Looking for motherboard

  • Thread starter Thread starter not_here.5.species8350
  • Start date Start date
I used the site you referenced and found the Vista OS link. However the WPA links back to XP.

When the time comes, I would rather run a free Linux OS than buy another copy of Vista.

That article, gives some indication how Microsoft
goes about tracking hardware changes for product
activation. Since it only "hints" at it, there aren't
any iron-clad rules publicly available. Even if
there were rules available, it probably wouldn't
be of much practical help to people. When there
is a problem, you still get to phone Microsoft.
They'll fix activation issues for you. Worst case,
the phone operator gives you a long string of
digits to type in, to override product activation,
so on a one-off basis, a problem can be fixed.

Paul
 
My motherbaoard has blown and am looking for a repalacement. It can be second-hand or refurbished - the chepaer the better.



It is to replace a motherboard in a Dell Dimension E 520 (LGA 775, Pentium D 915 2.8GHz processor, chipset probably G965, power supply output 305W. Form factor BTX



I am considering the DEll motherboard, but I'd like something cheaper, my machine is quite old.



Any references greatly received



PS. UK references

Thank you

'It's for situations where the hard drive died, you didn't make recovery
CDs like you were supposed to, then you can take a regular Windows installer
CD and use the COA license key. Then, the OS must be activated in the
same way as a home user would do it. If there's any problem, you just
phone Microsoft and they can help you.'

I did not make a recovery disk - I have the original VISTA disk - will this suffice?

Would I still need to make an installer, presumably I could make this prior to activation? I don't know how.

If I needed to phone Microsoft, will the number appear on the screen - hopefully toll free from the UK?

For Dell systems people usually contact them (Dell) - obviously not in this case.

Best wishes.
 
Thank you

'It's for situations where the hard drive died, you didn't make recovery
CDs like you were supposed to, then you can take a regular Windows installer
CD and use the COA license key. Then, the OS must be activated in the
same way as a home user would do it. If there's any problem, you just
phone Microsoft and they can help you.'

I did not make a recovery disk - I have the original VISTA disk - will this suffice?

Would I still need to make an installer, presumably I could make this prior to activation? I don't know how.

If I needed to phone Microsoft, will the number appear on the screen - hopefully toll free from the UK?

For Dell systems people usually contact them (Dell) - obviously not in this case.

Best wishes.

So what you're telling me, is you upgraded past the OS provided
with the Dell ?

That means you have an original COA from Dell on the computer case.

But, if you bought a copy of Vista, you will also have a COA with
license key, that came with the Vista disc. You would use that
license key, for installation. If the Vista disc is OEM (system builder),
then you'd activate the same way a home user would (not like a Dell
user - the Dell OS would not need to be activated).

You won't need to reinstall, but if you do, your separately
purchased Vista disc is more straight-forward. The OEM terms
say "for usage on one computer", and it can be tricky "upgrading"
a motherboard. If the motherboard is identical to the old one,
that is not a problem at all.

When I upgraded the motherboard on this WinXP machine with OEM
installation, the activation server had been relaxed enough, that
I didn't need to phone anyone. It just activated by clicking when
needed. Worst case, if you need to phone, they should provide a
phone number on the screen. And then, you can explain how your motherboard
failed, you've installed an identical motherboard (just the NIC MAC
address will be different), if they're at all concerned. I really
don't expect a problem here, if you're installing an identical motherboard.
It should just activate with a click, done.

If, on the other hand, you installed a completely different motherboard
that takes your processor, a different amount of RAM, and so on,
then you're more likely to need to phone. Strictly speaking,
you're not allowed to upgrade the motherboard on an OEM install,
as that constitutes "a different computer" from the Microsoft
bookkeeping perspective. But they're flexible, and explaining your
motherboard failed, and the new motherboard is the best you
could do for a replacement, they should help you out. They're
mainly there, to keep you from "stealing" an OS. With my
OEM WinXP, I didn't even need to do that when I changed from
an Asrock motherboard, to an Asus motherboard with entirely
different chipset (went from VIA chipset, to Intel chipset).
The VIA chipset had a bug that prevented my TV tuner
card from working properly, so I had to change motherboards
to fix it.

And yes, you're correct - if you buy your own Vista DVD,
with its own Vista license key, Dell has nothing to do
with it, and you would not be contacting Dell at all.
In fact, if you purchase an OEM (system builder) disc, in fact
*you* provide tech support :-) That's the whole idea with
OEM (system builder). Since a Retail disc (with limited
Microsoft support) costs so much more, it is more likely
that most people will get an OEM or upgrade edition of
some sort.

If it is an upgrade disc, then it should do a check for
a "qualifying OS". In such a case, if your hard drive
was blank, you could try the "double install method"
as a means to putting an upgrade version on an effectively
empty machine.

