Wooden PC

[UCLAN]

UCLAN has no electrical knowledge. Accused of posting as so many
computer techs who somehow are experts but do not even know how
electricity works. This is not intended for UCLAN. This post is
intended for others as a warning. Some routinely know; but cannot
bother to first learn even how electricity works.

Big boast, straw man. Are you saying that a capacitance in series with
a high impedance has an effect on that impedance? If so, we have just
learned that your basic electronics knowledge is seriously flawed. The
capacitance will look like a short to high frequencies, but the inductance
will still be a high impedance to those high frequencies. Now, if the
capacitance was in *parallel* with the inductance (which it wouldn't be
in a wooden case with no low impedance MB ground plane connection directly
to the chassis), it would be a different story. But it's not. Read a book.

 

[w_tom]

Are you saying that a capacitance in series with a high
impedance has an effect on that impedance? ... The capacitance
will look like a short to high frequencies, but the inductance will still
be a high impedance to those high frequencies.

Wow. UCLAN has no grasp. A bypass capacitor is in parallel. It
conducts no DC current AND is low impedance to high frequency
currents. Therefore high frequency currents don't need his low
impedance ground wire. UCLAN wants to solve a problem that does not
exist. This inductance - completely irrelevant and does not exist.

UCLAN tell us those bypass capacitors are in series. Each conducts
no DC current AND each is low impedance to high frequency currents.
No DC current flows from power supply to motherboard? UCLAN claims
'in series' electrolytic and tantalum capacitors conduct DC
electricity? He says those capacitors are 'in series'. Therefore
capacitors would block DC electricity from power supply to
motherboard. Somehow UCLAN is a computer expert. Electrical
knowledge need not be known. He need not understand why bypass
capacitors exist or even how capacitors are wired.

Trying to teach UCLAN science is futile. Instead, this post is a
warning to others about computer experts who don't even know how
electricity works. A board swapper needs no education to be a
computer expert. Even an A+ Certified Computer Tech needs no
electrical knowledge to be A+ Certified. Be wary of those who also
provide no numbers. When challenged with the 'whys', UCLAN invents
myths: such as bypass capacitors 'in series'. When challenged with
the existence of those electrolytic and tantalum capacitors, instead,
UCLAN avoided those posts. He did not understand. So he pretends the
science was never posted. He does not even know how electricity
works. He does not even know how bypass capacitors are connected. He
says capacitors are 'in series'. UCLAN then argues with Kony who
repeatedly posts technically accurate facts.

This is a warning to others about computer experts such as UCLAN who
do not know how electricity works. Who even provide no numbers to
justify speculation. Who would cure a problem he cannot even
identify. More indications of those who know - but cannot bother to
first learn the science.

 

[UCLAN]

Wow. UCLAN has no grasp. A bypass capacitor is in parallel.

Wow. w_tom has no grasp. A bypass cap on the MB - a MB without it's ground
plane tied to chassis via the MB standoffs - cannot be in parallel with the
DC power harness coming from the PSU. For a cap to be in parallel with that
harness, either one end of the cap would have to be attached to the PSU side
of that harness, or the MB standoffs would have to be attached to the metallic
chassis, which they are not. ANY high frequency noise being shunted to system
ground would HAVE TO go through the PSU DC harness to get there. There is
NOTHING in parallel with that harness, not even the bypass caps on the MB.

If we were discussing a metallic chassis with the ground plane connected
back to the PSU through the MB standoffs, it might be a different story.
But we're not. In this case, the bypass caps and the ground plane have
the SERIES impedance of the PSU DC power harness to deal with in the return
path to system ground.

No amount of babbling on your part will change this.

 

[Noozer]

Wow. w_tom has no grasp. A bypass cap on the MB - a MB without it's ground

plane tied to chassis via the MB standoffs - cannot be in parallel with
the
DC power harness coming from the PSU. For a cap to be in parallel with
that
harness, either one end of the cap would have to be attached to the PSU
side
of that harness, or the MB standoffs would have to be attached to the
metallic
chassis, which they are not.

