Booting Problem

S

species8350

When I try to start the tower the bios, represented by a horizontal bar, goes about one-quarter the way across and then stops.

How can I investigate and solve the problem?

Thanks

Ps Other details: tower is a Dell Dimension E520, with 1GB oif RAM, Pentium D 2.8GHZ

OS is Vista Home Premium, 32 bit
 
P

Paul

species8350 said:
When I try to start the tower the bios, represented by a horizontal bar, goes about one-quarter the way across and then stops.

How can I investigate and solve the problem?

Thanks

Ps Other details: tower is a Dell Dimension E520, with 1GB oif RAM, Pentium D 2.8GHZ

OS is Vista Home Premium, 32 bit

Are you referring to the OS boot bar, shown here at 25 seconds ?


Does the box have F8 Safe Mode ? Selecting "Last Known Good..."
is one thing you can try. If that didn't work, you can enable
boot logging, which is supposed to write out C:\windows\Ntbtlog.txt

http://pcsupport.about.com/od/fixtheproblem/ss/safe-mode-vista_2.htm

http://www.vistax64.com/vista-installation-setup/134105-boot-log.html

With the Ntbtlog.txt file on disk, now you need an environment to
read that text file. (A Linux LiveCD will do. I don't know what you
could use in the Vista recovery console, to read a text file.)

With Ntbtlog.txt, the annoying part is, the thing that's failing,
is "one line after" the last line in the file. The file is basically a
list of the things already handled, and where it is dying, is in something
which didn't make a log entry. So then you need to Google the last
good line, and see if anyone has figured out what the line after
that one might be.

The Vista recovery CD, may have the ability to attempt to repair the
boot. But really, your boot is working after a fashion (the bar showed up),
so it isn't a bad BCD file. It could be some driver that got messed up.

I guess at the moment, all I can suggest is "Last Known Good", or
at least give some indication what you did to the thing before
the last shutdown. Maybe what you're seeing, is the effects of
some Windows Update that just got applied.

I don't think the BIOS drew that bar. The last thing a legacy BIOS
does, is present a black screen with a flashing cursor in the upper
left hand corner. You're not supposed to see that screen, because
the OS has taken over. The OS is reading the disk using Extended INT 0x13
(no disk driver loaded yet), and graphically what is on the screen,
is the doings of the OS (OS boot animation).

Paul
 
P

philo 

Are you referring to the OS boot bar, shown here at 25 seconds ?


<snip>

I am pretty sure the OP actually did mean "BIOS"

On a Dell, the BIOS displays a horizontal progress bar.



If the BIOS freezes like that there is some type of hardware problem...

My first guess would be that it's trying to detect the hard drive...
and due to a defect, locks up.


Open the machine and disconnect the hard drive and see if the BIOS at
least loads to completion...(you'd get en error message such as "no boot
device" )
 
P

Paul

philo said:
<snip>

I am pretty sure the OP actually did mean "BIOS"

On a Dell, the BIOS displays a horizontal progress bar.



If the BIOS freezes like that there is some type of hardware problem...

My first guess would be that it's trying to detect the hard drive...
and due to a defect, locks up.


Open the machine and disconnect the hard drive and see if the BIOS at
least loads to completion...(you'd get en error message such as "no boot
device" )

I tried to find a video of E520 booting on Youtube, but couldn't find
anything relevant.

Paul
 
S

species8350

When I try to start the tower the bios, represented by a horizontal bar, goes about one-quarter the way across and then stops.



How can I investigate and solve the problem?



Thanks



Ps Other details: tower is a Dell Dimension E520, with 1GB oif RAM, Pentium D 2.8GHZ



OS is Vista Home Premium, 32 bit

Thanks for responding.

The video (after 25 secs) show the bars - mine locks up about one-quarter the way across.

I think I have a heardware problem, but don't know which piece is failing. I'll try disconneting the hard-drive, and see what happens. I assume that I just need to pull a cable (not done this before).

I was wondering if it could be be the motherboard or the RAM (have two 512MB modules).

Can I test the motherboard and RAMs?

At present, the systems just lock up at the BIOS stage, no keys work, nothing works.

Best wishes.
 
P

Paul

species8350 said:
When I try to start the tower the bios, represented by a horizontal bar, goes about one-quarter the way across and then stops.

How can I investigate and solve the problem?

