Bit of wire as good as an antistatic strap?

[kony]

Anything since a Pentium I has a problem - instead of asking PS makers to add
reliable1.2, 1.5, 2 volts, the idiot ATX standard calls for the lower voltages
to be produced by on-board regulators which 1) throw off half the waste heat in
the box and 2) leave large caps on the board that are not self-draining.

If the power supply had the lower voltages, it would have to
produce much higher amperage and much higher loss on the
long cables... it would be the poorer of the two
alternatives. Further, the power supply doesn't "know" the
exact voltage needed for a particular memory module,
northbridge, or CPU... at most it would get closer to the
right voltage but the board would still have to reduce that
voltage further- else an even more elaborate system of
digital logic to the power supply to set those voltages,
plus an additional wire to take readings, make adjustments-
again difficult because of the wire lengths involved and the
(relative) precision necessary - not a good thing to have a
1.3V CPU off by .2 volts.

Plus, ATX PSU are already pushing reasonable limits for
power density, squeezing even more parts into one, or rather
parts to accomodate additional functions unrelated to their
present design goals, would be even more difficult.

They do not "throw" half th waste heat into the box
currently. It would be quite impossible for the small
copper area the mosfets and inductors are mounted to, to
'sink that much heat.

What do you find problematic about the (supposed) lack of
self-draining caps? Take out a board, power it up then off
and tell us how long the caps retain any (reasonable
proportion of their regular) charge. This just isn't a
problem.

Additionally, many cheap power supplies (why I insist on PC Power&Cooling though
I don't get anything for it but higher bills) do not install bleeder resistors
across the 100 MFD high-voltage caps necessary in a modern switcher.

It would have to be an incredibly poor and low capacity
power supply to only have 100 mfd high-side caps, even the
incredibly poor junk ones are in the 330 mfd range.
Regardless, every PSU I've seen has one if not mutiple paths
to drain those caps. PCPower&Cooling makes good PSU, but
these simple (small, and very cheap to implement) things are
not usually the difference.

That's why mobo makers started putting LEDs on boards and telling you to unplug
and then turn "on" until the led goes out.

More likely it was simply to alert user that the PSU was
still plugged into AC, the LED will go out (as the power
drains) within a few seconds- less than 10, usually less
than 5.

 

[CBFalconer]

... but that's secondary.


Are you taking into account that your machine is connected to the
other machines on your LAN, and they are tied to Earth ground.

I hope not, because I would expect the network paths to be
electrically isolated from the nodes. And it is not the volts that
cause problems, but the charge. The volts reflect the density of
the charge, and provide the impetus to move it, creating current,
and attendant damage. Windhurst and Van de Graaf machines work by
moving charges around.

--
"Our enemies are innovative and resourceful, and so are we.
They never stop thinking about new ways to harm our country
and our people, and neither do we." -- G. W. Bush.
"The people can always be brought to the bidding of the
leaders. All you have to do is tell them they are being
attacked and denounce the pacifists for lack of patriotism
and exposing the country to danger. It works the same way
in any country." --Hermann Goering.

 

[Ralph W. Phillips]

Howdy!

Are you taking into account that your machine is connected to the
other machines on your LAN, and they are tied to Earth ground.

Are you taking into account that there's ground isolation on the LAN
links? And not always are all computers connected?

And in what way does that contradict what I said? B) I mean, if
you connect to the case, you're still at the same ground potential as the
CASE. Be it at absolute 0 V, or at 1 MEV ...

And it's being at the same relative potential that's important. NOT
the absolute potential.

Consider also - how about the birds that sit on a high tension phone
line? And fly off later? B)

RwP

 

[David Maynard]

I think that's the balast/inverter for the flourescent lamp that back-lights the
panel, nothing to do with the panel itslef.

The "panel itself" is of little use without illumination.

 

[David Maynard]

Anything since a Pentium I has a problem - instead of asking PS makers to add
reliable1.2, 1.5, 2 volts, the idiot ATX standard calls for the lower voltages
to be produced by on-board regulators

The co called 'idiots' spec it that way because that's the best place to do
it for just about every conceivable reason.

And besides all the electrical reasons, you'd obsolete every ATX PSU on the
market each time VRM requirements changed. Your "Pentium I" PSU would be
useless for an MMX, which would be useless for a P-II, which would be
useless for a P-III, which would be useless for a P4, and that doesn't
count AMD.

which 1) throw off half the waste heat in
the box

Nonsense. On-board regulators operate at well over 90% efficiencies and are
inconsequential to the overall case thermal load, and with potentially 200
to 400 watts dissipated in the case you're complaining about maybe 4 or 5
from the regulators.

and 2) leave large caps on the board that are not self-draining.

'Bleeder' resistors are unnecessary with circuits that naturally bleed of
in a couple of seconds.

Additionally, many cheap power supplies (why I insist on PC Power&Cooling though
I don't get anything for it but higher bills) do not install bleeder resistors
across the 100 MFD high-voltage caps necessary in a modern switcher.

That's why mobo makers started putting LEDs on boards and telling you to unplug
and then turn "on" until the led goes out.

I don't know which particular motherboard and, so, which LED but they're
typically the standby 5V indicator. Which has nothing to do with your
previous 'complaints'. And even then it'll bleed off in 4 or 5 seconds.
It's simply not an issue.

 

[David Maynard]

Of course it is when operating! when a "system" consists of a bare mobo, empty
case and assorted other bits and pieces, it is difficult to assume a system
ground.

