[Naki]

BIOS is Award. I have not been able to Google up anything useful yet.
PC specs:
* Motherboard: DFI AK-74 EC
* 900 MHZ Athlon CPU @ 855 MHZ.
* Via KT133 chipset.
* 512 MB of RAM (changing RAM doesn’t help)
* Tried 2 different AGP cards and one PCI, doesn’t help. AGP cards are GF2 MX and GF4 Ti4200. PCI card is Voodoo 3 2000.

Any idea what this means - when we turn it on, the PC sounds 3 long beeps, then turns off. Usually, after we turn it on for 20ish times, it turns on OK. A simple reboot produces the problem again.

[jaclaz]

Naki, on Dec 18 2008, 09:16 PM, said:

BIOS is Award.

Any idea what this means - when we turn it on, the PC sounds 3 long beeps, then turns off. Usually, after we turn it on for 20ish times, it turns on OK. A simple reboot produces the problem again.

http://www.wimsbios.com/index.jsp

http://web.archive.org/web/20020203070933/...rror_codes.html

jaclaz

[Naki]

jaclaz, on Dec 18 2008, 10:25 PM, said:

Naki, on Dec 18 2008, 09:16 PM, said:

BIOS is Award.

Any idea what this means - when we turn it on, the PC sounds 3 long beeps, then turns off. Usually, after we turn it on for 20ish times, it turns on OK. A simple reboot produces the problem again.

http://www.wimsbios.com/index.jsp

http://web.archive.org/web/20020203070933/...rror_codes.html

jaclaz

I don't think it is RAM as mentioned in the second link. Because we tried different RAM sticks in different slots (motherboard has 3 RAM slots).

The first link leads here
http://www.wimsbios....stbeepcodes.jsp
and the AWARD link is dead, as is the Eric's codes link.

This post has been edited by Naki: 18 December 2008 - 02:56 PM

[cluberti]

If I remember correctly, most AWARD post beep codes were video or RAM related. Anything that was 1 beep then another set of beeps was keyboard, video, etc - long beep codes were always motherboard or RAM issues. If you've swapped out video cards and used known-good RAM in here and it still won't POST, the motherboard is probably bad (specifically the memory controller is suspect).

[jaclaz]

Naki, on Dec 18 2008, 09:56 PM, said:

The first link leads here
http://www.wimsbios....stbeepcodes.jsp
and the AWARD link is dead, as is the Eric's codes link.

Sure, that's why I "translated" the link into a working archive.org cached page.

You asked if anyone had an idea to what the long beeps meant, I provided you a link to the not-so-easy-to-find- original page of the maker of the BIOS. Case closed.

You are free to think that they were lying or wrong, of course.

jaclaz

[puntoMX]

Naki, on Dec 18 2008, 02:56 PM, said:

I don't think it is RAM as mentioned in the second link. Because we tried different RAM sticks in different slots (motherboard has 3 RAM slots).

Well, if you tried the RAM and it had no effect, than your motherboard is dead enough; It would not see the RAM any more. I´m not surprised as it´s already a 9 year old system.

Naki, on Dec 18 2008, 09:56 PM, said:

The first link leads here
You asked if anyone had an idea to what the long beeps meant, I provided you a link to the not-so-easy-to-find- original page of the maker of the BIOS. Case closed.

You are right, just don´t finish him off .

[Naki]

Well, it seems it is a motherboard problem.
We put the two HDDs and videocard into another PC (NForce 2 chipset, Athlon XP 3000+, 2 GB of RAM).

We had to edit the system.ini file (VCache) and leave just 1 GB of RAM in order for the system to boot, because it is running Win98 SE. It is good that Win98 doesn't have the HDD Controller issues that WinXP Pro has when moving HDD to a new system.

Seems to be working fine now. Same PC has major problems with booting WinXP (motherboard damage problem it seems), but it boots Win98 SE fine. Some quirks AFTER booting into Windows, I will have to work them out...

This post has been edited by Naki: 19 December 2008 - 11:07 AM

[puntoMX]

There is nothing you can do, however, check the capacitors on the board; if they are faulty you could try to replace them.

[jaclaz]

puntoMX, on Dec 19 2008, 06:03 PM, said:

You are right, just don´t finish him off .

Naah, it would be cruel it must be a shock to post the same question on different boards :
http://www.neowin.ne...howtopic=711206
http://www.sudhian.c...wthread/105091/

and get everywhere the same bad news about the motherboard being somewhat fried.

jaclaz

[puntoMX]

Funny that I don´t know the last forum, but the guy there also says:

Quote

I’d check the capacitors on the mobo

EDIT: And teh guys at Neowin come up with this indeed:

Quote

sounds like a memory problem. that thing is old, lol

Funny...