Thanks

Ps Other details: tower is a Dell Dimension E520, with 1GB oif RAM, Pentium D 2.8GHZ

OS is Vista Home Premium, 32 bit

http://www.dell.com/support/Manuals/us/en/19/Product/dimension-e520

ftp://ftp.dell.com/Manuals/all-products/esuprt_desktop/esuprt_dimension_desktops/dimension-e520_service%20manual_en-us.pdf

The first time I downloaded the Service Manual, the file was truncated.
A second attempt to download, got a 2,429,989 byte file and I can
read that one OK.

On page 19, your computer uses diagnostic LEDs. Four of them. And
the manual has the codes for decoding what is showing. The LEDs
would only be driven intelligently, if the processor is running.
(And since the display is doing something, that is likely to be
the case.) The diagnostic LEDs cease to have meaning, once the
boot starts (as then, handoff is to the OS or boot CD).

So you can check for diagnostic LEDs as well.

The manual also mentions "Dell Diagnostics", but if those
are loading off the hard drive, that might not work either.

*******

OK, managed to find a boot sequence for E520. It's very fast.

"Windows 8 Pro 64bit -- Dell Dimension E520"


I selected three frames from the video, posted here.
As Philo notes, the first animation is early BIOS. The
second screen there, seems to be the Intel extension ROM for
running the SATA ports. The third picture, is Windows 8 starting
on the machine (on a Core2 processor). So if only a quarter of
the animation bar is showing, that must be a failure pretty early on.

http://img560.imageshack.us/img560/3660/dimensione520boot.jpg

The BIOS has an option, to set FastBoot to OFF, and maybe
that will extend the initial tests a bit.

I could find one poster who had a bad CMOS battery, and those
go bad faster, if you turn off the PC using the power switch
at the back, every day, after you finish using it. If the battery is
good, taking a reading off the top of the battery, with a multimeter,
and measuring with respect to chassis ground, should read at least 3 volts.
(Battery is probably a CR2032, as they're pretty common. And CR2032
is mentioned in the manual.)

Diagnostic facilities:

1) Diagnostic LEDs (4 LEDs, 16 possible codes, 12 documented)
2) BIOS beep codes, using motherboard beeper (documented in manual)
Beeper is near the main power connector.
3) Text underneath animation bar (but I don't see a way of turning
off the appearance of the initial animation bar. Try disabling
Fastboot, and hope something changes.
4) Dell diagnostic mode or software (don't know the details)

Paul
 
P

Paul

Grinder said:
E520 booting at 1:25 in this video:

That's quite the workshop the guy's got.

And he's getting good stuff from his dumpster, too.
If somebody left a PC out for the garbage here, it
would be a Pentium II.

Paul
 
S

species8350

When I try to start the tower the bios, represented by a horizontal bar, goes about one-quarter the way across and then stops.



How can I investigate and solve the problem?



Thanks



Ps Other details: tower is a Dell Dimension E520, with 1GB oif RAM, Pentium D 2.8GHZ



OS is Vista Home Premium, 32 bit

Let me clarify,

The green bars are the bars that I would expect to see if I can get past the BIOS stage.

The horizontal bar (BIOS) goes about one quarter way across the screen and then stops.

No key will function.

I would like to test the motherboard and the RAM (I have two 512MB modules).

Can anyone tell me how to test the above?

Thanks
 
S

species8350

When I try to start the tower the bios, represented by a horizontal bar, goes about one-quarter the way across and then stops.



How can I investigate and solve the problem?



Thanks

Paul

Thanks for the manual.

I am trying to identify the lights
 
S

species8350

When I try to start the tower the bios, represented by a horizontal bar, goes about one-quarter the way across and then stops.



How can I investigate and solve the problem?



Thanks



Ps Other details: tower is a Dell Dimension E520, with 1GB oif RAM, Pentium D 2.8GHZ



OS is Vista Home Premium, 32 bit

Just checked the diagnostic lights.

After they stopped flashing, the stable condition is no lights for the first two, and green for the last two.

According to the manual this means faulty RAM

Previously, I tried removing one RAM module and switching on, but it would not boot. I tried this module in the other slots, but no luck. I repeated this for the second module with the same result.

Can I confirm that it is a RAM problem (from my tests it looks doubtful - it would mean that both modules are faulty - unlikely).

Thanks
 
P

Paul

species8350 said:
Just checked the diagnostic lights.

After they stopped flashing, the stable condition is no lights for the first two, and green for the last two.

According to the manual this means faulty RAM

Previously, I tried removing one RAM module and switching on, but it would not boot. I tried this module in the other slots, but no luck. I repeated this for the second module with the same result.

Can I confirm that it is a RAM problem (from my tests it looks doubtful - it would mean that both modules are faulty - unlikely).