You do realize that there isn't a single mainboard out there that relies on
the chassis to connect any components to ground. Every single component that
goes to ground DOES connect to the ground cable on the PSU - EVEN IF THE
MAINBOARD IS SUSPENDED IN MID AIR BY NOTHING EXCEPT THE PSU CONNECTION.

 

[~misfit~]

Somewhere on teh interweb Noozer typed:

You do realize that there isn't a single mainboard out there that
relies on the chassis to connect any components to ground. Every
single component that goes to ground DOES connect to the ground cable
on the PSU - EVEN IF THE MAINBOARD IS SUSPENDED IN MID AIR BY NOTHING
EXCEPT THE PSU CONNECTION.

You do reaise that you're disagreeing with someone for whom arguing is a
hobby? For him, it's not about the facts, it's about getting a response.
Perhaps he wasn't breast-fed or he was bought up in a foster home. It's hard
to say. However, it's not hard to say he's a troll. He's admitted it
himself, although he phrases it like "I enjoy a good discussion".

It's sad to see you, w_tom and Kony wasting your valuable time with this...
serial arguer.

http://www.education-world.com/a_curr/shore/shore015.shtml

I bet his teachers wished they had this resource. At this stage in his life
I think only option 1 is viable. Also the last sentence is relevant.

 

[UCLAN]

You do realize that there isn't a single mainboard out there that relies on
the chassis to connect any components to ground. Every single component that
goes to ground DOES connect to the ground cable on the PSU - EVEN IF THE
MAINBOARD IS SUSPENDED IN MID AIR BY NOTHING EXCEPT THE PSU CONNECTION.

You do realize that we are discussing whether or not the high frequency
impedance (inductance) of the PSU DC MB harness would or would not affect
the amount of high frequency noise actually being shunted to chassis ground
in a system (wooden case) relying on this as the ONLY path to ground. We are
not discussing a component's DC connection to ground.

If the PSU DC power harness was all that was needed for optimum operation,
why do MB manufacturers go through the hassle and expense of ensuring that
the ground plane is connected to the solder ring around the standoff holes?
This effort on their part is useless if the standoffs are not grounded.

DC paths to ground and high frequency paths to ground are two different
things.

 

[GT]

Wow. w_tom has no grasp. A bypass cap on the MB - a MB without it's ground
plane tied to chassis via the MB standoffs -

Plastic standoffs won't tie anything to the chassis!!

 

[UCLAN]

Plastic standoffs won't tie anything to the chassis!!

No kidding! Your point?

 

[GT]

No kidding! Your point?

A motherboard operating outside a metal case is grounded in exactly the same
way as a motherboard that is mounted in a metal case, so any mention of the
standoffs in this discussion is irrelevant. Not all standoffs are brass, so
grounding cannot be guaranteed and is therefore not achieved via the
standoffs.

 

[UCLAN]

A motherboard operating outside a metal case is grounded in exactly the same
way as a motherboard that is mounted in a metal case, so any mention of the
standoffs in this discussion is irrelevant. Not all standoffs are brass, so
grounding cannot be guaranteed and is therefore not achieved via the
standoffs.

Sigh....

You're rehashing an old thread. Measure the HF noise on the many bypass caps
with and without the ground plane attached to chassis with at least one
standoff. The difference should cause a "hmmm..."

 

[GT]

Sigh....

You're rehashing an old thread. Measure the HF noise on the many bypass
caps
with and without the ground plane attached to chassis with at least one
standoff. The difference should cause a "hmmm..."

But a plastic standoff won't make any difference - the motherboard might as
well be out of the case. Plastic doesn't conduct!

 

[kony]

Sigh....

You're rehashing an old thread. Measure the HF noise on the many bypass caps
with and without the ground plane attached to chassis with at least one
standoff. The difference should cause a "hmmm..."