This is how you use an upgrade disc, and the double install
method, to install an upgrade disk to an empty hard drive.
This even works for Windows 8 (conceptually), but the
details are just slightly different. The Windows 8 method
is still a double install. The first install, with no
license key, is to "look like an upgrade OS".

http://www.vistax64.com/tutorials/68767-clean-install-upgrade-vista.html

*******

The same guy runs several tutorial servers, vistax64.com,
sevenforums.com, eightforums.com . When you have a problem,
you can go to Google and do:

site:vistax64.com tutorial upgrade double install

or the like. That searches the tutorial server only,
for related articles. The word tutorial is there,
to try to find the articles written by Shawn.

Paul
 
My motherbaoard has blown and am looking for a repalacement. It can be second-hand or refurbished - the chepaer the better.



It is to replace a motherboard in a Dell Dimension E 520 (LGA 775, Pentium D 915 2.8GHz processor, chipset probably G965, power supply output 305W. Form factor BTX



I am considering the DEll motherboard, but I'd like something cheaper, mymachine is quite old.



Any references greatly received



PS. UK references

Thanks yet again.

I think that I confused you.

What I meant to say is that I did not create any disks because I have the VISTA OS on a disk supplied by Dell. I have not purchaed any other software.

I am looking for a motherboard at present, the main criteria, other than fit and it must work, is the price, the cheaper the better - I am looking at used boards.

As a matter of interest, people who sell motherboards claim they are fully tested. If a motherboard came from a fully working pc, then I suppose that that is the test. But otherwise, how do they test these boards?

If I get my system up and running, I have not made a final decision about whether I will re-install VISTA (if necessary), or the free UBUNTU OS.

Previously, when I asked you about those black fins on the graphics card, they did not ring a bell. I think that they may be part of the card cooling system. If I am right, I wonder why they are not made of aluminium. I can'tsee what they are cooling. I found a similar card on ebay, not quite the same as mine, but it shows the black fins. I give the link below.

http://www.ebay.co.uk/itm/Dell-CH48...Computing_CPUs_Processors&hash=item257d394e5a

Best wishes.
 
My motherbaoard has blown and am looking for a repalacement. It can be second-hand or refurbished - the chepaer the better.



It is to replace a motherboard in a Dell Dimension E 520 (LGA 775, Pentium D 915 2.8GHz processor, chipset probably G965, power supply output 305W. Form factor BTX



I am considering the DEll motherboard, but I'd like something cheaper, my machine is quite old.



Any references greatly received



PS. UK references

Forgot to mention.

I will read the links that you kindly provided.
 
Thanks yet again.

I think that I confused you.

What I meant to say is that I did not create any disks because I have the VISTA OS on a disk supplied by Dell. I have not purchaed any other software.

I am looking for a motherboard at present, the main criteria, other than fit and it must work, is the price, the cheaper the better - I am looking at used boards.

As a matter of interest, people who sell motherboards claim they are fully tested. If a motherboard came from a fully working pc, then I suppose that that is the test. But otherwise, how do they test these boards?

If I get my system up and running, I have not made a final decision about whether I will re-install VISTA (if necessary), or the free UBUNTU OS.

Previously, when I asked you about those black fins on the graphics card, they did not ring a bell. I think that they may be part of the card cooling system. If I am right, I wonder why they are not made of aluminium. I can't see what they are cooling. I found a similar card on ebay, not quite the same as mine, but it shows the black fins. I give the link below.

http://www.ebay.co.uk/itm/Dell-CH48...Computing_CPUs_Processors&hash=item257d394e5a

Best wishes.

As far as I know, that's an anodized aluminum heatsink.
It's aluminum underneath. The coating on the outside, helps
protect the aluminum. If you take a sharp tool, and scratch
off the black finish, shiny aluminum should be underneath.

http://en.wikipedia.org/wiki/Anodizing

The "fully tested" means what you'd expect - the motherboard
was removed from a working system. Since you can't see them
doing the work, from 3000 miles away, they're under no
obligation to even lift a finger to test it.

Imagine the labor cost component, if they actually did
test them. They're not going to hire extra guys, to
sit around testing stuff all day. No profit in that.

Paul
 
My motherbaoard has blown and am looking for a repalacement. It can be second-hand or refurbished - the chepaer the better.



It is to replace a motherboard in a Dell Dimension E 520 (LGA 775, Pentium D 915 2.8GHz processor, chipset probably G965, power supply output 305W. Form factor BTX



I am considering the DEll motherboard, but I'd like something cheaper, mymachine is quite old.



Any references greatly received



PS. UK references

Im my search for motherboards I am only looking for the motherboard, no RAM, or CPU. I have a CPU and RAM modules on the old board and tend to use these on the next board.