Thanks

It's unlikely both DIMMs would be bad at the same time, if
it was a DIMM fault.

Standard practice would be, to find another compatible DIMM,
and test it in the motherboard, and see if the symptoms change.

For both DIMMs, tested individually, to give faults, it's
more likely the VDimm regulator isn't working, or something
related to the Northbridge (Vnb for example). In other words,
a bad motherboard.

But the fact that the screen lights up, and starts to paint
the animated bar, means much of the chips is working. That's
the symptom which doesn't make much sense. A "solid" RAM failure,
should have given a black screen at startup. Not the green bar
on your screen. To get the green bar on the screen, the system
would have to be 99.9% working.

If the BIOS is running a RAM test, it could certainly find an
error in the upper areas of memory (above where the frame buffer
image is stored, at some location like 0xA0000). But what are
the odds, that both DIMMs have bad locations, suddenly detected
at the same point in time, such that either of the DIMMs
tested individually, give the same symptoms ? That part, just
doesn't make sense.

Maybe the error code is "bad RAM", but in fact it's some other
fault entirely, and the BIOS is just... nuts.

*******

You could take the side off the machine and inspect the
capacitors. Those are cylinders with a plastic sleeve over
them. Look for signs of bulging (related to internal pressure)
or leaking of electrolyte from inside the capacitor. Older
motherboards use aluminum electrolytic capacitors. More modern
boards use solid polymer capacitors. The solid polymer ones,
don't have pressure relief seams on the top of the capacitor,
implying there is less chance of pressure buildup with the
solid electrolyte.

There is nothing really wrong with the usage of aluminum
electrolytics. They're perfectly good. It's because billions
of them were made improperly, that we have a problem with them.
The introduction of solid polymer ones, to take their place,
is a "face saving" move by the motherboard manufacturer, a way
of saying "well, you can see we don't have any bad caps, because
they aren't the wet electrolyte type any more". If all the capacitors
were properly made, the older style cap with the pressure seam
on top, could have continued to be used. Some Dell models, had
a tremendous dropout - virtually all the motherboards failed
due to that problem. To such an extent, certain Dell models
are not safe to buy, used, on Ebay.

Here is a background article on the issue.

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

Your machine is relatively modern, and the newer the machine
is, the less likely it is to be a capacitor problem. There are
still motherboard companies using bad materials, but becoming
less and less every day. There was a large inventory of bad
product out there, and the motherboard makers have to invest
effort in preventing the bad capacitors from going into products.

Paul
 
S

species8350

When I try to start the tower the bios, represented by a horizontal bar, goes about one-quarter the way across and then stops.



How can I investigate and solve the problem?



Thanks



Ps Other details: tower is a Dell Dimension E520, with 1GB oif RAM, Pentium D 2.8GHZ



OS is Vista Home Premium, 32 bit

Thanks for replying.

I did correct my post.

I am not seeing the green bars, I do not get to that stage. I switch on the tower and get the BIOS indicator that travels about one quarter the way across, then stops. No keys function.

The diagnostics indicate bad RAM, but bearing in mind what I said regarding removing RAM modules and testing, and being unable to boot, that my problem is most likely to be caused by the motherboard.

Is there a way to confirm that the motherboard is faulty? Could the diagnostic indicate RAM when in fact the problem is the motherboard.

Can I prove either RAM or motherboard failure.

Thanks
 
S

species8350

When I try to start the tower the bios, represented by a horizontal bar, goes about one-quarter the way across and then stops.



How can I investigate and solve the problem?



Thanks



Ps Other details: tower is a Dell Dimension E520, with 1GB oif RAM, Pentium D 2.8GHZ



OS is Vista Home Premium, 32 bit

I have opened the case and noticed what I thought was a capacitor that was slanting. I pushed it into the vertical position.

I placed the two memory modules in position, and obtained the same result (BIOS to about one quarter way across).

I removed one of the RAM modules and re-started the tower (from the front button). The BIOS indicator crawled across the screen. I got the message: 'Ammount of memory changed...'.
Alert! OS install mode enabled... memory limited to 256 MB.

I removed this memory module and used the second module. I obtained the same result.

Are these observations indicative of a faulty motherboard?

Thanks
 
P

Paul

species8350 said:
I have opened the case and noticed what I thought was a capacitor that was slanting. I pushed it into the vertical position.

I placed the two memory modules in position, and obtained the same result (BIOS to about one quarter way across).

I removed one of the RAM modules and re-started the tower (from the front button). The BIOS indicator crawled across the screen. I got the message: 'Ammount of memory changed...'.
Alert! OS install mode enabled... memory limited to 256 MB.