No it shouldn't, because the level of external HF noise
making it's way into the system onto the ground plane is
trivial compared to the HF noise the system ITSELF PRODUCES!

It's not keeing noise out, it's keeping noise IN.

 

[UCLAN]

But a plastic standoff won't make any difference - the motherboard might as
well be out of the case. Plastic doesn't conduct!

Sigh...at least one METALLIC standoff. [...shaking head...]

 

[UCLAN]

No it shouldn't, because the level of external HF noise
making it's way into the system onto the ground plane is
trivial compared to the HF noise the system ITSELF PRODUCES!

It's not keeing noise out, it's keeping noise IN.

What's keeping noise in? We're not discussing the chassis keeping in RFI.
I'm talking about keeping high frequency noise (spikes) out of MB circuitry.
The amount of high frequency noise entering Vcc or logic inputs of ICs is
lower with the ground plane directly grounded to the chassis with at least
one metallic standoff than it is relying on the PSU DC return path as the
only ground plane connection to chassis ground. Measure it.

 

[kony]

What's keeping noise in? We're not discussing the chassis keeping in RFI.
I'm talking about keeping high frequency noise (spikes) out of MB circuitry.

Yes, and that's wrong.
The high frequency noise already present in the motherboard
circuitry, produced by the motherboard and plugged in parts,
is orders of magnitude more than anything you'd pick up by
not having the board connected to the motherboard tray by
standoffs.

The amount of high frequency noise entering Vcc or logic inputs of ICs is
lower with the ground plane directly grounded to the chassis with at least
one metallic standoff than it is relying on the PSU DC return path as the
only ground plane connection to chassis ground. Measure it.

It's not even close to enough to matter, and unmeasurable.

If you had actually measured it, you would realize that the
only reason a computer works at all is that it is a digital
system quite a bit more immune to stray RF than an analog
system is, because, again, it produces far more noise than
anything close enough to matter, unless you had duct-taped
it to the side of a radio tower.

 

[kony]

It's not even close to enough to matter, and unmeasurable.

.... unless you were trying to only measure dB at any given
frequency, which is fairly irrlevant since it won't effect
operation and is largely eclipsed by noise at the frequency
of the system parts.. being far higher and more significant,
including the noise produced by the PSU itself and the
ripple of the powered parts.

There just isn't a noise pickup factor present, the system
IS the noise!

Instead of going on and on about this, it's about time you
found a reproducible problem in not having the standoffs and
earth grounded tray. Go ahead and do that, if you don't
then your theory falls through from lack of evidence.

We can sit and daydream about anything, but without evidence
it is only folly.

 

[UCLAN]

Yes, and that's wrong.
The high frequency noise already present in the motherboard
circuitry, produced by the motherboard and plugged in parts,
is orders of magnitude more than anything you'd pick up by
not having the board connected to the motherboard tray by
standoffs.

I'm not referring to what I'd "pick up" by not using the standoffs,
I'm saying that the bypass caps do a better job when the ground plane
is directly connected to the chassis rather than having to go through
the PSU DC cable in its return to chassis.

It's not even close to enough to matter, and unmeasurable.

The high frequency noise spikes as measured with an oscilloscope are quite
measurably of higher amplitude when the ground plane is not directly grounded.
Look and see for yourself.

If you had actually measured it, you would realize that the
only reason a computer works at all is that it is a digital
system quite a bit more immune to stray RF than an analog
system is, because, again, it produces far more noise than
anything close enough to matter, unless you had duct-taped
it to the side of a radio tower.

Yes, it produces much high frequency noise. And keeping this high frequency
noise out of logic IC Vcc and inputs can improve the lifetime of that IC
(if the spikes are higher than the maximum recommended Vcc voltage), or in
extreme cases eliminate an intermittent "glitch" (if spike is of long enough
duration on a logic input.) Having the MB ground plane tied directly to the
chassis reduces the noise spikes in these areas. Please stop trying to confuse
the issue by making statements like your above "radio tower" statement. You
and I both know we are discussing internally generated noise.