Regarding my previous post and the question about motherbpoard testing. I had in mind that perhaps there was some form of diagnostic tool that could be plugged into the board. Grenn light fully functional; red light faulty - something like that. It appears not to be the case. However, I have read many posts where people talk of refurbished motherboards, this does suggest that faulty boards have beeen repaired, hence faults must have been identified.

Reagarding the cooling fins on the card mentioned, I assume that there mustbe a chip under the fins that requires cooling.

Best wishes.
 
Im my search for motherboards I am only looking for the motherboard,
no RAM, or CPU. I have a CPU and RAM modules on the old board and tend
to use these on the next board.

Regarding my previous post and the question about motherbpoard testing.
I had in mind that perhaps there was some form of diagnostic tool that
could be plugged into the board. Grenn light fully functional; red light
faulty - something like that.

There was a video of production testing at one of the motherboard
companies, and their test setup was more complicated than that.

The closest thing to a test card, is a PCI Port 80 test card.
And it doesn't give much more information, than a beep test would.
Such a card is fine if you already own one, but potentially a
waste of money if you go looking to buy one.

There are loopback test devices, for testing communications ports.
So there are a few other small test tools of that type.

At the factory, there are two kinds ot testing: structural and functional.
Structural can be done with a bed-of-nails tester. It basically
verifies wires and signals. The bed-of-nails is less concerned
with what the motherboard is supposed to do. Where I worked, a lot of
our testing was done that way (boundary scan, continuity testing).

A motherboard manufacturer also gives the motherboard a two minute functional
check. Several removable assemblies are plugged into the motherboard,
and then it is tested. If you make three million motherboards a
month, you can't spend very long on each one. Of the two minutes,
some of the time is wasted just assembling and disassembling the
test setup.
It appears not to be the case. However, I
have read many posts where people talk of refurbished motherboards, this
does suggest that faulty boards have beeen repaired, hence faults must
have been identified.

A "tested" board, came from a working system. It's expensive to verify
the boards, if you're only charging $50 for each one. And if the companies
are in North America, the wage rate sees to it, that they can't afford
to do much work on each board.

A "refurbished" board, maybe they blew the dust off it. Or wiped it
with a damp cloth. (While boards can be washed in a thing that looks
like a dish washer, I doubt many small companies would bother. There
is also a risk of forcing dirt, underneath mechanical assemblies. You
have to install rubber dams on components that cannot take a washing.
You cover the Ethernet port, cover the CPU socket, and so on. Who is
going to go to that much trouble ?)

It would take an exceptional company, to have a repair
facility of any sort. Repairing boards in North America
simply costs too much. To get a Southbridge replaced here
(with 2.5D xray verification), costs about $1000. If you get someone
in Taiwan to do it, maybe $25, because they do so many of them, and
the staff are never idle. And nobody in a used motherboard
business, is going to shuttle products back and forth to a
country that can "do it for cheap". That would be a logistics
nightmare. Think of the paperwork at the borders.

If a board comes from one of the failure-prone Dells, then
you're looking for the words "re-capped", meaning capacitors
have been removed and new ones installed. The word "refurbished"
really is not specific enough, to conclude a board has been
re-capped, and a major source of unreliability, removed.

It's not like someone rebuilding a car alternator, and changing
the brushes and bearings. They more likely to just blow the
dust off, and put it in a box.

Many companies went through stuff like the following. Including one
motherboard firm that is no longer in business. This is
just a sampling of the cases. Some companies had a small number
of failures, while others had huge failure rates. And for those
motherboards, if you're looking for a spare one of the same
model, then you want confirmation it has been re-capped. If you want
someone to re-cap it for you, it costs $50 to $100 or so. (There
was a guy doing them for $50, but he got sick of it and quit.)

http://www.theregister.co.uk/2011/07/08/dell_optiplex_lawsuit_expanded/

For a motherboard with that kind of reliability problem,
simply looking at your own motherboard, and looking for
signs of leakage on the caps, may be enough to conclude there
was a risk the replacement boards would be just as bad.
If your board appears to be in good cosmetic shape, that's
a quick check maybe the replacement will be OK too.

I had a power supply here, an Antec (made by ChannelWell), and
the capacitors in that leaked, without the power being on.
That supply went bad, sitting in the box. For the capacitors with
bad internal chemistry, the capacitors can start to leak on their
own. They don't need to be stressed, to start the ball rolling.

http://en.wikipedia.org/wiki/Capacitor_plague
Reagarding the cooling fins on the card mentioned, I assume that
there must be a chip under the fins that requires cooling.

Yes :-) Up to a certain power level, you can use passive cooling, and
the fins are enough. Above that level, you need a fan blowing onto
the fins. With a typical fan, the performance of the heatsink
improves by around a factor of three. Even more is possible,
if heatpipes help distribute the heat into the arrays of fins.
The very best coolers, use heatpipes. A heatpipe can transfer
more heat flux, than an equivalent solid piece of copper.

