Wooden PC
GT
Seriously - what is wrong with a wooden PC case? I'm thinking of making
myself one to replace my ageing metal case, which has more holes cut in it
than a colander. I would use the metal, slide out backing 'tray' for the
motherboard. Do I need to consider earthing everything, or is everything
earthed via its various feeding power cables (I suspect not). Where should I
earth things to - the power supply outer casing, or a point inside the PSU?
I expect it will muffle noise better than a metal case. There will be no
vibrations. It will be easier (for me) to work with when expanding /
adapting to new equipment. I can design a careful airflow. I'm thinking of
putting the hard disks in a compartment with the PSU. I'll have a front
vent(s), so the PSU fan at the rear will draw air over the drives and out
through the back. I'll have a second compartment for the motherboard area,
with the CPU fan drawing air in directly from the room, through a duct from
the side of the case, then I'll need just 1 large, quiet exhaust fan or
maybe even just a vents/holes as the CPU fan will draw in room temperature
air and force the compartment to fill with air, thereby pressurising the
compartment and forcing air out through an exhaust point.
Any thoughts / ideas / considerations?
kony
<ContactG...@hotmail.com> wrote:
>I'm thinking of building a wooden PC: Wooden case, Wooden CPU - Wooden go!!
>
>Seriously - what is wrong with a wooden PC case? I'm thinking of making
>myself one to replace my ageing metal case, which has more holes cut in it
>than a colander. I would use the metal, slide out backing 'tray' for the
>motherboard. Do I need to consider earthing everything, or is everything
>earthed via its various feeding power cables (I suspect not). Where should I
>earth things to - the power supply outer casing, or a point inside the PSU?
Everything is grounded by the power leads. Using wood you
will not have EMI trapping but if no other devices nearby
have a problem with this it should be an acceptible result.
>
>I expect it will muffle noise better than a metal case. There will be no
>vibrations.
True
> It will be easier (for me) to work with when expanding /
>adapting to new equipment.
?? How so when a metal case is designed for standard
components?
>I can design a careful airflow. I'm thinking of
>putting the hard disks in a compartment with the PSU. I'll have a front
>vent(s), so the PSU fan at the rear will draw air over the drives and out
>through the back. I'll have a second compartment for the motherboard area,
>with the CPU fan drawing air in directly from the room, through a duct from
>the side of the case, then I'll need just 1 large, quiet exhaust fan or
>maybe even just a vents/holes as the CPU fan will draw in room temperature
>air and force the compartment to fill with air, thereby pressurising the
>compartment and forcing air out through an exhaust point.
>
>Any thoughts / ideas / considerations?
>
Over time the heat will dry out the wood, consider oiling it
but that might make it smell funny for awhile. Otherwise
the main issue is the large # of calculations and precise
cuts to make everything fit, though using an existing
motherboard try will make that easier.
Fergus
"GT" <ContactG...@hotmail.com> wrote in message
news:00a76327$0$30353$c3e...@news.astraweb.com...
http://www.silentpcreview.com/article237-page1.html
GT
"kony" <sp...@spam.com> wrote in message
news:mfc0l3h4j3cis7kq9...@4ax.com...
> On Fri, 30 Nov 2007 13:59:44 -0000, "GT"
> <ContactG...@hotmail.com> wrote:
>
>>I'm thinking of building a wooden PC: Wooden case, Wooden CPU - Wooden
>>go!!
>>
>>Seriously - what is wrong with a wooden PC case? I'm thinking of making
>>myself one to replace my ageing metal case, which has more holes cut in it
>>than a colander. I would use the metal, slide out backing 'tray' for the
>>motherboard. Do I need to consider earthing everything, or is everything
>>earthed via its various feeding power cables (I suspect not). Where should
>>I
>>earth things to - the power supply outer casing, or a point inside the
>>PSU?
>
>
> Everything is grounded by the power leads. Using wood you
> will not have EMI trapping but if no other devices nearby
> have a problem with this it should be an acceptible result.
>
>
>
>>
>>I expect it will muffle noise better than a metal case. There will be no
>>vibrations.
>
> True
>
>> It will be easier (for me) to work with when expanding /
>>adapting to new equipment.
>
> ?? How so when a metal case is designed for standard
> components?
Was simply thinking long term - as I evolve the PC and move fans / intakes
etc around the case (as I have with my metal case over the last 12+ years),
I can just replace a side if there are too many holes! My metal case is now
littered with holes and folded bits of metal (don't have any snips - just a
hack saw!).
>>I can design a careful airflow. I'm thinking of
>>putting the hard disks in a compartment with the PSU. I'll have a front
>>vent(s), so the PSU fan at the rear will draw air over the drives and out
>>through the back. I'll have a second compartment for the motherboard area,
>>with the CPU fan drawing air in directly from the room, through a duct
>>from
>>the side of the case, then I'll need just 1 large, quiet exhaust fan or
>>maybe even just a vents/holes as the CPU fan will draw in room temperature
>>air and force the compartment to fill with air, thereby pressurising the
>>compartment and forcing air out through an exhaust point.
>>
>>Any thoughts / ideas / considerations?
>>
>
> Over time the heat will dry out the wood, consider oiling it
> but that might make it smell funny for awhile. Otherwise
> the main issue is the large # of calculations and precise
> cuts to make everything fit, though using an existing
> motherboard try will make that easier.
This isn't to make it look nice, its for fun!. If it takes ages (and it will
with a baby in the house and no time to even breath!), then I'm not
bothered! I'm thinking of MDF, so don't think that will dry out. I'm even
considering building it onto the underside of my desk. It is a corner desk
with metal underframe, so has good, strong mounting points. The project
would be well out of potential knee banging zone, would clear a bit of
floorspace, tidy cables up off the floor. I might even extend the front
panel lights/buttons and get them onto the top of the desk, perhaps even
recessed!. Its just a win-win really! Now where do I locate that internal
DVD drive...
wm_w...@hotmail.com
> Seriously - what is wrong with a wooden PC case?
If designed properly and well thought out, absolutely nothing. And
you'll have a computer case that is all your own and stands out from
the crowd. I have some wooden cased computers and they get a lot of
comments.
I would go ahead and run a ground wire from the case of the PSU to the
other metal components in the case when you have built it. It
certainly won't hurt anything to do this and chances are your current
case is completely grounded right now (if it is plugged into a
properly grounded outlet) by way of the power supply's metal case
being held against the case metal.
> I expect it will muffle noise better than a metal case.
> There will be no vibrations.
Wood tends to conduct sound fairly well and does so differently from
metal. I agree with your assessment that the wood case will be quieter
than a metal one, but I think you might hear noises you haven't heard
before.
> I can design a careful airflow.
It sounds like you've thought that out very carefully.
> Any thoughts / ideas / considerations?
There is really only one thing I could suggest adding or thinking
about. Sometimes computer hardware fails, and sometimes those failures
are violent. I'm sure many posters here could talk about hardware that
has for some reason or another gone "pop" or released its magic smoke.
I really doubt all but the most serious component failures would cause
a fire, but with a wood case this is something you will have to think
about. Air flow might only fan the flames if a fire gets started. I
would try to shield the motherboard at the outer edges, or mount a
sheet metal square on the side of the case that will face the
motherboard's component when it is all closed up and built. That way
if something does go "pop" or gets too hot, it will direct that heat
toward the metal and not the wood.
