ESD

In the GTS book it says you MUST unplug the Computer and clip your
anti-static band to it, to balance out the voltage in you and on the
computer. Generic advice seems to be to leave the computer plugged in
but switched off to ground the voltage in you and in the computer I am
quite willing to go with "computer must be unplugged and voltages
balanced" for the sake of any test. I was just wondering if they say
computer must be unplugged as Americans usually don't have switched
sockets. Then you rely on the switch working to shield you from 240
volts, but any voltage between you and the computer can be leaked away.

soup

***IF*** the computer has a true master power switch, leaving it plugged
in but turned off (MASTER switch (usually on the back of the power
supply) is better, because the power cord grounds the chassis (and you,
if you are using a wrist strap). However, many computers (especially
models made by the large OEMs) do not have a master power switch, and
the only way to truly turn them COMPLETELY off is to unplug them.

The issue is not the AC power line voltage (in the US, typically 115-120
volts), but the 5 volts standby, and sometimes 3.3 volts, that is still
present when a modern computer is "shut down".


soup wrote:
> In the GTS book it says you MUST unplug the Computer and clip your
> anti-static band to it, to balance out the voltage in you and on the
> computer. Generic advice seems to be to leave the computer plugged in
> but switched off to ground the voltage in you and in the computer I am
> quite willing to go with "computer must be unplugged and voltages
> balanced" for the sake of any test. I was just wondering if they say
> computer must be unplugged as Americans usually don't have switched
> sockets. Then you rely on the switch working to shield you from 240
> volts, but any voltage between you and the computer can be leaked away.

Barry Watzman

Barry Watzman wrote:
> ***IF*** the computer has a true master power switch, leaving it plugged
> in but turned off (MASTER switch (usually on the back of the power
> supply) is better,
> The issue is not the AC power line voltage (in the US, typically 115-120
> volts), but the 5 volts standby, and sometimes 3.3 volts, that is still
> present when a modern computer is "shut down".

I am not talking about the switch on the computer but the socket on the
wall usually in the UK they are switched but in the US they are usually
unswitched .

I would never leave the power cord plugged in to an unswitched socket as
standby voltages can still be present even if the computer's power
supply is switched off.

soup

Re: "I would never leave the power cord plugged in to an unswitched
socket as standby voltages can still be present even if the computer's
power supply is switched off."

Not if the computer is turned off with a switch on the REAR of the power
supply itself. That switch is the same as unplugging the computer.


soup wrote:
> Barry Watzman wrote:
>> ***IF*** the computer has a true master power switch, leaving it
>> plugged in but turned off (MASTER switch (usually on the back of the
>> power supply) is better, The issue is not the AC power line voltage
>> (in the US, typically 115-120 volts), but the 5 volts standby, and
>> sometimes 3.3 volts, that is still present when a modern computer is
>> "shut down".
>
> I am not talking about the switch on the computer but the socket on the
> wall usually in the UK they are switched but in the US they are usually
> unswitched .
>
> I would never leave the power cord plugged in to an unswitched socket as
> standby voltages can still be present even if the computer's power
> supply is switched off.

Barry Watzman

Barry Watzman wrote:

> Not if the computer is turned off with a switch on the REAR of the power
> supply itself. That switch is the same as unplugging the computer.

I must have misunderstood what switching the computer off meant (i.e.
the switch on the computer rather than the switch on the power supply)
The piece I had read was:-

•An ATX power supply provides soft power . This is a condition where
the motherboard always has power even when the computer is turned off.

I must have mistaken that *computer is turned off* to mean the switch on
the power supply rather than the switch on the computer.


The issue of the switches being faulty still would worry[1] me . Is
this balancing the charge (ie you at 100 volts the chassis of the
computer you are working at also at 100 volts(note, not real figures)),
rather than earthing the entire system, theory a goer?

[1] 'Worry' is maybe too strong a word 'thoughtful off' too weak
probably something in between (can you tell I am no linguist?).

soup

soup wrote:
> Barry Watzman wrote:
>
>> Not if the computer is turned off with a switch on the REAR of the
>> power supply itself. That switch is the same as unplugging the computer.

Except that the ground is still there (the chassis is
still connected to a zero volt reference--generally
a ground rod or cold water pipe somewhere).

> I must have misunderstood what switching the computer off meant (i.e.
> the switch on the computer rather than the switch on the power supply)
> The piece I had read was:-
>
> •An ATX power supply provides soft power . This is a condition where
> the motherboard always has power even when the computer is turned off.
>
> I must have mistaken that *computer is turned off* to mean the switch on
> the power supply rather than the switch on the computer.
>
>
> The issue of the switches being faulty still would worry[1] me .