This is an example of a video card cooler, where the heat from the
GPU, goes down those vapor-filled pipes, and then the heat
flows into the fins. The heatpipe allows a long thermal path,
while having a low thermal resistance, so you actually get
some benefit from the fins. If it weren't for the heatpipes on
this one, the outer fins would be cold to the touch. The heatpipe
ensures the heat, gets to the fins, and all the fin surface is
working. That grey material on the main heatsink body, aids
thermal conductivity to the top of the GPU chip. This thing
works better, than a solid aluminum (anodized) heatsink.

http://images17.newegg.com/is/image/newegg/35-186-054-Z03?$S300W$
Best wishes.

HTH,
Paul
 
My motherbaoard has blown and am looking for a repalacement. It can be second-hand or refurbished - the chepaer the better.



It is to replace a motherboard in a Dell Dimension E 520 (LGA 775, Pentium D 915 2.8GHz processor, chipset probably G965, power supply output 305W. Form factor BTX



I am considering the DEll motherboard, but I'd like something cheaper, mymachine is quite old.



Any references greatly received



PS. UK references

Thanks again.

Am I right in assuming that the definition of a motherboard is the board (mobo)where all removables have been removed, or in the case of a new mobo not installed. In other words, a motherboard does not include memory modules,cards, the processor, or the battery. Fortunateley, I have these on my existing board. Sometimes I note when viewing motherboards that the processsorsocket has a plate in place. I assume that there is no processor underneath if they are advertising a motherboard. Are my assumptions correct?

I did not see any leaking, bulging, or split capacitors (an experienced engineer might see someting), but any faults could easily be internal.

Best wishes
 
On Sun, 28 Apr 2013 05:44:07 -0700 (PDT),
Am I right in assuming that the definition
of a motherboard is the board (mobo)where all removables have been
removed, or in the case of a new mobo not installed. In other words, a
motherboard does not include memory modules, cards, the processor, or
the battery. Fortunateley, I have these on my existing board.
Sometimes I note when viewing motherboards that the processsor socket
has a plate in place. I assume that there is no processor underneath
if they are advertising a motherboard. Are my assumptions correct?
I did not see any leaking, bulging,
or split capacitors (an experienced engineer might see someting), but
any faults could easily be internal.

The MB definition, mine, is something that lasts 5 years at a minimum
to 12 years, or more. Comes with a flimsy tin plate to reveal what
isn't directly connected on the backplane's potential I/O connections,
a battery already installed, some storage cables, pin jumper blocks
installed so indicated, default pin jumpers, possibly a cute,
itty-bitty piezo speaker, drivers on an optical disc, and a protector
plate over the CPU socket when and especially if warranted (something
that can be painfully apparent after a successful repair of bent
miniscule pins within Intel's newer Land Grid Array sockets -- to say
vrs AMD's conventional Pin Grid Array. I suggest at a minimum a
pinhead and x30 lighted magnifier.)

Solid state capacitors don't bulge, not that I'm entirely sure my SS
mbs are conclusively impervious. But, I'll get back with you first
thing, after another 20 years, when I know otherwise. For sure.

Part of buying a MB is carefully researching supported CPUs, both on
the manufacturer site and if necessary outside from a catchbag of
keywords over forums, to include and to a lesser extent memory modules
supported. Video and such as USB, features, then, MBs vie on their
ostentatious boxes to marketing competition, should as well be
apprised with a mind to future operability and satisfaction. As only
deserving, since apples are apples and CPUs and memory are equally
equitable among orange flavorings. I like eggs in the raw in mine.
 
Thanks again.

Am I right in assuming that the definition of a motherboard is the board (mobo)
where all removables have been removed, or in the case of a new mobo not installed.
In other words, a motherboard does not include memory modules, cards, the processor,
or the battery. Fortunateley, I have these on my existing board. Sometimes I note
when viewing motherboards that the processsor socket has a plate in place. I assume
that there is no processor underneath if they are advertising a motherboard. Are my
assumptions correct?

I did not see any leaking, bulging, or split capacitors (an experienced engineer
might see someting), but any faults could easily be internal.

Best wishes

The motherboard is the square or rectangle of material.
It does not include a processor, or memory modules. It
does include the small CMOS battery (the CR2032 coin cell).
That is included. Those batteries last anywhere from
three to ten years, depending on computer usage pattern.
For an older board, you may end up going to Radio Shack,
or just about any drug store, and buying another CR2032.

The CPU socket, depending on type, comes with a PNP cap.
That's to keep dirt out of it. An LGA775 might come with
such a cap. For boards pulled from old computers, they
might not have any matching caps for them. In such a case, the
outer packaging materials would need to be robust, to protect
the socket area from getting crushed.