For power supplies and disk drives, mounting them in appropriate metal
cages that hold them a few inches from the wood case in every
direction should be fine.
GT
news:ZsmdncTZCMiYpc3a...@bt.com...
That is the Ferarri to the Skoda I was planning!!
Fergus
sorry the Skoda is
http://www.overclockers.com/tips1233/
GT
Hmm - Perhaps I was thinking 'Fiat' then, but definitely not Ferrari and
whatever it is, it will be fixed to the underside of my corner desk - out of
sight!
GT
"Fergus" <T...@Catbtinternet.com> wrote in message
news:ZsmdncTZCMiYpc3a...@bt.com...
>
http://www.ecogeek.org/content/view/62/1/
Fergus
pass me a Sanguka please.
philo
There would be no shielding...so you'd might want to line the case with
tin-foil and ground it
UCLAN
>>I'm thinking of building a wooden PC: Wooden case, Wooden CPU - Wooden go!!
>>
>>Seriously - what is wrong with a wooden PC case? I'm thinking of making
>>myself one to replace my ageing metal case, which has more holes cut in it
>>than a colander. I would use the metal, slide out backing 'tray' for the
>>motherboard. Do I need to consider earthing everything, or is everything
>>earthed via its various feeding power cables (I suspect not). Where should I
>>earth things to - the power supply outer casing, or a point inside the PSU?
>
> Everything is grounded by the power leads. Using wood you
> will not have EMI trapping but if no other devices nearby
> have a problem with this it should be an acceptible result.
The inductance of the power leads make them look like a very high
impedance to high frequency noise generated by the motherboard. The
MB should be grounded to a mounting plate with screws as usual. That
mounting plate can be returned to the PSU chassis with a short wire.
Ian D
grease the wooden spindle to wooden bearing contact points, otherwise
you'll have the hottest PC in town.
Keeping the original motherboard mounting plate and back panel is a
good idea. The plate provides grounding at specific points, and a
ground plane for the motherboard. Also grounding for the back panel
is required, and the physical tolerances for card brackets are tight.
Another suggestion is to use manufactured drives cages from your,
or other scrapped PC cases. If the various components are isolated
by wood, it's best to ground them all to a specific point, i.e.,the
motherboard plate, using copper braid. Copper braid provides an
excellent grounding route for high frequency RF. If you want to
shield the case panels, metal window screen is much better, and
easier to use than aluminum foil.
kony
<ContactG...@hotmail.com> wrote:
>>> It will be easier (for me) to work with when expanding /
>>>adapting to new equipment.
>>
>> ?? How so when a metal case is designed for standard
>> components?
>
>Was simply thinking long term - as I evolve the PC and move fans / intakes
>etc around the case (as I have with my metal case over the last 12+ years),
>I can just replace a side if there are too many holes! My metal case is now
>littered with holes and folded bits of metal (don't have any snips - just a
>hack saw!).
I think you're allowed to buy a new case at some point. I'm
not so sure that if you design this well, with airflow in
the right places to start with, that there would be any need
to change it later. Just cover all the basics like intake
for HDD rack, side intake across from video card, "maybe" a
side intake for CPU (mixed feelings about that, it really
shouldn't need this and unless you make the case extra wide,
many of today's very tall heatpipe coolers stick out so far
that there's barely if any room for a side panel fan across
from them, and even if there is, the fan is practically
butted up against a solid sheet of Al on the top of the
heatpipes.
I suppose if you later went with a different format like
BTX, that would require more changes, but how much time is
it worth to reuse a case? I often like to keep a system in
the same case and when upgrading, take my time building into
a new or new-used case without interrupting operation of the
existing system, plus if I want to give away or sell at some
point it has to have a case anyway.
>>>Any thoughts / ideas / considerations?
>>>
>>
>> Over time the heat will dry out the wood, consider oiling it
>> but that might make it smell funny for awhile. Otherwise
>> the main issue is the large # of calculations and precise
>> cuts to make everything fit, though using an existing
>> motherboard try will make that easier.
>
>This isn't to make it look nice, its for fun!.
I think you can have both?
>If it takes ages (and it will
>with a baby in the house and no time to even breath!), then I'm not
>bothered! I'm thinking of MDF, so don't think that will dry out.
Unfinished MDF? I suppose if that's what you want, but with
the airflow from the fans, it would seem the roughness would
catch, buildup more dust. It also seems like it will be
more difficult to create sunken holes for drive bays/etc in
MDF, though I suppose you could just use really long screws
instead of standard short ones.
>I'm even
>considering building it onto the underside of my desk. It is a corner desk
>with metal underframe, so has good, strong mounting points. The project
>would be well out of potential knee banging zone, would clear a bit of
>floorspace, tidy cables up off the floor. I might even extend the front
>panel lights/buttons and get them onto the top of the desk, perhaps even
>recessed!. Its just a win-win really! Now where do I locate that internal
>DVD drive...
You could just design the side of the case to VESA-mount an
LCD, make it about the width of an optical drive so it sits
up on the desk with the optical, buttons, indicators and
USB/etc ports right under the monitor frame.
kony
It makes no difference to the function of the system. I
can't tell you how many times I and millions of others have
boards out on a bench for testing... and I have one in front
of me at the moment as well.
kony
wm_w...@hotmail.com wrote:
>Hi!
>
>> Seriously - what is wrong with a wooden PC case?
>
>If designed properly and well thought out, absolutely nothing. And
>you'll have a computer case that is all your own and stands out from
>the crowd. I have some wooden cased computers and they get a lot of
>comments.
>
>I would go ahead and run a ground wire from the case of the PSU to the
>other metal components in the case when you have built it. It
>certainly won't hurt anything to do this and chances are your current
>case is completely grounded right now (if it is plugged into a
>properly grounded outlet) by way of the power supply's metal case
>being held against the case metal.
There is no need to do this. All components that need case
grounded, have it grounded through their power ground. This
is effective.
philo
"Ian D" <tau...@nowhere.com> wrote in message
news:Sf-dnV13m7hS583a...@giganews.com...
>
> "GT" <ContactG...@hotmail.com> wrote in message
> news:00a76327$0$30353$c3e...@news.astraweb.com...
> > I'm thinking of building a wooden PC: Wooden case, Wooden CPU - Wooden
> > go!!
> >
> > Seriously - what is wrong with a wooden PC case? I'm thinking of making
> > myself one to replace my ageing metal case, which has more holes cut in
it
> > than a colander. I would use the metal, slide out backing 'tray' for the
> > motherboard. Do I need to consider earthing everything, or is everything
> > earthed via its various feeding power cables (I suspect not). Where
should
> > I earth things to - the power supply outer casing, or a point inside the
> > PSU?
> >
>
here's a nice, wooded computer or something:
http://www.schmarder.com/radios/tube/30bb.htm
UCLAN
>>The inductance of the power leads make them look like a very high
>>impedance to high frequency noise generated by the motherboard. The
>>MB should be grounded to a mounting plate with screws as usual. That
>>mounting plate can be returned to the PSU chassis with a short wire.
>
> It makes no difference to the function of the system. I
> can't tell you how many times I and millions of others have
> boards out on a bench for testing... and I have one in front
> of me at the moment as well.
Testing? Fine. Day-to-day glitch-free operation? No.
kony
There is no distinction, the very purpose of testing is to
assure the proper operation.