I removed this memory module and used the second module. I obtained the same result.

Are these observations indicative of a faulty motherboard?

Thanks

OK, you had 2x512MB installed, then when one module is installed, it says
it sees 256MB. It should have said 512MB.

Now, that tells you something.

The BIOS has two ways to evaluate the RAM. The first method it uses, is
it reads the tiny SPD chip on the DIMM, and gets the details from there.
The SPD will tell the BIOS "this module contains 512MB". The BIOS assumes
that is the amount in question.

But hidden in the BIOS code, is a second method. It runs too. It's the
"I'm a skeptic" piece of code. It doesn't trust the SPD declaration.
The BIOS uses the old "peek n' poke" algorithm, to size the RAM. It
tries to write to address 512MB, goes back moments later and attempts
to read the value, and sees a different value. This tells the skeptic code,
that the RAM really isn't 512MB long. That code does a binary search of
sorts, and it crawls through the DIMM until it determines the locations
that are successfully storing data.

So the legacy algorithm ran, and determined it isn't a 512MB module.
This could happen, if an address line wasn't making good contact
on the DIMM. Simply reseat the memory (all power removed!!!).
Do not play with RAM, unless all power is removed! Once you're
absolutely sure the DIMM is snapped into place, only then plug in
the power cord, turn on the power and test it. The DIMM slot has
standby voltage on it, if you leave the thing powered. And the DIMM
could be damaged by that voltage, if you "hot insert".

And no matter how much RAM it finds, let it try to boot. If there
isn't enough RAM for the OS, let's hope it BSODs without
making too much of a mess :)

*******

It's obligatory for me to mention memtest86+ . This is a tool
you can use for testing computer memory. The program is self
booting, and needs no hard drive connected to the computer.
Scroll half-way down this page, to find the downloads.

http://www.memtest.org

Memtest86+ V4.20 (25/01/2011)

It can only test the amount of memory the computer "sees". In
your current situation, all it can test is the 256MB detected
as "good" by the BIOS.

If you cannot get more than 256MB detected, use a magnifying
glass and a good source of light, to inspect the DIMM slots
for a bent pin.

*******

This assumes you didn't pull a fast one, and plug a 256MB
DIMM in for testing :) I'm assuming you've plugged in a
single 512MB, and it says the thing is only 256MB in size.

HTH,
Paul
 
S

species8350

When I try to start the tower the bios, represented by a horizontal bar, goes about one-quarter the way across and then stops.



How can I investigate and solve the problem?



Thanks

After pushing that capacitor (near the RAM) upright the BIOS scrolled slowly across the screen. I then put the cover back on and inverted the tower back into the upright position. Later, I switched the computer on and found Iwas back where I started, the BIOS was stuck about one quarter the way across the screen.

I know that you think its a RAM problem, but I can't understand how I get the same result after swapping the 512MB module for the other 512 MB of RAM.

I assumed that the reading of 256MB was a result of failure of some sort.

You mention Memtest. The computer does not run, I can't get beyond the BIOS- nothing works.

As I reacall, all the connectors on the RAM modules looked fine.

I have re-inserted the RAM modules on many occasions.

Thanks
 
P

Paul

species8350 said:
After pushing that capacitor (near the RAM) upright the BIOS scrolled slowly across the screen. I then put the cover back on and inverted the tower back into the upright position. Later, I switched the computer on and found I was back where I started, the BIOS was stuck about one quarter the way across the screen.

I know that you think its a RAM problem, but I can't understand how I get the same result after swapping the 512MB module for the other 512 MB of RAM.

I assumed that the reading of 256MB was a result of failure of some sort.

You mention Memtest. The computer does not run, I can't get beyond the BIOS - nothing works.

As I reacall, all the connectors on the RAM modules looked fine.

I have re-inserted the RAM modules on many occasions.

Thanks

Don't bend that capacitor back and forth too much, because it could
eventually snap off.

Normally, a capacitor would be seated onto the board, rather than sitting
above it. If the rubber plug in the bottom of the capacitor works loose,
that's one way capacitors get tilted. When pressure builds up inside
a "bad" capacitor, either the seams in the lid open (stamped pattern on
top is for pressure relief), or the rubber plug in the bottom can work
itself loose.

Are all the fans working normally on the system ? An overheating
processor though, wouldn't account for all your symptoms.

The "insert 512MB, detect 256MB" still bothers me. Either the
module is really a 256MB module, or an electrical connection
is not making contact. That's all I can think of there.

