Typical mains power for mid-range PC?

On Mon, 31 Jul 2006 03:05:46 -0700, "Ed Light" Gave us:

You still seemed to miss the fact that you used more than the supply is rated at in your example. :-]

Your assessment of my intelligence is probably as inaccurate as mine of yours..

However, IMHO, relationships do best when both parties are totally convinced that they are each unworthy of the other - and fail when they prove that they were right.

Also, as to propensity for crime between the sexes - remember who did the first one.

I have actually met far more intelligent men than intelligent women, but, perhaps, that came with the territory.

Still, men never learn - Lucrezia Borgia's husbands probably thought she was intelligent...

OK, the Seasonic 330w.

Add up 12v 1 and 2 and you get 22 amps.

Of course if it actually had to do that it would limit its output on the 5 and 3.3 a bit, like you said.

In this test

They made it put out, it appears, 330 watts consisting of 19A of 12v, 11.5A of 5v, 12A of 3v, and those little ones, at 78% efficiency, temp. rise of

8C.

On Mon, 31 Jul 2006 04:17:20 -0700, "Ed Light" Gave us:

No ripple or regulation spec?

Thro the board vias & indeed thermal isolation from HD casing, heat conducts to the air - rough surface gives more surface area. Same reason P-911 air-cooled engines have a textured finish.

Seagate U6 were not a reliable drive - cheap, but not reliable. The rubber was to stop them marking the wall too severely.

Backups :-)

Jon D wrote

Yep.

Nope, it was there to make the drive quieter.

Only that it wont be as quiet.

While some PSU are more efficient when nearer their full load than the 500W would be, there are other aspects beyond power usage that will effect PSU longevity.

With the higher wattage PSU, if designed appropriate to it's wattage same as the 250W would be, it will tend to have at least a few of the following features;

- Larger capacitors, and/or higher quality (if they need to fit in a limited amount of space).

- Thicker heatsinks

- Larger die BJTs, diodes, for handling the higher current, and with 250W PSU in mind, may be a larger component package size like TO247 instead of TO220.

These aspects spread out the per-descrete-component thermal density so regardless of whether the whole PSU is more or less efficient, the individual parts that run hottest, are running cooler.

The larger or higher quality capacitors in the larger PSU may also help reduce ripple, but it can depend a lot on the quality of the 250W PSU, too, as some 250W are low end even from quality manufacturers and others are using reasonble quality parts to fill up the available space over much of the PCB, merely a smaller transformer and high side, HV DC filter caps. Seldom do aftermarket PSU get designed this well though, as it results in the lower capacity unit being not that much cheaper. Industrial PSU, or those for older workstations and servers are some market segments where the HQ lower wattage PSUs usually appeared... in past years.

No, most definitely not. As an example I had two HDDs out last night, lying upside down on a desk. One was a maxtor 40GB, the other a Samsung

250GB. Their frames were moderately warm but I touch-tested several parts on the circuit boards. Several chips were significantly hotter than the frame.

Granted you wrote "PCB", not "Chips" but just as with the idea about cooling the sides of the drive frame ultimately cooling the rest of the frame more too, likewise cooling the PCB more will cool the chips more.

Yes, and it fails at it because it can't cool the chips on the PCB. The drive still needs an amount of airflow to cool them that is sufficient for cooling the drive frame too. IF the chips had individual heatsinks, "maybe" the situation would change, the chips would then need a lower, possibly low enough aiflow rate that this rate would be insufficient to cool the frame too but we can ignore this since drives don't typically have 'sinks on the chips.

Yes, I'd already acknowledged it would be better than nothing, but have you taken a good look at the sides of modern drives? Not only are they not flat, they deliberately machine them so the areas where the mounting screws are tapped out, are raised above the rest. There is not a good thermal path from the hottest areas of the drive to the passive cooling product, these hottest areas will still run quite warm with this expensive passive cooler relative to even a low flow, quiet fan.

Not always what? It is fairly easy for them to be cheaper, and to work better.

The key to cooling is to overcome the thermal gradient, have the most expediant path from high heat density to lowest.

