"safe mode" stops clock?!

[Dave Gower]

As I normally do after a surfing session, I did a restart, and the went for
supper. About an hour later I came back and found a screen I haven't seen in
the 5 months since I acquired my new computer with XP. It was a message box
telling me I was in "safe mode" and gave no details as to why except that I
had some kind of problem. I did another restart and it came back to normal
except that both the distributed computing project I run as a screensaver
and my clock had not budged. Now I can understand not running a screensaver,
but stopping the clock? I though that ran by itself.

Like I said everything seems normal now but what happened? I've been reading
this NG for months now and can't remember anyone else having this problem.
Is there something wrong with my new computer?

 

[johnf]

Possibly it's time to replace the battery on the MOBO?
They only cost a couple of $ at the Supermarket.
You didn't say whether the clock date & time were current, as an
intermittent battery will reset the clock back several years.

 

[Dave Gower]

Possibly it's time to replace the battery on the MOBO?
They only cost a couple of $ at the Supermarket.
You didn't say whether the clock date & time were current, as an
intermittent battery will reset the clock back several years.

The computer was new last September, and the clock simply stopped, it did
not go back in time (I remember when my old Toshiba laptop went back to Jan
1 1980 on Jan 1 2000).

 

[Sharon F]

Now I can understand not running a screensaver,
but stopping the clock? I though that ran by itself.

The Windows clock picks up the time from the RTC at startup. After that it
uses CPU cycles to advance. If a program locks up all of the CPU cycles, it
is possible for the clock to "get stuck."

 

[J&P Steele]

They only cost a couple of $ at the Supermarket.
You didn't say whether the clock date & time were current, as an
intermittent battery will reset the clock back several years.

If he used the same PC to send his message then date and time are current
and correct?

Regards
Joe Steele

 

[johnf]

If your clock's wrong the first thing you do is reset it to the correct
time - so, not a valid comment.
Anyway, that turns out not to be the problem. Any constructive ideas, JS?

 

[Parish]

The Windows clock picks up the time from the RTC at startup. After that it
uses CPU cycles to advance. If a program locks up all of the CPU cycles, it
is possible for the clock to "get stuck."

Every (AFAICT) time XP crashes with a BSOD (I've got a dodgy sound codec
I think) it screws my RTC clock. I've seen changes ranging from 10 hours
fast to 4 *years* slow!

Since I've got 3 machines running XP here and this is the only one that
does this I reckon it's partly/wholly related to the mobo/BIOS (Asus
A7M266-D/Award)

Oh, and for the benefit of anyone tempted to trot out the usual
"check/replace the CMOS battery", this machine runs 24/7 and the longest
it's ever been powered down in the 3 months since I built it is about
1/2 hour - and the clock didn't change then.

Regards,

Parish

 

[J&P Steele]

"johnf" <> wrote in message>

If your clock's wrong the first thing you do is reset it to the correct

time - so, not a valid comment.
Anyway, that turns out not to be the problem. Any constructive ideas, JS?

Wasn't suggesting it was the problem.

But if you have a read of the paragraph headed

"5/12/2003 RTC Word Attachments"

on the following site :-

http://nkcsystems.home.comcast.net/qacolumn.html

you may find a suitable explanation, perhaps.

Regards,
Joe Steele

 

[Sharon F]

Every (AFAICT) time XP crashes with a BSOD (I've got a dodgy sound codec
I think) it screws my RTC clock. I've seen changes ranging from 10 hours
fast to 4 *years* slow!

Since I've got 3 machines running XP here and this is the only one that
does this I reckon it's partly/wholly related to the mobo/BIOS (Asus
A7M266-D/Award)

Oh, and for the benefit of anyone tempted to trot out the usual
"check/replace the CMOS battery", this machine runs 24/7 and the longest
it's ever been powered down in the 3 months since I built it is about
1/2 hour - and the clock didn't change then.

Regards,

Parish

That's what makes computers interesting. There is a standard rule of thumb
to guide us and then exceptions to be considered as well.

I have yet to change the CMOS battery on any of my systems and some of them
are years old too. I have replaced batteries for others. On one system, the
battery was maintaining a charge when the computer was left plugged in.
Unplug the machine from an electrical source for any length of time and
that sustained charge dissipated quickly and settings were lost.

Software is usually at the root of the type of time changes but it's not
the only possible explanation. When XP was first released there was time
related trouble with one version of BIOS shipped on Dell machines. Time
jumped around forward and backwards, many hours at a time and while the
computer was in use. A registry change was created that rectified the
problem.

 

[Parish]

That's what makes computers interesting. There is a standard rule of thumb

Interesting? Bah, frustrating you mean. You want to see the bruises on
my forehead from repeatedly banging it on the desk

to guide us and then exceptions to be considered as well.

I have yet to change the CMOS battery on any of my systems and some of them
are years old too. I have replaced batteries for others. On one system, the
battery was maintaining a charge when the computer was left plugged in.
Unplug the machine from an electrical source for any length of time and
that sustained charge dissipated quickly and settings were lost.

Now that is strange, since the CMOS batteries aren't rechargeable.

Software is usually at the root of the type of time changes but it's not
the only possible explanation. When XP was first released there was time
related trouble with one version of BIOS shipped on Dell machines. Time
jumped around forward and backwards, many hours at a time and while the
computer was in use. A registry change was created that rectified the
problem.

This machine dual boots XP and FreeBSD, the latter (which is a
bleeding-edge version) changes the RTC in a bizarre, yet predictable,
way every time I shutdown. Since I haven't found anyone else with the
same problem (in FreeBSD) and XP does the same thing (albeit only when
it crashes) it kind of confirms that the root of the problem lies with
the BIOS which is a beta release. The last stable release doesn't
correctly identify Athlon MP2800s and claims they are not SMP capable. I
did read that this is only erroneous info on the boot screen and that
they run at the correct clock/voltage and both of them run so I might
ask in the Asus NG about flashing down to the release version.

Regards,

Parish

 

[johnf]

Now that is strange, since the CMOS batteries aren't rechargeable.

I seem to recall that quite a few board manufacturers used a Supercap in
lieu of a CMOS battery, which maintained its charge via a little bit of
low-voltage detection circuitry.
Many home-phones use the same system to hold the stored-number memory up and
people are actually told that their phone has a battery, as it causes too
many problems trying to explain what a Supercap is - (which is simply an
extremely efficient capacitor that retains its charge for a much longer
period than a normal capacitor).

For anyone interested, you can have a look at an example on -
http://www.st.com/stonline/products/families/memories/nvram/support/nvcl_idx.htm

johnf

FWIW, terminology nowadays is really up the spout. The actual CMOS 'battery'
comes with packaging referring to it as such, when in fact it is a single
cell - not a battery. The term 'battery' only should refer to a string of
cells, e.g. a 12v (nominal) car battery consists of 6 x 2v cells.