*******

There are two types of cylindrical capacitors. Aluminum electrolytic (wet)
and polymer caps. The electrolytics have a stress relief pattern
stamped into the top of the metal body. If excessive pressure
builds up inside, the metal fails along the stamp marks in
the top. That's to prevent the pressure from becoming too high.

In school, we got a demo of what happens. And the professor
doesn't do this, one of your fellow students does. If you
reverse the polarity on an electrolytic (plug it into the
board rotated 180 degrees from correct position), the
capacitor immediately pressurizes. It pressurizes so fast,
the stress relief mechanism can't work in time. It explodes,
filling the air with black sooty looking streamers. One of
the students next to me, did something like that on purpose,
to another student protoboard. So I've actually seen one
of them fail. What a mess.

The cap is an aluminum can, with a rubber plug in the bottom
end of it, to seal it. If the seams on top don't open when
they're supposed to, internal gas pressure can also force
the rubber plug to dislodge from the bottom. When doing visual
inspection, if you see a cap "tilted" at an angle, and see
something black wedging its way out of the cap, that's the
rubber plug working loose.

In a normal universe, we simply wouldn't know all this
stuff, or have experienced it. Only the "exploding cap"
in a student lab, that would be the only outward sign they
have a failure mechanism. But this article, describes how
some industrial espionage went wrong, and someone stole a
formula for electrolyte, and got it wrong. The pH of the
liquid inside the cap ends up wrong, and then things start
to corrode.

http://en.wikipedia.org/wiki/Capacitor_plague

The "plague" caps, tend to end up with the rusty colored
stuff on top. My Antec ATX power supply, the caps had
this look to them, on top.

http://upload.wikimedia.org/wikipedia/commons/f/f5/Al-Elko-bad-caps-Wiki-07-02-17.jpg

The Nichicons in the Dell systems that failed, that
probably wasn't the bad electrolyte, but some
manufacturing problem that I haven't seen described.
Nichicon is Japanese, and the Japanese didn't use
that electrolyte (because they had their own formula).
It was a number of Chinese manufacturers that decided
to use the badly made electrolyte, in their products.

Polymer caps, don't have the seams on top.

http://www.gigabyte.com/microsite/270/images/feature_04.jpg

( http://www.gigabyte.com/microsite/270/images/features.html )

Boards exist, where all capacitors are polymer type.
The next tier down of motherboard, the capacitors
under some stress are polymer, and the others
remain electrolytic (that's a "half n' half" board).
The cheapest of boards, to this day, still use all
electrolytic caps. And there is really nothing wrong
with that. One of the North American companies making
electrolytics, say they can last anywhere from 15 to
17 years, under ideal circumstances. And that the end
of life is caused by the rubber plug in the bottom, no
longer providing a hermetic seal, and the moisture in
the cap evaporates. If you were in a high ozone environment,
I expect the rubber plug would fail a lot faster, leading
to a good electrolytic, drying out.

There is an electrolytic capacitor in my electric lawn
mower, which functions as a "boost cap" for the motor.
And my lawn mower is over twenty years old and still
works fine. So if the electrolyte formulation is good,
they remain stable inside.

Paul
 
My motherbaoard has blown and am looking for a repalacement. It can be second-hand or refurbished - the chepaer the better.



It is to replace a motherboard in a Dell Dimension E 520 (LGA 775, Pentium D 915 2.8GHz processor, chipset probably G965, power supply output 305W. Form factor BTX



I am considering the DEll motherboard, but I'd like something cheaper, my machine is quite old.



Any references greatly received



PS. UK references

I see the aluminium solid state capacitor - What is the blue flash on top for?

On the video card there are some light green things with a different burst pattern on top. I assume these are also capacitors?

Are capacitors for regulating the current? There a quite a lot of them.

Are those things with copper turning wires voltage regulators?
 
I see the aluminium solid state capacitor - What is the blue flash on top for?

On the video card there are some light green things with a different burst
pattern on top. I assume these are also capacitors?

Are capacitors for regulating the current? There a quite a lot of them.

Are those things with copper turning wires voltage regulators?

In this picture, the polymer caps have a blue polarity marking on
the top. That helps tell the installer, which leg is positive
and which leg is negative. The polymer caps have no pressure
relief pattern on the top. The electrolytic with the "X" stamped
in the top, that one is set up for pressure relief, so it won't
explode in a way that hurts someone.

http://www.ixbt.com/video3/images/g92-9/ecs-9500gt-front.jpg

The regulation functions on the video card, are similar to
those on the motherboard. One difference, might be the amount
of power that needs to be handled.