If a motherboard can't operate properly due to the noise IT
creates, the ship has sunk before it's left the dock.
UCLAN
>>>It makes no difference to the function of the system. I
>>>can't tell you how many times I and millions of others have
>>>boards out on a bench for testing... and I have one in front
>>>of me at the moment as well.
>>
>>Testing? Fine. Day-to-day glitch-free operation? No.
>
> There is no distinction, the very purpose of testing is to
> assure the proper operation.
If you are leaving the MB out on the bench for weeks at a time, yes.
If you are only checking its operation for a short period of time, no.
Noise problems do not always present themselves immediately.
sandy58
> "GT" <ContactGT_remo...@hotmail.com> wrote in message
>
> news:00a28854$0$32354$c3e...@news.astraweb.com...
>
>
>
> > "Fergus" <T...@Catbtinternet.com> wrote in message
> >news:ZsmdncTZCMiYpc3a...@bt.com...
>
Hey, I like it!! I have a huge grin plastered across my "boat". Just
look out for termites etc. Nice one.
kony
Then what about people who have already built into wooden or
plexiglass cases?
The supposed noise problem just does not exist, except noise
escaping the non-metal case and effecting some OTHER device
that receives it.
Ken Maltby
All reasonable points, but there is no reason to avoid
any potential benefits from a common ground plane, or
maintaining the same condition, in that regard, as exists in
a normal metal case. A little wire or strap shouldn't be
that big a deal.
I like the idea of "remoting" the bays on the front of a
traditional case to the desktop. I would also install this
wood box on slides and/or hinges, so it would be more
accessible. The MB could be mounted "backwards"
as well, so all the rear connections would be handy.
With the PSU and internal HHD occupying their own
compartment, and airflow, you should factor that into
the design of the airflow across the MB.
Luck;
Ken
UCLAN
>>Noise problems do not always present themselves immediately.
>
> Then what about people who have already built into wooden or
> plexiglass cases?
What about them? What speed CPU are they using? What speed RAM? What
speed FSB? No intermittent "glitches" that had them scratching their
heads?
> The supposed noise problem just does not exist,
You mean *you* haven't experienced it. Others have and solved it with a
better ground. Don't believe me? I don't really care.
kony
<nom...@thanks.org> wrote:
>kony wrote:
>
>>>Noise problems do not always present themselves immediately.
>>
>> Then what about people who have already built into wooden or
>> plexiglass cases?
>
>What about them? What speed CPU are they using? What speed RAM? What
>speed FSB? No intermittent "glitches" that had them scratching their
>heads?
>
I forgot to take that poll or log statistics, since nobody
could know you'd come along with this idea so easily
disproven by all the systems out there running fine, but
generally speaking they're using the same parts as everyone
else except the case.
The noise just doesn't cause any problems. The motherboard
/other PCBs themselves are designed to provide stable
operation without an additional ferrous ground plane a few
millimeters behind them.
Since plenty of systems sitting on benches also run fine,
you must be thinking of some kind of rare event. A better
description of this event would be useful.
>> The supposed noise problem just does not exist,
>
>You mean *you* haven't experienced it. Others have and solved it with a
>better ground. Don't believe me? I don't really care.
"Better ground" is sufficiently vague to be unuseful.
Nobody with a properly designed PC motherboard needs a
"better ground" in the form of a metal chassis behind it,
UNLESS their area had very server leakage from some other
device which should not be operating because that other
device needed a grounded metal chassis to contain it's EMI.
The catch-22 is that if the computer is not then grounded,
it may effect some other devices. This leads to the
possibility that if you had several motherboards very
closely spaced on non-metal shelves, "maybe" they would
interact with each other, but that is not the scenario
posted in this thread, and that is also unlikely because
even when a system is inside a metal case, that only keeps
external signals out, all the parts in same case are still
interacting. Some particularly sensitive audio circuits
might be effected the most, since the nature of analog is in
signal levels instead of absolute thresholds.
~misfit~
> On Sat, 01 Dec 2007 22:19:55 -0800, UCLAN
> <nom...@thanks.org> wrote:
>> What about them? What speed CPU are they using? What speed RAM? What
>> speed FSB? No intermittent "glitches" that had them scratching their
>> heads?
>>
>
> I forgot to take that poll or log statistics, since nobody
> could know you'd come along with this idea so easily
> disproven by all the systems out there running fine, but
> generally speaking they're using the same parts as everyone
> else except the case.
Kony, unless you have spare time to dedicate to the pastime of arguing
you're best leaving this guy alone. I've had him killfiled ever since he
admitted that he was trolling (although he phrased it like "I enjoy a good
argument" or similar).
IMO it's best to keep this group for the discussion of comp.hardware and let
him go elsewhere for his perverse thrills. As always, your choice tho. :-)
--
TTFN,
Shaun.
kony
I'm just looking for specifics. Maybe there's one place on
the moon where it matters but it shouldn't. RF equipment
needs to keep stray external signals out but not our digital
computer gear which is required to be wholey self-grounded,
autonomous of case grounding with the exception of the PSU
itself in that it's circuit board is meant to stay
permanently inside the PSU casing and not touched by the
assembler or end-user... but even then, the PSU case need
not be grounded to the system case for functionality
purposes.
UCLAN
>>What about them? What speed CPU are they using? What speed RAM? What
>>speed FSB? No intermittent "glitches" that had them scratching their
>>heads?
>
> I forgot to take that poll or log statistics, since nobody
> could know you'd come along with this idea so easily
> disproven by all the systems out there running fine, but
> generally speaking they're using the same parts as everyone
> else except the case.
And you're knowledgeable of "all the systems out there running fine"...how?
As I said, just because *you* haven't seen the problem (?) doesn't mean
it doesn't exist. It does, and I've seen it.
> "Better ground" is sufficiently vague to be unuseful.
> Nobody with a properly designed PC motherboard needs a
> "better ground" in the form of a metal chassis behind it,
Never said anything like that. All I suggested was a short cable back to the
PSU, and to NOT rely solely on the relatively high impedance of the DC power
cable. A "better ground" in this discussion is just what has been said - one
that is a lower impedance to high frequency noise than is the PSU DC power
cable.
UCLAN
> Kony, unless you have spare time to dedicate to the pastime of arguing
> you're best leaving this guy alone. I've had him killfiled ever since he
> admitted that he was trolling (although he phrased it like "I enjoy a good
> argument" or similar).
Heh, heh. Still sore that you lost the America's Cup argument, I see.
And "enjoying a good argument (or similar)" is the same as trolling in
your book?
> IMO it's best to keep this group for the discussion of comp.hardware and let
> him go elsewhere for his perverse thrills. As always, your choice tho. :-)
I doubt many would term suggesting a proper ground a "perverse thrill",
but maybe things are different down in New Zealand.
kony
<nom...@thanks.org> wrote:
>kony wrote:
>
>>>What about them? What speed CPU are they using? What speed RAM? What
>>>speed FSB? No intermittent "glitches" that had them scratching their
>>>heads?
>>
>> I forgot to take that poll or log statistics, since nobody
>> could know you'd come along with this idea so easily
>> disproven by all the systems out there running fine, but
>> generally speaking they're using the same parts as everyone
>> else except the case.
>
>And you're knowledgeable of "all the systems out there running fine"...how?