Once a system is up and running, it is virtually in a Faraday Cage and a nearby
lightning bolt will have to put out one hell of an EMP to effect it.

That would have a chance of being true if it weren't for all those lovely
conductive wires going in and out of it.

 

[David Maynard]

Howdy!




Are you taking into account that there's ground isolation on the LAN
links? And not always are all computers connected?

And in what way does that contradict what I said? B) I mean, if
you connect to the case, you're still at the same ground potential as the
CASE. Be it at absolute 0 V, or at 1 MEV ...

And it's being at the same relative potential that's important. NOT
the absolute potential.

Consider also - how about the birds that sit on a high tension phone
line? And fly off later? B)

RwP

The ground strap on an airplance is a real bitch.

 

[Kinell]

Once a system is up and running, it is virtually in a Faraday
Cage and a nearby lightning bolt will have to put out one hell
of an EMP to effect it.

I'm sure w_tom would have something to say about that ;-)

 

[AT]

AT wrote:

That would have a chance of being true if it weren't for all those lovely
conductive wires going in and out of it.

You are right, I should have been discussing the box, if segregated from
everything else except a good ground, or just sitting isolated on rubber feet,
and not connected to nothing.

I just get bugged sometimes at people installing everything with the greatest of
care, conductive carbon-fiber gloves @$100 per, etc, and then either a)
continuing to be afraid of operating a proper system if it might rain or b) just
plugging it in to an untested outlet.

Both extremes bug me, and I should have better explained what I meant and left
Faraday out of it.

<the concept of a faraday box is the stuff *inside* a grounded or isolated box
is safe from even intense fields around it - why the safest place to be in a
thunderstorm on a wide-open stretch of road is in your car, cushioned and
isolated by an additional layer of plastic/leather/whatever from the steel
around you. Probably even enough metal in a Corvette to do the job.>

Tr Al

 

[AT]

And the suicidal terrorist squirrils who take out neighborhoods every year by
stepping on a hot and ground line - especially near a transformer! (no, do not
try to remove for a quick meal, even if the hind quarters, only edible part on
the things, isn't burned to a crisp, you WON'T like it, and, even if you are
into road-kill rodent souffle, you might go the way of the tree-rat yourself.

TrAl

 

[Ed Medlin]

You are right, I should have been discussing the box, if segregated from
everything else except a good ground, or just sitting isolated on rubber
feet,
and not connected to nothing.

I just get bugged sometimes at people installing everything with the
greatest of
care, conductive carbon-fiber gloves @$100 per, etc, and then either a)
continuing to be afraid of operating a proper system if it might rain or
b) just
plugging it in to an untested outlet.

Agreed. One easy way to assure grounding that is cheap too, is to cut the
two power plugs off a power cable and plug the system in. Most of us have a
few of those around if you have been working on these things for awhile.
This, of course, assumes your outlet and home ground is good. All you have
to do is touch the case to assure your body is discharged and you are fine.
As far as gloves are concerned, never used them and never saw them used in
any shops around here. Ground straps are sometimes mandatory, but by no
means in every shop.

Ed

 

[Fisher]

YES and no - the metal's OK, but attach the strap to the GROUND line of the
power/mains *after* assuming the house wiriing is right and you are not about to
incinerate yourself good. Attach any ground strap human-to-case does NOTHING.
Attaching both to a good GROUND works.

This is exactly what I said and I got called a fool for it.

 

[Ed Cregger]

This is exactly what I said and I got called a fool for it.

Attaching oneself to the case (unplugged/ungrounded) equals the charge
between the operator and the device, hence, no current can flow, thus no
damaging discharge through sensitive devices. That is all we were saying.

If you want to ground yourself, feel free, but don't be surprised if
something not properly grounded in the area knocks you on your butt, should
you touch it.

Ed Cregger

 

[kony]

Attaching oneself to the case (unplugged/ungrounded) equals the charge
between the operator and the device, hence, no current can flow, thus no
damaging discharge through sensitive devices. That is all we were saying.

THe problem is that there is not only the operator and the
device, there is an entire world around them so having both
at same potential does nothing to prevent discharge, current
flow, through the components to a closer-ground object like
a desk or (whatever). Attaching oneself to the device is
only of limited usefulness since common parts are put in or
removed from such device.

 

[Ed Cregger]

THe problem is that there is not only the operator and the
device, there is an entire world around them so having both
at same potential does nothing to prevent discharge, current
flow, through the components to a closer-ground object like
a desk or (whatever). Attaching oneself to the device is
only of limited usefulness since common parts are put in or
removed from such device.

I'll think about this one for a while before answering. If I answer. Good
point.

Ed Cregger

 

[David Maynard]

THe problem is that there is not only the operator and the
device, there is an entire world around them so having both
at same potential does nothing to prevent discharge, current
flow, through the components to a closer-ground object like
a desk or (whatever). Attaching oneself to the device is
only of limited usefulness since common parts are put in or
removed from such device.

Connecting one's self, and the case, to 'ground' doesn't address that
situation either. All that does is ensure that any object with a static
charge prefers shorting to 'ground' through the equipment, or the person.

Part of the problem is people are talking about where to 'connect' the
'person' and case (I.E. to each other or both to 'ground') but the thing is
to create an equal potential *workspace*.

Parts entering or leaving the equal potential workspace are to be in ESD
protective containers and brought to potential before opening (as in set on
the work space conductive mat).

Safety grounds are a completely different issue.