (CPU not peripherals). The | ONLY time it is less, is \when it is put to sleep, which is not often | on a PC that actually IS being used.

The largest swing in PC power usage is when heavily seeking a disk. That can increase power usage by as much as 10 watts, and make the drive as much as 20C to 40C hotter.

| If my PC sits for an hour, the cpu is the same temp as it is when | two seti units are being crunched down, and I am playing a full Open | GL game app, and DLing a new Knoppix ISO in the BG. | | Temperature is the perfect indicator of activity. No change means | that the cpu uses the same whether doing work, or not.

Motherboard CPU temp measurements are inaccurate. They also take a long time to see change, and while the changes are real, they are small. You could easily have missed it. But it is not so large as the previous poster is making it out to be.

| If you want something more accurate, you'd probably need to buy a | wattmeter.

Pretty much, yes.

Not sure where this comes from, but it makes some sense.

...which is why I made the comment about "backup supplies". It's current (VA) that matters to the source, and Watts to the load. When the two don't match, someone is going to get "mad". ;-)

I had my Fluke out today and measured .11 ohms on the leads...that would be .01 ohms more than your suggested .1 resistor.....And the digital accuracy of the 8062A....So it seems that my DMM would be just as accurate...just a comment here....

I wish I had more time to participate in this one!....Even though I am retired, I still do a little consulting in Silicon Valley and took on a job for most of the week....but it is tempting!!!

This time of year that's spelled "Misery"!!

I think this just boils down to what accuracy the OP was looking for....He wanted an approximation because, I assume, he had breakers tripping in an old building. If he can throw an approximate number on every device, he should easily be able to decide what circuits to connect and what combination of devices....1 amp one way or the other should not make a difference! I will have to check my system too....now you got my interest up!....take care....

Just top post to DarkMatter and it REALLY irritates him...I would call it a guaranteed flame!...He is a self proclaimed Usenet Cop....and has still not figured out how to ask nicely.....I hope the posts have helped you solve your problem...take care....Ross

You might have missed the idea DarkMatter posted about using a resistor - it is buried in the posts. It is the best answer in the thread to give an easy way to determine amperage use on your own.

It will let you determine the amperage used by any of your devices with pretty good accuracy and at low cost.

They're doing their processing but they need not be spinlooping.

IIRC, there's a HALT buried way down deep in some Windows VXD.

Most multitaksing OSes have their equivalent.

OTOH, if you're still running DOS....

Definitely, that's how SpeedStep works.

The Pentium 4 CPU has an on-die thermal sensor and it will change significantly in seconds.

By it's measurement (Intel says it's an approximate measurement) I can run a spinloop for couple of minutes and it will go from 95 to 111 degrees F.

As was said, "Temperature is the perfect indicator of activity." By that rule, running a spinloop results in more activity at least on my machine.

You are assuming he has a DMM to finish the "low cost" job....Not to mention the ability to do the conversions... An amp clamp is a simple and easy solution for an "approximation" that he was looking for...After all, this was not a current study on his PC but just a power problem he has in an old building with several pieces of hardware.

For some reason, that made me think of "Green Acres" even though I haven't watched re-runs of it in years.

Brings back some old memories......