*******

To explain what the L's and C's do in the circuit, you can try
an article like this one. It's mainly focused on the output
side of the circuit (output L and output C, and the switching
MOSFET).

http://www.radio-electronics.com/in...step-down-buck-regulator-converter-basics.php

The capacitor cans, give the "C" in the diagram. The powered iron toroids
with the copper wire wound on them, are inductors or L.

Neither of them are "regulators" by themselves. They are
analog circuit primitives, to be used in groups, wired
a particular way.

The capacitor smooths the output, and provides a reservoir. The
capacitor even ends up with an AC current flowing through it. And that
current, is related to the ripple rating (how much ripple current
it can handle). They use multiple capacitors in a group, to enhance
the total ripple current rating. If you used only one cap, it might
heat up and die prematurely.

The inductor is a reservoir as well, only it stores energy in the
magnetic field. So in a sense, like some kind of engine, you
go from pumping up the inductor, to transferring what the inductor
has achieved, into the capacitors. And the chip controlling it all,
monitors how the circuit is doing, whether the pulse width needs
to be extended on the next cycle, and so on. This allows the conversion
of a high voltage (like 12V on the yellow wires), to around 1.0 to 1.5V
for the processor.

The "switch" symbol in the diagram, is performed by MOSFETs. Like a
lot of other transistor types, there are three contacts on it. The gate
has a fairly high input resistance. The source to drain path, when the
device is turned on, has a very low resistance, making devices like this,
close to being ideal switches.

http://commons.wikimedia.org/wiki/File:D2PAK.JPG

To see what a "real" circuit looks like, try page 11 here. This is the
regulator for the processor on my P4 motherboard. It is a four phase
type, like a four cylinder engine. Each phase contributes a quarter
of the required current. So it's like the Radio Electronics diagram,
only times four. The little buffer chips, one per phase, amplify the
switching control signals from the main chip (so the main chip doesn't
get too hot, and each buffer chip warms up instead).

ADP3180 Switching Regulator - sample schematic page 11

http://web.archive.org/web/20040331...oadedFiles/Data_Sheets/129783535ADP3180_0.pdf

The filter capacitors are in two banks. C1 through C6 are the input bank.
The output bank is C21 through C28. Filtering on the input bank,
is to smooth the load on the ATX power supply. Filtering on the output,
improves transient response (when the regulator can't respond in time,
and the capacitors fill in the gaps).

This is the regulator from my first motherboard. The schematic is on
page 12. This is a single phase circuit, with only one inductor on
the output side. Once you collect enough of these, and read the
description, you get some sense of how they work.

http://web.archive.org/web/20010613105224/http://www.intersil.com/data/FN/FN4/FN4567/FN4567.pdf

It's pretty hard to find an EE class quality description.
So those datasheets will have to do.

*******

Paul
 
Reagarding the cooling fins on the card mentioned, I assume that there must be a
chip under the fins that requires cooling.

No, they just put the fins there to make the card look pretty.
 
I see the aluminium solid state capacitor - What is the blue flash on top for?

On the video card there are some light green things with a different burst
pattern on top. I assume these are also capacitors?

Are capacitors for regulating the current? There a quite a lot of them.

Are those things with copper turning wires voltage regulators?

Why don't you Google for answers to your questions, instead of expecting
them to be handed to you on a plate?
 
En el art�culo <e986fdd9-5ecd-4567-9024-ff323192280.com>,

.net escribi�:







Why don't you Google for answers to your questions, instead of expecting

them to be handed to you on a plate?



--

(\_/)

(='.'=)

(")_(")

It has nothing to do with you.
 
In this picture, the polymer caps have a blue polarity marking on

the top. That helps tell the installer, which leg is positive

and which leg is negative. The polymer caps have no pressure

relief pattern on the top. The electrolytic with the "X" stamped

in the top, that one is set up for pressure relief, so it won't

explode in a way that hurts someone.



http://www.ixbt.com/video3/images/g92-9/ecs-9500gt-front.jpg



The regulation functions on the video card, are similar to

those on the motherboard. One difference, might be the amount

of power that needs to be handled.



*******



To explain what the L's and C's do in the circuit, you can try

an article like this one. It's mainly focused on the output

side of the circuit (output L and output C, and the switching

MOSFET).



http://www.radio-electronics.com/in...step-down-buck-regulator-converter-basics.php



The capacitor cans, give the "C" in the diagram. The powered iron toroids

with the copper wire wound on them, are inductors or L.



Neither of them are "regulators" by themselves. They are

analog circuit primitives, to be used in groups, wired

a particular way.



The capacitor smooths the output, and provides a reservoir. The

capacitor even ends up with an AC current flowing through it. And that

current, is related to the ripple rating (how much ripple current

it can handle). They use multiple capacitors in a group, to enhance

the total ripple current rating. If you used only one cap, it might

heat up and die prematurely.