>As I said, just because *you* haven't seen the problem (?) doesn't mean
>it doesn't exist. It does, and I've seen it.
So you claim, and yet no specifics.
I know they all do because that is how they are designed.
There are no PC parts, short of a TX/RX radio instrument
(like a wifi NIC, except if/when those need shielding they
have the shield on them already) that need it.
This is a necessary part of design because as I'd already
mentioned, the parts are inside a case that does not shield
one part from the next besides that shielding built into the
card (ground planes as needed).
>
>> "Better ground" is sufficiently vague to be unuseful.
>> Nobody with a properly designed PC motherboard needs a
>> "better ground" in the form of a metal chassis behind it,
>
>Never said anything like that. All I suggested was a short cable back to the
>PSU, and to NOT rely solely on the relatively high impedance of the DC power
>cable. A "better ground" in this discussion is just what has been said - one
>that is a lower impedance to high frequency noise than is the PSU DC power
>cable.
How is it you feel a short cable back to the PSU is somehow
better than the existing 9 ground leads on the ATX wiring
harness? There simply is no need for this. Show us a
motherboard that needs this, a reproducible problem when
this board doesn't have the case as ground plane behind it.
w_tom
> How is it you feel a short cable back to the PSU is somehow
> better than the existing 9 ground leads on the ATX wiring
> harness?
Reasons to justify that short ground cable are correct. That wire
back to power supply does have higher impedance. Then we add missing
facts. That high wire impedance is made irrelevant by electrolytic
and tantalum capacitors located on the motherboard that are necessary
to create a lower impedance. Impedance lower than anything provided
by a short ground cable.
One problem with a wooden chassis. Switches are rated for 20,000
volts so that human generated static electricity does not flow into
computer circuits. But that 20,000 volt isolation assumes the switch
body is chassis grounded so that static electric current has a path to
the room's floor without going through electronics. Current must flow
somewhere. A switch mounted on wood would no long provide that
alternative current path. Therefore 20,000 volts would pass into the
computer through switch causing a computer crash (maybe even damage).
Many external parts (switches, cable connectors) provide electronics
protection if their body (ie D connector shell) is grounded to a
chassis; not via motherboard. This protection is lost if wood
mounting surface does not have a conductive material (ie conductive
paint) connected back to AC safety ground wire.
As Kony notes, incoming (external) noise does not mal affect any
properly constructed motherboard. That shielding is required to meet
FCC emission standards. A solution to outgoing RF; not for incoming
noise.
~misfit~
> ~misfit~ wrote:
>
>> Kony, unless you have spare time to dedicate to the pastime of
>> arguing you're best leaving this guy alone. I've had him killfiled
>> ever since he admitted that he was trolling (although he phrased it
>> like "I enjoy a good argument" or similar).
>
> Heh, heh. Still sore that you lost the America's Cup argument, I see.
First sign that the person you're talking to is a troll; They rewrite
history.
Second sign; They try to engage you in an argument, often by rewriting
history.
> And "enjoying a good argument (or similar)" is the same as trolling in
> your book?
Yes, when it's in an inappropriate group. As I'm sure you are aware, there
are groups where 'arguments' are all they do.
>> IMO it's best to keep this group for the discussion of comp.hardware
>> and let him go elsewhere for his perverse thrills. As always, your
>> choice tho. :-)
>
> I doubt many would term suggesting a proper ground a "perverse
> thrill", but maybe things are different down in New Zealand.
I refuse to play your silly games. I'm here to discuss hardware and share
knowledge. Not play "argument" and one-upmanship. There are other groups for
that. However, I suspect that they are above your league, leaving you to
plague this group with your penchant for disharmony.
Now I'll consign this new return-to addy you adopted to defeat my killfile
(third sign of a troll) into the bozo bin, along with your last.
--
TTFN,
Shaun.
~misfit~
<snip>
>> Kony, unless you have spare time to dedicate to the pastime of
>> arguing you're best leaving this guy alone. I've had him killfiled
>> ever since he admitted that he was trolling (although he phrased it
>> like "I enjoy a good argument" or similar).
>>
>> IMO it's best to keep this group for the discussion of comp.hardware
>> and let him go elsewhere for his perverse thrills. As always, your
>> choice tho. :-)
>
>
> I'm just looking for specifics.....
<snip>
I know that you're looking for specifics and persuing the truth. The trouble
is, he knows it too. That's why he'll string you along for as long as he
can, all the while rubbing himself and "enjoying the argument".
--
TTFN,
Shaun.
GT
news:0e222f73-405e-4291...@s36g2000prg.googlegroups.com...
> On Dec 2, 5:07 pm, kony <s...@spam.com> wrote:
>> How is it you feel a short cable back to the PSU is somehow
>> better than the existing 9 ground leads on the ATX wiring
>> harness?
>
> Reasons to justify that short ground cable are correct. That wire
> back to power supply does have higher impedance. Then we add missing
> facts. That high wire impedance is made irrelevant by electrolytic
> and tantalum capacitors located on the motherboard that are necessary
> to create a lower impedance. Impedance lower than anything provided
> by a short ground cable.
>
> One problem with a wooden chassis. Switches are rated for 20,000
> volts so that human generated static electricity does not flow into
> computer circuits. But that 20,000 volt isolation assumes the switch
> body is chassis grounded so that static electric current has a path to
> the room's floor without going through electronics. Current must flow
> somewhere. A switch mounted on wood would no long provide that
> alternative current path. Therefore 20,000 volts would pass into the
> computer through switch causing a computer crash (maybe even damage).
My switch is currently mounted into the plastic front of my case - the most
static collecting material in this room (I'm not wearing any polyester)!. It
is not grounded, but connected directly to the motherboard via 2 wires
forming a basic, unearthed DC circuit. There is no ground connection from
any of the front panel LEDs or the 2 switches, so what difference would it
make if I replace the static-gathering front plastic panel with a non-static
gathering wooden one?
GT
news:ZLydncgQfY4A_s3a...@athenet.net...
Shielding from what? I have been running with the side off the PC for about
6 months now, so a wooden side will offer more shielding than it has at
present.
kony
<w_t...@usa.net> wrote:
>On Dec 2, 5:07 pm, kony <s...@spam.com> wrote:
>> How is it you feel a short cable back to the PSU is somehow
>> better than the existing 9 ground leads on the ATX wiring
>> harness?
>
> Reasons to justify that short ground cable are correct. That wire
>back to power supply does have higher impedance.
If there were only one, but there are 9 of them at least.
Remember also that having a motherboard ground plane to a
trace going to a standoff pad to a brass standoff to a metal
tray is also impedance.
>Then we add missing
>facts. That high wire impedance is made irrelevant by electrolytic
>and tantalum capacitors located on the motherboard that are necessary
>to create a lower impedance. Impedance lower than anything provided
>by a short ground cable.
... plus the ceramics quite close to the chips.
>
> One problem with a wooden chassis. Switches are rated for 20,000
>volts so that human generated static electricity does not flow into
>computer circuits. But that 20,000 volt isolation assumes the switch
>body is chassis grounded so that static electric current has a path to
>the room's floor without going through electronics. Current must flow
>somewhere. A switch mounted on wood would no long provide that
>alternative current path. Therefore 20,000 volts would pass into the
>computer through switch causing a computer crash (maybe even damage).