There's often an LED on the motherboard to indicate
standby power is still present and that you shouldn't
be working inside the case. If there's no LED or you
suspect that it or the power supply switch is faulty, the
routine should be to either check for the presence of
the standby voltage (+5 VDC between the 20/24 pin connector
purple wire and ground) or unplug the AC cable at the
power supply and connect a clip lead between the grounded
connector on the AC cable and a bare metal point on the
case/chassis. That same point or area is where you should
also connect your wrist strap (if you believe in using
one--which is another debate). This is a good procedure
in situations where the power supply doesn't have a switch
(assuming you believe that connecting the chassis to the
reference is necessary--more on that below).

> Is
> this balancing the charge (ie you at 100 volts the chassis of the
> computer you are working at also at 100 volts(note, not real figures)),
> rather than earthing the entire system, theory a goer?

The point is to establish and maintain a voltage
reference. The common way to do that is to start
with a patch of ground (or Earth) near the building.
That's done through contact between a piece of metal
with a relatively large surface area and the ground
itself. That contact area is said to be at zero volts
DC (zero frequency). Since it has no frequency the
extra considerations that apply to AC don't apply.
To put it simply, this zero volt reference can be
transported with only the resistance of the conductor
to raise it above zero. So, if we use heavy conductors
relative to the expected amount of current flow--which
should be zero unless there's a fault which should
only be momentary as ground faults trip protection
circuits that quickly stop the flow of current, we'll
have a close to zero volt reference anywhere along any
of the conductors in the circuit.

When it comes to computer/electronic servicing the
zero volt reference is a convenience more than it really
is a necessity. For safety and charge balancing purposes
something has to be the reference. It doesn't have to
be a patch of ground near the building, however that's
what has been settled on.

> [1] 'Worry' is maybe too strong a word 'thoughtful off' too weak
> probably something in between (can you tell I am no linguist?).

What's more confusing than the reference when considering
ESD is the purpose of the megohm resistors in the wrist
straps and other ESD devices. The idea is the resistors
limit the current so that the person wearing them doesn't
get shocked if they contact a source of high voltage
relative to the reference (which is supposed to be zero).
The problem is that allows the person to be at a higher
potential than everything else that is connected directly
to the reference (like the computer). So, using the
common ESD devices is of no value when it comes to
preventing ESD damage to computer/electronic components.
Further, if the person touches the chassis or anything
else that is connected to the reference while in contact
with a source of high voltage they will still be shocked.

When you really think about it logically the point really
is simply charge balancing between you and the chassis
of the computer you're working on. At the A+ level where
all that is done is removing and replacing block-level
devices the technician can simply unplug the AC cable at
the power supply and physically touch the bare metal
chassis to balance the charge between himself and the
chassis. Handle devices by the places that connect to the
chassis (in the case of processors and other semiconductors,
handle them by their edges and don't touch any of the
electrical contacts) and touch the chassis occasionally
to balance any static charge built up by walking around
and handling other objects. That's really all there is
to the ESD thing at the A+ level.
--

Bill Eitner

The ground still being there is BENEFICIAL. You WANT the ground to be
there. The reason you wear a wrist strap is so that YOU are at ground
potential. For this to be as effective as possible, the computer should
also be at ground potential, which is best achieved if it, too, is grounded.

Re: "The issue of the switches being faulty still would worry[1] me"

and

"If there's no LED or you suspect that it or the power supply switch is
faulty, the routine should be to either check for the presence of the
standby voltage (+5 VDC between the 20/24 pin connector purple wire and
ground) or ....."

Look, if you are worried that the computer might still be active, just
push the power switch (power button)(on the computer), as if you were
trying to turn it on. If it doesn't turn on ..... then you know that
the power really is killed. No need to start measuring things, although
the motherboard standby power warning light is also a perfectly good
indicator of the power supply status, except that not all motherboards
have such an LED.

Barry Watzman

Barry Watzman wrote:
> The ground still being there is BENEFICIAL.

Why?

> You WANT the ground to be
> there.

Why?

> The reason you wear a wrist strap is so that YOU are at ground
> potential.

No, you're not--you're at a megohm above ground.
That's a MILLION ohms above ground. You're at
every potential EXCEPT GROUND.

> For this to be as effective as possible, the computer should
> also be at ground potential, which is best achieved if it, too, is
> grounded.

For what to be as effective as possible?

You're babbling.

> Re: "The issue of the switches being faulty still would worry[1] me"

It wouldn't worry me.

> and
>
> "If there's no LED or you suspect that it or the power supply switch is
> faulty, the routine should be to either check for the presence of the
> standby voltage (+5 VDC between the 20/24 pin connector purple wire and
> ground) or ....."

You edited my quote.
That's bad form.

> Look, if you are worried that the computer might still be active, just
> push the power switch (power button)(on the computer), as if you were
> trying to turn it on. If it doesn't turn on ..... then you know that
> the power really is killed. No need to start measuring things, although
> the motherboard standby power warning light is also a perfectly good
> indicator of the power supply status, except that not all motherboards
> have such an LED.

You suck.
--

Bill Eitner