I suppose the module could have two ranks, and one rank
is not responding (chip select line not working). Which
is another way that perhaps the problem could arise.

Paul
 
S

species8350

When I try to start the tower the bios, represented by a horizontal bar, goes about one-quarter the way across and then stops.



How can I investigate and solve the problem?



Thanks



Ps Other details: tower is a Dell Dimension E520, with 1GB oif RAM, Pentium D 2.8GHZ



OS is Vista Home Premium, 32 bit

The top of the capacitor looks fine.

The fan was working when the tower was working.

the RAM's are 512 MB, from previous information when the tower was working.

Previously, you mentioned pins. The memory modules consist of many golg plated pins, as a strip - no damage.

You did not mention the motherboard. Is it not possible that the motherboard close to the RAM is damaged and gives the impression that the RAM is damaged (that capacitor is close to the RAM). Is there a way that I can distingusih motherboard damage from RAM damage. Removing RAMs, or someting.

Previously, before the system achieved the state of not completing the BIOS, I could run an OS from the CD drive (VISTA would not start), however, I usually noticed kernel panics. I could also run a OS from a USB thumb drive,but again panics occured.

I mention the above cases in case they give any clues.

Best wishes.
 
P

Paul

species8350 said:
The top of the capacitor looks fine.

The fan was working when the tower was working.

the RAM's are 512 MB, from previous information when the tower was working.

Previously, you mentioned pins. The memory modules consist of many golg plated pins, as a strip - no damage.

You did not mention the motherboard. Is it not possible that the motherboard close to the RAM is damaged and gives the impression that the RAM is damaged (that capacitor is close to the RAM). Is there a way that I can distingusih motherboard damage from RAM damage. Removing RAMs, or someting.

Previously, before the system achieved the state of not completing the BIOS, I could run an OS from the CD drive (VISTA would not start), however, I usually noticed kernel panics. I could also run a OS from a USB thumb drive, but again panics occured.

I mention the above cases in case they give any clues.

Best wishes.

Panics could be bad memory.

Panics could be mis-adjusted memory as well. On a retail
motherboard, there are a multitude of timing adjustments,
and if you set those manually, you can mess up the memory.
A copy of memtest86+ can highlight this for you.

Your kernel panics are good enough evidence, that there is a
memory problem. If you boot a Linux LiveCD under moderate memory
issues, individual programs tend to die and the kernel continues
to run. I've found the kernel tips over when things are really bad.
(I tested that a bit, on an overclocked system, just to see what
would happen.)

A test procedure, is to find another DIMM and test whether
things work better with a different DIMM. Try the test DIMM
in different slots, to see if the symptoms change. Sometimes
it is one slot which is bad. I suspect something else is wrong
with the motherboard, but using a multimeter to check the
voltages isn't exactly easy. I've only seen one motherboard
design, which was clever enough to install monitor points so
a user can easily probe with a multimeter (check VDimm, check Vtt).
More typically, none of the critical voltages, are even labeled
on the motherboard.

And when inserting the test DIMM, make sure all power is off on
the system. I recommend unplugging it, in cases where the user
isn't sure about the power. If it's unplugged, you know for
sure there is no power present in the DIMM slot, when plugging
in the DIMM. My Asus motherboards, they all have a green warning
LED, and as long as the LED is lit, it is not safe to change
a DIMM. My Asrock motherboard, they don't put that LED. And on
that system, it's just safer to unplug it.

Paul
 
S

species8350

When I try to start the tower the bios, represented by a horizontal bar, goes about one-quarter the way across and then stops.



How can I investigate and solve the problem?



Thanks



Ps Other details: tower is a Dell Dimension E520, with 1GB oif RAM, Pentium D 2.8GHZ



OS is Vista Home Premium, 32 bit

I did try moving the memory modules into the other slots. I tried each RAM module in each slot, but it made no difference. Does this indicate a faulty motherboard? Unfortunately, I only have these two RAM's
 
S

species8350

When I try to start the tower the bios, represented by a horizontal bar, goes about one-quarter the way across and then stops.



How can I investigate and solve the problem?



Thanks



Ps Other details: tower is a Dell Dimension E520, with 1GB oif RAM, Pentium D 2.8GHZ



OS is Vista Home Premium, 32 bit

Regarding safety.

I do switch off the tower and unplug it from the socket before working inside the case.

I assume that when you were checking the voltages on the motherboard with your multimeter that had the tower switched on.

Can the motherboard be checked if voltages are unknown and there are no obvious checking points?
 

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments. After that, you can post your question and our members will help you out.

Ask a Question

Top