>
Most switches are not chassis grounded anymore, they are
small plastic bodied and held in place by an additonal
plastic frame which clips or screws into the front wall of
the chassis. Front panel USB and audio are still typically
grounded through PCB they're mounted on, to metal chassis.
> Many external parts (switches, cable connectors) provide electronics
>protection if their body (ie D connector shell) is grounded to a
>chassis; not via motherboard. This protection is lost if wood
>mounting surface does not have a conductive material (ie conductive
>paint) connected back to AC safety ground wire.
This would be my main concern, that ESD flows through them.
kony
True, front plastic switches are typically ungrounded to
chassis because of cheapness and convenience. Something
like a USB port would still be a reasonable candidate to
ground, it's metal frame to motherboard ground or to the PSU
IF you'll be having and plugging these ports.
Static generation need not come from wearing polyester nor
the case panel unless something were rubbed across it,
simply walking across some carpeting can cause it to some
degrees if humidity levels are not high enough.
GT
Absolutely, but my point was that the switches are not grounded at present,
so it won't make any difference if they are moved from a plascic housing to
a wooden one and therefore 'w_tom's point was.. er.. pointless!
UCLAN
> I know they all do because that is how they are designed.
> There are no PC parts, short of a TX/RX radio instrument
> (like a wifi NIC, except if/when those need shielding they
> have the shield on them already) that need it.
??? They are designed with being grounded to a metal chassis in mind,
not with a wood or plexiglass chassis.
> This is a necessary part of design because as I'd already
> mentioned, the parts are inside a case that does not shield
> one part from the next besides that shielding built into the
> card (ground planes as needed).
The individual ground planes all rely on a low impedance path back
to common ground. The power supply DC return does NOT provide a low
impedance for high frequency noise. Why do you think the MB ground
plane is grounded with metallic screws to the metallic chassis?
> How is it you feel a short cable back to the PSU is somehow
> better than the existing 9 ground leads on the ATX wiring
> harness?
I don't care if there are 100 ground leads if they all look like a
high impedance to high frequency noise. If you *really* believe that
the PSU DC return path is sufficient, do you use fiber washers between
all standoffs and the MB? Of course not. But with a wooden chassis,
that low impedance ground path is not possible. Hence, the short ground
wire back to the PSU.
UCLAN
>>>Kony, unless you have spare time to dedicate to the pastime of
>>>arguing you're best leaving this guy alone. I've had him killfiled
>>>ever since he admitted that he was trolling (although he phrased it
>>>like "I enjoy a good argument" or similar).
>>
>>Heh, heh. Still sore that you lost the America's Cup argument, I see.
>
> First sign that the person you're talking to is a troll; They rewrite
> history.
>
> Second sign; They try to engage you in an argument, often by rewriting
> history.
Killfile malfunctioning?
Rewriting history? Sure you want to revisit that thread? I didn't think so.
>>And "enjoying a good argument (or similar)" is the same as trolling in
>>your book?
>
> Yes, when it's in an inappropriate group. As I'm sure you are aware, there
> are groups where 'arguments' are all they do.
And what better group to discuss computer MB grounding in a wooden chassis
than a computer hardware group?
> I refuse to play your silly games. I'm here to discuss hardware and share
> knowledge. Not play "argument" and one-upmanship. There are other groups for
> that. However, I suspect that they are above your league, leaving you to
> plague this group with your penchant for disharmony.
I see. Disagreeing with Kony over MB grounding is "one-upmanship" ??
> Now I'll consign this new return-to addy you adopted to defeat my killfile
> (third sign of a troll) into the bozo bin, along with your last.
No "return-to" address was used. The complete headers can be viewed at:
http://groups.google.com/group/alt.comp.hardware/msg/e6a0d014569a5493?dmode=source
I replied to the group. No amount of fabrication on your part can change that.
It just shows that - once again - those that boast about placing somebody in
a killfile usually don't. And that liars are usually easily exposed.
kony
<nom...@thanks.org> wrote:
>kony wrote:
>
>> I know they all do because that is how they are designed.
>> There are no PC parts, short of a TX/RX radio instrument
>> (like a wifi NIC, except if/when those need shielding they
>> have the shield on them already) that need it.
>
>??? They are designed with being grounded to a metal chassis in mind,
>not with a wood or plexiglass chassis.
They designed it to be wholey grounded already by the power
connector(s).
>
>> This is a necessary part of design because as I'd already
>> mentioned, the parts are inside a case that does not shield
>> one part from the next besides that shielding built into the
>> card (ground planes as needed).
>
>The individual ground planes all rely on a low impedance path back
>to common ground. The power supply DC return does NOT provide a low
>impedance for high frequency noise. Why do you think the MB ground
>plane is grounded with metallic screws to the metallic chassis?
Actually, some have been plastic, and some people put fiber
washers on... and it still works fine.
The main reason they're metal is that's the best way to make
a standoff. Same applies for any screw in anything that
doesn't even screw into anything metal.
>
>> How is it you feel a short cable back to the PSU is somehow
>> better than the existing 9 ground leads on the ATX wiring
>> harness?
>
>I don't care if there are 100 ground leads if they all look like a
>high impedance to high frequency noise. If you *really* believe that
>the PSU DC return path is sufficient, do you use fiber washers between
>all standoffs and the MB? Of course not.
There is no impedance lower than the 9 return power leads
needed. This is by design. The board has ground planes
which handle this. Perhaps if it did not, THEN your idea
would have merit.
> But with a wooden chassis,
>that low impedance ground path is not possible. Hence, the short ground
>wire back to the PSU.
Would it interest you to know that some people even use
isolation mats on their PSU mounting so their case isn't
even grounded at all except through those 9 ground return
wires to the PSU? Again their systems work fine.
I suppose you'll just have to stop idealizing and see for
yourself. Plenty of us have already, there simply is no
need and it is easily proven.
Again I ask for specifics of a reproducible problem. If you
have no specifc scenario that can be reproduced, it is an
idea clearly without proof, while on the other hand there
are plenty of systems running right now showing the
opposite.
w_tom
> Absolutely, but my point was that the switches are not grounded at present,
> so it won't make any difference if they are moved from a plascic housing to
> a wooden one and therefore 'w_tom's point was.. er.. pointless!
You magically know plastic inside surface is not coated with a
conductive material? If that switch is pressed by a long plastic
material, then low conductive plastic would be sufficient to conduct
static elsewhere. But if your finger touches the switch plastic
button, then its body must be mounted on a more conductive surface.
That shorter plastic is sufficiently conductive to discharge into
computer electronics - as was accurately posted previously. Careful
how you just know only because you automatically know.
This static problem also existed in the earliest IBM PC and some
Apples. Keyboard designs were changed by coating inside plastic with
a conductive material. Why the different cable for IBM AT - or did you
never ask such questions - did you always just somehow know? Yes, GT,
a direct challenge to your snide remarks because you just
automatically know?
The point was pointless to those who did not solve those infrequent
problems over many decades; who do not discover why strange problems
occurred. One example of static electric: a wax on a linoleum tile
floor that crashed only some computer when switch was isolated from
conductive materials by epoxy paint. Ground the switch. Then all
computers worked without the rare and infrequent failure.
UCLAN
>>??? They are designed with being grounded to a metal chassis in mind,
>>not with a wood or plexiglass chassis.
>
> They designed it to be wholey grounded already by the power
> connector(s).