For example, take a CPU and put a high quality heatsink with large surface area directly over it. Now compare that with taking same heatsink but a low-surface-area, flatted spherical chunk of aluminum 2" thick between the CPU and the heatsink. Which will run the CPU cooler if the rest of the environment, including fan, is the same? The first will.

Would you buy a poor case and then expensive ineffective passive coolers for every drive? Any sane person wouldn't, and either way, the passive cooler does almost nothing to cool the chips on the PCB.

I'm sure they'd claim it was, and in doing so they'd reinforce the other aspect of implementing it, that it's a marketing gimmick. User sees drive has a thing added that the others don't have, assumes this makes it better else why would it have it. Like parachute pants, some trends come and go.

What matters is whether those temps were outside what is acceptible.

But they dont necessarily need cooling more.

Doesnt necessarily fail at all.

Have fun explaining how a drive can stay at a perfectly reasonable temp without any deliberate airflow at all.

But they do normally have chips that dont get that warm even without deliberate airflow, particularly when there is significant conduction of heat to the metal drive bay stack.

Bullshit they are.

But may well still be at a perfectly adequate temperature.

Only necessary if the drive gets too hot without that.

CPUs are completely irrelevant, they have a MUCH higher power density than modern mass market hard drives do.

Irrelevant to what is necessary with a hard drive.

They dont necessarily need any cooling, just like most of the chips on the motherboard dont either.

kony wrote

It wasnt a marketing gimmick, those drives were one of the quietest drives around at that time.

Irrelevant to whether those drives were one of the quietest drives around at that time.

Yes, Seagate doesnt bother so much about quiet drives anymore.

Others still do tho, and those who want quiet drives buy those instead.

Would ripple be the voltage variation at a constant output? No, they don't show that. They do show variations between various output levels. 11.82 to 11.95, for instance.

On Mon, 31 Jul 2006 16:57:35 GMT, "Dorothy Bradbury" Gave us:

It is for IR emissivity gains, and the minor increase in surface are is not the reason. It has to do with IR rays diffracting upon contact with the surface texture from within the medium. That is how emissivity works.

A perfectly smooth surface like a mirror or polished steel surface would stay hot when heated longer due to a LOT of IR energy reflecting back down into the medium as opposed to "radiating". A matte finish on nearly any medium will significantly increase it's IR emissivity, not due to surface area increases, but due to the way IR energy operates as it reaches the surface of a hot medium. Hitting that matte finish diffuses the IR and allows it to radiate in many more directions. Far more efficient.

Here is a good test. Take a teflon coated aluminum skillet. Get it hot on the range top (at least 150F or so too hot to touch).

Pick it up by the handle and turn it upside down and bring the bottom surface near your cheek. Feel the heat radiating. Now turn it around so the teflon surface points at your cheek. You should feel a notable increase in how "hot" the radiation "feels" to your cheek. This is due to the difference in the IR emissivity of the surfaces.

The type of teflon coating they use is very dull finished, and they even "mottle" it, and it has micro-features that let it radiate better still.

Aluminum ass of skillet after heat soak: e= 0.62 (maybe)

Same aluminum pan with same heat soak on teflon side: e= 0.95 or better.

On Mon, 31 Jul 2006 16:57:35 GMT, "Dorothy Bradbury" Gave us:

Hahaha... skid marks instead?

On Mon, 31 Jul 2006 17:19:41 -0400, kony Gave us:

They are switchers.

On Mon, 31 Jul 2006 17:41:08 -0400, kony Gave us:

You means these things won't work!!!?

D A M N I T !!! ;-]

On Tue, 1 Aug 2006 10:09:36 +1000, "Rod Speed" Gave us:

Bullshit. They ARE the quietest drives around.

On Mon, 31 Jul 2006 20:58:28 -0700, "Ed Light" Gave us:

The spec is measured throughout the power range of a given rail while the other rails are at their rated output. As the rated output of the rail being tested is approached, a good supply will still have low ripple. A shit supply will be very high in ripple at rated output, which cancels the validity of the rating.

It is very similar to THD ratings of audio amplifiers.

The ATX12V spec for PC supplies is very specific as to what is acceptable. The lesser form factor specs are as well.

There's still a transformer.