The inductor is a reservoir as well, only it stores energy in the

magnetic field. So in a sense, like some kind of engine, you

go from pumping up the inductor, to transferring what the inductor

has achieved, into the capacitors. And the chip controlling it all,

monitors how the circuit is doing, whether the pulse width needs

to be extended on the next cycle, and so on. This allows the conversion

of a high voltage (like 12V on the yellow wires), to around 1.0 to 1.5V

for the processor.



The "switch" symbol in the diagram, is performed by MOSFETs. Like a

lot of other transistor types, there are three contacts on it. The gate

has a fairly high input resistance. The source to drain path, when the

device is turned on, has a very low resistance, making devices like this,

close to being ideal switches.



http://commons.wikimedia.org/wiki/File:D2PAK.JPG



To see what a "real" circuit looks like, try page 11 here. This is the

regulator for the processor on my P4 motherboard. It is a four phase

type, like a four cylinder engine. Each phase contributes a quarter

of the required current. So it's like the Radio Electronics diagram,

only times four. The little buffer chips, one per phase, amplify the

switching control signals from the main chip (so the main chip doesn't

get too hot, and each buffer chip warms up instead).



ADP3180 Switching Regulator - sample schematic page 11



http://web.archive.org/web/20040331...oadedFiles/Data_Sheets/129783535ADP3180_0.pdf



The filter capacitors are in two banks. C1 through C6 are the input bank.

The output bank is C21 through C28. Filtering on the input bank,

is to smooth the load on the ATX power supply. Filtering on the output,

improves transient response (when the regulator can't respond in time,

and the capacitors fill in the gaps).



This is the regulator from my first motherboard. The schematic is on

page 12. This is a single phase circuit, with only one inductor on

the output side. Once you collect enough of these, and read the

description, you get some sense of how they work.



http://web.archive.org/web/20010613105224/http://www.intersil.com/data/FN/FN4/FN4567/FN4567.pdf



It's pretty hard to find an EE class quality description.

So those datasheets will have to do.



*******



Paul

Paul,

Why do designers use electrolytic and polymer based capacitors on single boards. I would expect them to use one or the other?
 
Why do designers use electrolytic and polymer based capacitors on single boards.
I would expect them to use one or the other?

If the polymer caps cost a penny more, then the motherboard designer
will only use them were absolutely necessary.

The VCore regulator is a stressful circuit. With a need for a high
ripple rating, and multiple capacitors. "Amps" of current, flow
through those caps. That means, if Polymer caps are to be used at all,
they'll first appear around the CPU socket. Or, perhaps
for the VDimm regulator circuit.

Now, there are also caps on the board, for "filtering of power rails".
These caps are sprinkled around the board, and their location is
less critical. They might be termed "bulk capacitance". The amount
of ripple current flowing through them, is small. And, it is small,
because the ATX power supply already regulated out most of the AC behavior.
So DC is placed across the bulk bypass caps, and they aren't really
stressed at all. If a PCI card draws a slug of current all the sudden,
the bulk bypass cap fills the void, giving the ATX supply time (maybe
a millisecond), to respond in a useful way.

For those "less stressed" or "bulk" capacitors, you can use whatever
you want. If an electrolytic is a penny cheaper, then in they go.

On enthusiast motherboards, the marketing department will tell
engineering, to use nothing but polymer.

If the motherboard only retails for $50, then you'll find it
filled with electrolytics.

And to repeat, there is nothing wrong with electrolytics. The reputation
of electrolytics, was ruined by the "bad electrolyte" episode. And that's
why we have such a mix of implementations.

*******

Even on video cards, you can find the same kind of split thinking. But
video cards at least, were using polymer before the plague happened,
So in fact their introduction was driven by implementation issues. If
the product is high end, you could well find polymer, or even ceramic,
used for power regulation circuits.

*******

On this video card, an unconventional regulator is used for the GPU
core power. There are Volterra regulator chips. And notice the lack of
tin cans near it. Just a few tin cans, for bulk bypass. The energy
storage in this case, may be done with relatively low profile ceramic
caps. Which can be very expensive. (I managed to buy some recently,
for $0.27 a piece, but they can be a lot more than that.)

http://images.ht4u.net/reviews/2009/amd_radeon_hd_4890/front_blank.jpg

http://hothardware.com/image_popup.aspx?image=big_5970_slide-3.jpg&articleid=1417&t=a

And this motherboard, uses a Volterra setup and ceramic caps, just above
the CPU socket.

http://images.bit-tech.net/content_images/2008/11/first-look-dfi-lanparty-x58-t3eh8/mobo-s.jpg

Another Volterra setup.

http://www.techpowerup.com/reviews/Computex2006/DFI/images/pwm.jpg

And I found something interesting here. How embarrassing.

http://ixbtlabs.com/articles3/mainboard/evga-nforce-790i-sli-ftw-digital-pwm-nforce790i-sli-p1.html

I thought the chips with the "C" logo, were regulators, when in fact
they contain the coils. The Volterra stuff is above that. That's why
you can't identify any traditional coils. Those circuits run at a much
higher frequency than a regular VCore regulator, a frequency so high, that
electrolytics would no longer be appropriate (electrolytics are really
crude devices, and ill-suited to life above 1MHz). Even ceramics have
their limitations (there is some variation in quality and circuit parameters
- the ones I bought for $0.27 have rather poor parameters and high ESR).