Really? Seems like a lot of trouble and expense for them to ensure all
necessary layers make contact with the mounting hole, including circuit
etching, if the boards didn't need it. It looks like they wanted better
ground as well. And manuals need not instruct direct contact with the
standoff if only the power connector was needed.
> The main reason they're metal is that's the best way to make
> a standoff. Same applies for any screw in anything that
> doesn't even screw into anything metal.
That doesn't explain the MB ensuring contact with the standoff
via the mounting hole.
> There is no impedance lower than the 9 return power leads
> needed.
For DC current, yes. For high frequency stuff, no.
> Would it interest you to know that some people even use
> isolation mats on their PSU mounting so their case isn't
> even grounded at all except through those 9 ground return
> wires to the PSU? Again their systems work fine.
Right! And I suppose they use special plastic screws to mount their PSU
as well. Get real.
> I suppose you'll just have to stop idealizing and see for
> yourself. Plenty of us have already, there simply is no
> need and it is easily proven.
You've operated a complete computer in a wooden or plexiglass case
for an extended period of time? Really? How long?
I'm not saying it will never work, just that poor high frequency grounding
will cause intermittent "glitches" in some systems where a metal chassis
is not used, or the MB is not mounted on a metallic tray, or a HF ground strap
between MB and PSU is not used. Lots of people smoke and don't get cancer, but
that doesn't mean nobody does.
w_tom
> I'm not saying it will never work, just that poor high frequency grounding
> will cause intermittent "glitches" in some systems where a metal chassis
> is not used, or the MB is not mounted on a metallic tray, or a HF ground strap
> between MB and PSU is not used.
Those high frequency currents are, instead, made completely
irrelevant by what was posted previously:
> That wire back to power supply does have higher impedance.
> Then we add missing facts. That high wire impedance is made
> irrelevant by electrolytic and tantalum capacitors located on the
> motherboard that are necessary to create a lower impedance.
> Impedance lower than anything provided by a short ground
> cable.
Not mentioned are even higher frequency currents. A motherboard is
chock full of ceramic capacitors adjacent to various ICs. Yes
adjacent as in less than one inch; that distance also may be critical.
No significant high frequency currents flow back to the power supply.
In fact, we might put more inductance on those wires just to further
restrict those high frequency currents on that power supply wire - for
reasons that do not apply here.
The hows and whys of grounding motherboard to chassis may involve
something completely different - also involving impedance. For
example (and only one example), put a computer on a glass table. Yes,
glass because other table top materials can be too conductive.
Generate massive static electricity in your body. Shock various parts
of the computer chassis including all eight corners. Those static
discharges should be so strong as to hurt. Computer software must not
crash. If computer does crash, then chassis grounding to motherboard
may be defective. These currents are completely different from
currents into power supply. Those currents are why motherboard ground
is typically different from chassis ground. Yes, both grounds share
an electrical connection - and are electrically different.
GT
> On Dec 3, 10:24 am, "GT" <ContactGT_remo...@hotmail.com> wrote:
>> Absolutely, but my point was that the switches are not grounded at
>> present,
>> so it won't make any difference if they are moved from a plascic housing
>> to
>> a wooden one and therefore 'w_tom's point was.. er.. pointless!
>
> You magically know plastic inside surface is not coated with a
> conductive material?
No magic involved - unlike you clearly did, I had a look!
> If that switch is pressed by a long plastic
> material, then low conductive plastic would be sufficient to conduct
> static elsewhere. But if your finger touches the switch plastic
> button, then its body must be mounted on a more conductive surface.
Nope - its just a plastic button with a small microswitch connected to the
motherboard via 2 wires. No magic conductive materials, so low conductive
plastic pointers, just a plastic switch that you press with your finger!
> That shorter plastic is sufficiently conductive to discharge into
> computer electronics - as was accurately posted previously. Careful
> how you just know only because you automatically know.
Nothing automatic - just had a look!
[blah blah blah]
kony
<w_t...@usa.net> wrote:
>On Dec 3, 10:24 am, "GT" <ContactGT_remo...@hotmail.com> wrote:
>> Absolutely, but my point was that the switches are not grounded at present,
>> so it won't make any difference if they are moved from a plascic housing to
>> a wooden one and therefore 'w_tom's point was.. er.. pointless!
>
> You magically know plastic inside surface is not coated with a
>conductive material? If that switch is pressed by a long plastic
>material, then low conductive plastic would be sufficient to conduct
>static elsewhere. But if your finger touches the switch plastic
>button, then its body must be mounted on a more conductive surface.
>That shorter plastic is sufficiently conductive to discharge into
>computer electronics - as was accurately posted previously. Careful
>how you just know only because you automatically know.
Here is a typical, example switch that is used in an OEM or
aftermarket case,
http://www.e-switch.com/pdf/TL2201.pdf
Certainly some systems use different switches than others,
and especially so with this custom build, the OP is free to
choose any switch he desires. Personally, I think a switch
like the following would look rather sharp mounted in wood,
and it's long threaded body would more easily accomdate
that. Since that body is metal, grounding it would be easy.
Unfortunately they are out of stock but I vaguely recall
some other electronics surplus 'site has them too, perhaps
bgmicro.com or allelectronics.com
http://www.mpja.com/prodinfo.asp?number=16093+SW
Yes, upon checking bgmicro has these in both flat and
rounded dome style.
http://www.bgmicro.com/index.asp?PageAction=VIEWPROD&ProdID=9358
http://www.bgmicro.com/index.asp?PageAction=VIEWPROD&ProdID=9359
>
> This static problem also existed in the earliest IBM PC and some
>Apples. Keyboard designs were changed by coating inside plastic with
>a conductive material.
That was back when things were built to last and cost a bit
more... today most are just made with cheapest method
possible which precludes using conductive coatings or
liners.
kony
<nom...@thanks.org> wrote:
>kony wrote:
>You've operated a complete computer in a wooden or plexiglass case
>for an extended period of time? Really? How long?
I've operated a computer with no case at all hanging on the
wall like a picture for years. I always have some system on
a bench as well. How long? I've been doing it for quite a
few years.
>
>I'm not saying it will never work, just that poor high frequency grounding
>will cause intermittent "glitches" in some systems where a metal chassis
>is not used, or the MB is not mounted on a metallic tray, or a HF ground strap
>between MB and PSU is not used. Lots of people smoke and don't get cancer, but
>that doesn't mean nobody does.
I'm still looking for examples. We can find examples of
people who smoke and have cancer, but what about systems
that didn't work properly until placed on/in this grounded
chassis? They do not need it, and you could test this
yourself instead of just a theory with no scientific method
to support resolution of the theory.
w_tom
> Here is a typical, example switch that is used in an OEM or
> aftermarket case,http://www.e-switch.com/pdf/TL2201.pdf
This switch is only rated:
> Dielectric Strength: 500 VAC
Assuming that is voltage from button to switch conductors (and not
its switching voltage), then static electric discharge could pass from
finger directly into electronics. Some manufacturer solutions include
a long plastic connector between where the switch is pressed and the
switch body. That separation distance provides the 20,000 volt
isolation.
Shielding performs many solutions. That shielding is not always
obvious. Even the conductive material coating inside of a plastic case
(ie keyboard) is not obvious. Other techniques are also used.
Shielding is not for protecting electronics from exterior noise. As
Kony notes, that motherboard must be designed with noise margins so
that exterior noise causes no problems. A chassis connection from
motherboard to mounting plate is not to reduce connection impedance to
the power supply.
Downside of a Wooden PC is that the exterior electrostatic shield may
not exist so that human interface points may put electronics at risk
due to static electric dishcharge and so that excessive RF radiation
exists.
Up side of a wooden PC - good wood (when not painted) can make a
pretty case. A wooden case can be constructed to provide proper
shielding from these rare and unacceptable problems.
UCLAN
> They do not need it, and you could test this
> yourself instead of just a theory with no scientific method
> to support resolution of the theory.
You want me to run my computer out of its case or in a plexiglass
case for a few months? No thanks. That would only prove it to *me*.
You would still doubt the results. An exercise in futility.
Post your phone number, and the next time I run into someone having
the problem I'll have him give you a call.
kony
<w_t...@usa.net> wrote:
>On Dec 4, 6:52 am, kony <s...@spam.com> wrote:
>> Here is a typical, example switch that is used in an OEM or
>> aftermarket case,http://www.e-switch.com/pdf/TL2201.pdf
>
> This switch is only rated:
>> Dielectric Strength: 500 VAC
> Assuming that is voltage from button to switch conductors (and not
>its switching voltage), then static electric discharge could pass from
>finger directly into electronics. Some manufacturer solutions include
>a long plastic connector between where the switch is pressed and the
>switch body. That separation distance provides the 20,000 volt
>isolation.
Yes, the above switch always has a button in front of it,
that button usually being significantly longer than the
stock button they sell for the power switch, but perhaps
about the same as their stock buttons for the reset switch.
In some OEM cases the distance is even further as there is a
plastic lever assembly between the outer pushbutton and the
inner switch. These levers or extended length buttons have
become normal now that cases have a fairly deep air cavity
between the front of the bezel and the front metal wall of
the case frame.
kony
You claim something with no evidence, while there is ample
evidence of systems running w/o problems having no metal
case ground. Seems you have an argument without any weight
behind it, only an idea which you have not shown has any
_significant_ effect, only that in theory it might have some
tiny effect which is outside of a level which would actually
effect the operation of a system.
Similarly, in theory, having magnets in a hard drive could
create magnetic fields that also interfere with a computer's
operation. In practice, we see through proof of systems
running fine that this also has far too small an effect to
matter.
~misfit~
> <nom...@thanks.org> wrote:
>
>> kony wrote:
>
>> You've operated a complete computer in a wooden or plexiglass case
>> for an extended period of time? Really? How long?
>
> I've operated a computer with no case at all hanging on the
> wall like a picture for years. I always have some system on
> a bench as well. How long? I've been doing it for quite a
> few years.
LOL! I did that too! ATX board and had the HDD mounted on a (generous-sized)
mobo tray with the PSU on a shelf next to it. It was my firewall/ethernet
switch with all PCI slots populated with NICs. Made a great talking point
with folks that weren't familiar with a computer's insides.
--
TTFN,
Shaun.
UCLAN
>>Post your phone number, and the next time I run into someone having
>>the problem I'll have him give you a call.
>
> You claim something with no evidence, while there is ample
> evidence of systems running w/o problems having no metal
> case ground. Seems you have an argument without any weight
> behind it, only an idea which you have not shown has any
> _significant_ effect, only that in theory it might have some
> tiny effect which is outside of a level which would actually
> effect the operation of a system.
As I wrote above, I'll have the owner of the computer having this problem
call you directly next time, but he'll probably simply add the ground
wire first.
kony
IMO, having the details here would be far more effective...
if there really is some board badly enough designed that it
needs this measure it would be good for others to be able to
find it through a search engine.
Meanwhile, if you insist on grounding yours, go right ahead
since it won't hurt anything, but as already shown time and
time again, neither will not doing it - except for the EMI
that then escapes to potentially effect some other equipment
nearby.
w_tom
>> evidence of systems running w/o problems having no metal
>> case ground. Seems you have an argument without any weight
>> behind it, only an idea which you have not shown has any
>> _significant_ effect, only that in theory it might have some
>
> As I wrote above, I'll have the owner of the computer having this problem
> call you directly next time, but he'll probably simply add the ground
> wire first.
"Doctor. When I bend this elbow, then its hurts."
"Then don't bend the elbow and pay the nurse on your way out."
That is ECLAN's logic. As Kony repeatedly and accurately notes, "You
claim something with no evidence ...". But speculation from UCLAN is
somehow presented as if fact.
This fact does exist and is ignored by UCLAN:
> That wire back to power supply does have higher impedance.
> Then we add missing facts. That high wire impedance is made
> irrelevant by electrolytic and tantalum capacitors located on the
> motherboard that are necessary to create a lower impedance.
> Impedance lower than anything provided by a short ground
> cable.
UCLAN demonstrates that 'speculation promoted as fact' is alive and
still frequent. To keep replying (he never once provides numbers), he
must ignore what he does not understand: those "electrolytic and
tantalum capacitors located on the motherboard that create lower
impedance".
A complaint here is about 'computer experts' who don't even know how
electricity works. To become A+ Certified Computer Tech, one needs no
electrical knowledge. UCLAN demonstrated what gets posted when one
does not have that basic electrical knowledge AND when one cannot
differentiate between speculation and facts.
Any low impedance created by that wire is made completely irrelevant
by something that provides even lower impedance: components on the
motherboard. A+ Certified Techs would not even know why those
components exist or what those components do. But those who do have
electrical knowledge can say confidently and with numerical facts that
his lower impedance connection accomplishes nothing.
Again, what UCLAN ignores because he does not understand it: filters
may be installed on those power supply wires to INCREASE that
reactance. Why do the opposite? Not relevant here. But also done
elsewhere for the same reasons that make UCLAN's solution useless,
ineffective, and an example of 'speculation promoted as fact'.
UCLAN - if your arguments here had merit, then you provided
numbers. You do not provide those numbers for two reasons. First -
that wire impedance is completely irrelevant as demonstrated even by
tantalum capacitors. Second - you don't demonstrate electrical
knowledge to explain the 'whys'. Nowhere do you cite what is solved
by that lower impedance connection. No 'whys' and no numbers; your
claims come from speculation. Kony has repeatedly posted technical
fact. Why, UCLAN, do you keep replying with speculation - and no
numbers? You have a solution to a problem that you cannot even
identify. The problem does not exist which UCLAN would know from
numbers and some fundamental electrical concepts. Do you even know
what a tantalum capacitor is?
"So don't bend the elbow". ECLAN cures a symptom without even
knowing what (or if) a problem exists. ECLAN does not even state what
lower impedance would accomplish. He demonstrates how other computer
'experts' also know without first learning how electricity works.
Mount that motherboard with plastic screws and it works just fine.
Somehow UCLAN knows those metal mounting screws must do something
electrical only because the screws are metal. If the screws are
metal, then an electrical reason must exist? That speculation is his
proof? UCLAN - those mounting holes must be coated in metal for
reasons mechanical. Only one hole - usually one closest to the DC
power connector - makes a necessary electrical connection for
different reasons discussed elsewhere. Reasons that are irrelevant to
anything posted by UCLAN.
Please first identify a problem before recommending a solution.
Your speculation results in a mythical problem and in a solution that
accomplishes nothing. Kony has repeatedly and accurately answered
your posts.