I got a hint, from glancing at the first page of this paper. It
says "digital VRM", but the functional block still shows the usage
of L and C. And then, looking at the info in the ixbtlabs photo
and caption, I can see I was looking at the wrong components.

http://www-power.eecs.berkeley.edu/publications/xiao_peterchev_sanders_pesc.pdf

Paul
 
My motherbaoard has blown and am looking for a repalacement. It can be second-hand or refurbished - the chepaer the better.



It is to replace a motherboard in a Dell Dimension E 520 (LGA 775, Pentium D 915 2.8GHz processor, chipset probably G965, power supply output 305W. Form factor BTX



I am considering the DEll motherboard, but I'd like something cheaper, my machine is quite old.



Any references greatly received



PS. UK references

Hi,

The diagrams show clearly the ceramic capacitors.

I am looking at a WG864 motherboard at the moment.

All Wg864 motherboards that I have seen have a black SATA connector (SATA4). However, this one has a white connector. Do I need to worry about this?

I have checked the the DS/N number, it agrees with mine up to CN 0WG864-4811 - then disagrees. I assume that this disagreement is due to the fact that it is a different motherboard (unique number). Am I correct?

The cpu coverplate looks a little different to mine, it is solid and has what looks like white dashes around the edges! I don't think I need to be concerned.

Best wishes
 
Hi,

The diagrams show clearly the ceramic capacitors.

I am looking at a WG864 motherboard at the moment.

All Wg864 motherboards that I have seen have a black SATA connector (SATA4).
However, this one has a white connector. Do I need to worry about this?

I have checked the the DS/N number, it agrees with mine up to CN 0WG864-4811 -
then disagrees. I assume that this disagreement is due to the fact that it
is a different motherboard (unique number). Am I correct?

The cpu coverplate looks a little different to mine, it is solid and has
what looks like white dashes around the edges! I don't think I need to be
concerned.

Best wishes

A quick Google, says the DS/N is the first part of the PPID, also
known as a serial number. So the reason the trailing digits would
differ, is that is part of the serial number that is unique. Try a
Google, and see if you can get any more info about the formulation
of "Dell PPID".

The SATA ports are color coded, for functions such as RAID. Say there
were two controllers, one controller had dual RAID ports, then they'd be
the same color.

When you see a Dell with four different SATA colors for the four ports,
that implies to me they are non-RAID, and limited to IDE or AHCI modes.
Motherboards with six SATA connectors, are more likely to be RAID capable.

If they change colors in midst of production, all that matters is whether
the replacement connectors match colors the same way. If two ports were
black on a previous production run, then it would be OK for the same
two ports to be yellow. It's to help you identify things like RAID ports,
if they're present. Or, if the system has an ESATA port, ESATA adapter plate
and cable, an "odd" color may be used to identify the sixth port reserved
for that function (some AMD boards do that).

The Pick and Place Cap (PNP cover) is intended to protect Intel LGA
sockets from damage. When the CPU is not present, you put the cover
back in place. So don't throw it away. For example, if sending a
motherboard to Asus for repair/warranty, if it arrives with no PNP
cap in place, the warranty claim may be rejected. Same goes for
warranty claims, where the LGA socket springs are bent. That's
considered to be "customer damage", even though there are plenty of
cases where the shipper is what caused the damage.(Even with a PNP
cap, there have been cases of the socket getting damaged in transit.
Somehow.)

It doesn't matter if the PNP cover is a different color or not, as
long as it is the right size to protect the socket.

The older "ZIF" sockets, don't need protection, because the top
plastic layer on those provides a good deal of protection on its own.
You can rip the top off a ZIF, but it takes an effort.

The Intel LGA is surprisingly good, for what it is. When it
first came out, I was predicting it would be a disaster, but
it turned out fine. Some other company tried a similar idea,
and theirs stunk (lots of field service trips for techs,
when the processor crashed). So Intel did some good work when
they did theirs. I was expecting the LGA2011 would be a disaster,
due to the large number of contacts, but I haven't heard any serious
complaints there. Perhaps because not too many people can afford them.

Paul
 
Back
Top