UCLAN
> This fact does exist and is ignored by UCLAN:
>
>> That wire back to power supply does have higher impedance.
>>Then we add missing facts. That high wire impedance is made
>>irrelevant by electrolytic and tantalum capacitors located on the
>>motherboard that are necessary to create a lower impedance.
>>Impedance lower than anything provided by a short ground
>>cable.
A capacitance in series with a high impedance has absolutely no effect
on the high impedance. It would be a different story if the caps were
in parallel with the high impedance, as the caps would act as a low
impedance to the high frequency noise. But, they're NOT. So, they DON'T.
> A complaint here is about 'computer experts' who don't even know how
> electricity works. To become A+ Certified Computer Tech, one needs no
> electrical knowledge. UCLAN demonstrated what gets posted when one
> does not have that basic electrical knowledge AND when one cannot
> differentiate between speculation and facts.
pot -> kettle -> black
Learn the difference between series and parallel circuits.
kony
<nom...@thanks.org> wrote:
>w_tom wrote:
>
>> This fact does exist and is ignored by UCLAN:
>>
>>> That wire back to power supply does have higher impedance.
>>>Then we add missing facts. That high wire impedance is made
>>>irrelevant by electrolytic and tantalum capacitors located on the
>>>motherboard that are necessary to create a lower impedance.
>>>Impedance lower than anything provided by a short ground
>>>cable.
>
>A capacitance in series with a high impedance has absolutely no effect
>on the high impedance.
They're not in series, these are decoupling caps.
>It would be a different story if the caps were
>in parallel with the high impedance, as the caps would act as a low
>impedance to the high frequency noise. But, they're NOT. So, they DON'T.
You need to look at a few more chip schematics or bust out a
multimeter and see for yourself, they're parallel,
decoupling. Series caps on a motherboard are a tiny
minority of those present.
UCLAN
>>A capacitance in series with a high impedance has absolutely no effect
>>on the high impedance.
>
> They're not in series, these are decoupling caps.
Doesn't matter *what* you call them. Their electrical position puts them in
series with the PSU power return lines, which are a high impedance return to
chassis ground. They *would* be a low impedance in parallel with those return
lines if the MB ground plane were tied directly to the chassis, which it's not
in a wooden case.
> You need to look at a few more chip schematics or bust out a
> multimeter and see for yourself, they're parallel,
> decoupling. Series caps on a motherboard are a tiny
> minority of those present.
*Any* cap to the ground plane on the MB is in series with the PSU ground
return wires. Only if the MB standoffs were directly grounded to the
chassis - which they're not in your scenario of a wooden case - would they
be in parallel. They may be in parallel *on the MB*, but they are in series
with the return lines *from* the MB to the PSU.
Your suggestion of testing this with a multimeter is useless. We are not
discussing DC here. A multimeter will only register resistance, *not*
impedance.
kony
<nom...@thanks.org> wrote:
>kony wrote:
>
>>>A capacitance in series with a high impedance has absolutely no effect
>>>on the high impedance.
>>
>> They're not in series, these are decoupling caps.
>
>Doesn't matter *what* you call them. Their electrical position puts them in
>series with the PSU power return lines, which are a high impedance return to
>chassis ground.
No it does not. Either you have a misunderstanding of the
word series, or you have not done much if any work with
computer electronics.
>They *would* be a low impedance in parallel with those return
>lines if the MB ground plane were tied directly to the chassis, which it's not
>in a wooden case.
Non-applicable
>
>> You need to look at a few more chip schematics or bust out a
>> multimeter and see for yourself, they're parallel,
>> decoupling. Series caps on a motherboard are a tiny
>> minority of those present.
>
>*Any* cap to the ground plane on the MB is in series with the PSU ground
>return wires.
No, the cap are generally across a VCC and Gnd. They are
expressly there for power smoothing which includes HF noise
when of a type capable of that frequency like the ceramics
all over the board. This is absolutely manditory circuit
design on everything these days from hard drives to
motherboards, to the processor carrier itself.
>Your suggestion of testing this with a multimeter is useless. We are not
>discussing DC here. A multimeter will only register resistance, *not*
>impedance.
A multimeter would easily determine the parallel or series
placement of the caps in the circuit... if only you knew
where to probe.
This subtopic has reached it's fruitful conclusion. Dance
around a fire swinging a chicken over your head too if you
feel like it, which will of course be similarly unnecessary.
w_tom
> A capacitance in series with a high impedance has absolutely no effect
> on the high impedance. It would be a different story if the caps were
> in parallel with the high impedance, as the caps would act as a low
> impedance to the high frequency noise. But, they're NOT. So, they DON'T.
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.
How does one know UCLAN is posting myths? First, he does not post
numbers for his 'low impedance' solution. Second, he posts a
solution for a problem he cannot even define. Third, he knows only
because he wired something and it worked; as if that is proof that
some problem even existed. Above three are characteristic of those
who proclaim as only a lying politician would; who need not learn
facts, numbers, and reasons 'why'. UCLAN cannot even define the
problem. But he has a solution.
UCLAN tells us those electrolytic and tantalum capacitors are in
series. Of course not. If in series, then no DC power could flow
from power supply to motherboard. UCLAN is that confused.
As Kony accurately notes, those capacitors are in parallel. They
are bypass capacitors. Their function is to create a low impedance
connection. If ULCAN had learned, then he knows what capacitors do,
how those capacitors were connected, what their purpose is, and why
his low impedance connection solves nothing. UCLAN had to be ignorant
on every point to post his reply. It explains why he ignored repeated
referenced to those tantalum, et al capacitors. To make his claims,
UCLAN had to know zero of four above functions. He had to be that
electrically naive.
Nothing posted here will be comprehended by UCLAN. To learn, he
must acknowledge he was wrong. He does not understand how electricity
works and denies electrical facts. UCLAN is too devoted to myths to
acknowledge reality. But again, this is not for UCLAN's benefit.
This post warns others that numerous myths like this are posted by
'self proclaimed' computer experts. This post is a warning to others
about those who are computer experts, but do not even know how
electricity works.
UCLAN does not even understand what a multimeter can measure. But
again, he somehow knows without understanding the significance of
numbers. He does not even understand his solution solves a mythical
problem. And then he overtly denies Kony's electrical facts. Those
capacitor are not in series no matter how voraciously UCLAN denies
electrical concepts.
UCLAN
>>A capacitance in series with a high impedance has absolutely no effect
>>on the high impedance. It would be a different story if the caps were
>>in parallel with the high impedance, as the caps would act as a low
>>impedance to the high frequency noise. But, they're NOT. So, they DON'T.
>
> 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
> 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
>> 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.
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
> 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~
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.
--
TTFN,
Shaun.
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
news:p1q6j.1647$Rf5...@newsfe13.phx...
> w_tom wrote:
>
>>> 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.
>
> 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
>>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!!
No kidding! Your point?
GT
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
But a plastic standoff won't make any difference - the motherboard might as
well be out of the case. Plastic doesn't conduct!
kony
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
>>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!
Sigh...at least one METALLIC standoff. [...shaking head...]
UCLAN
>>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.
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
<nom...@thanks.org> wrote:
>kony wrote:
>
>>>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.
>
>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
wrote:
>>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.
... 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
>>>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.
>
> 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.
>>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.
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.