[cluberti]

And I said it yesterday at 5PM. You're all LATE! to the party lol . Seriously though, hopefully the OP will be able to get a motherboard in that will work, and get back to using his machine.

[Naki]

PC is not mine, it's my mother's.
Hmm, both boards are DFI, both have issues...
The currently used mobo boots into Win2K and Win98 fine, but boots only one of about 10 times into WinXP, the other times it freezes on the logo screen. Once it boots, it seems to work fine.
(My mother uses Win98, thus avoiding the WinXP issue.)
There is also a problem with the ATA HDDs DMA (random, seems to work for long periods fine) and also problem with built-in LAN (so using PCI LAN card).

Maybe I should avoid DFI as a mobo brand. My own PC current mobo is Elitegroup (not the best of brands, I know...).

This post has been edited by Naki: 20 December 2008 - 04:45 PM

[puntoMX]

Naki, on Dec 20 2008, 04:44 PM, said:

Maybe I should avoid DFI as a mobo brand.

Well, they are just not made to work that long, nor are other brands. Remember that a computer-life is about 3 to 4 years, sure it still works after that but there is no warrantee it will work well...

[jaclaz]

puntoMX, on Dec 21 2008, 01:16 AM, said:

Remember that a computer-life is about 3 to 4 years, sure it still works after that but there is no warrantee it will work well...

Really?

Do you have any source for this?

If you do, please keep it for yourself.

I have a number of PC's working 24/7 since 2003 that know nothing about this fact and are still going strong......

@Naki
The symptoms you describe being intermittent, it is also possible that one or more solderings on the board have "cracked" and thus they don't make a good contact or the contact is on/off depending on "sheer luck", temperature, amount of current/voltage going through them.
In these cases, if you have LOTS of time and patience, an accurate inspection of the board with a mgnifying glass may reveal which ones have to be "re-flowed".


jaclaz

[cluberti]

Jaclaz, it's called MTBF, or mean time before failure, and all computer parts have a rating for this. Basically, you can expect most destkop-grade consumer parts to last for 3 - 5 years before they start to fail. Some parts fail before this, and some fail LONG after this. However, the expected useful life of most desktop consumer-grade computer parts (including parts ON the motherboard, for instance) is between ~5,000 and ~10,000 hours, depending on the vendor, the quality of the components, and how many subcomponents a part has (each of those would factor into the MTBF equation of the overall device as well). I've seen vendors claim 50,000, 100,000, even 1,000,000 MTBF hours for a component, but in reality about 8,000 hours is as long as most computer equipment (especially ones on for longer than 8 hours a day) should be expected to last.

Again, MTBF doesn't guarantee a part will last as long as it's MTBF number, and it also doesn't guarantee it'll fail anywhere near it if it lasts longer either (remember, it's an average, not a hard line in the sand). I choose to replace my main machines every 2 - 3 years or so, and yes I have some machines that are on year 5 or 6 of running, but I don't rely on these machines for anything critical because they are technically running longer than the manufacturer likely envisioned, and are out of warranty on pretty much everything as well.

Buying good quality components usually assures you of a long life (5 years+), but it's not a guarantee. Also, buying cheaper components (or getting components from a vendor that barely met QC) is usually a recipie for failure on or before a device's MTBF.

[jaclaz]

cluberti, on Dec 21 2008, 06:45 PM, said:

Jaclaz, it's called MTBF

Sure , being a bit technically knowledgeable, and an engineer by education and trade, I do know what a MTBF is.

But what I wished to see (and keep hidden to my PC's ) is data about actual MTBF declared by manufacturers AND the way they calculated this AND an actual database of past events to be able to analyze it's data statistically and find out what the actual Life of the object would be/have been, which is a totally different thing from MTBF.

Since you like Wikipedia articles, in computer matters, very often Service Life:
http://en.wikipedia....ki/Service_life
is shorter than expected MTBF but actual Life can greatly outnumber BOTH.

Besides, there is also the Planned Obsolence problem:
http://en.wikipedia....ed_obsolescence

At least theoretically a computer should be a Durable good:
http://en.wikipedia....ki/Durable_good

but MTBF does not make much sense or it is impossible to be calculated reliably on a complex system like the PC is.
It can - at most - be applied to parts of it:
http://stason.org/TULARC/pc/pc_hardware_fa...t_34_years.html
like an hard drive, but you don't really espect MTBF to be reliable and applicable to the particularly small sample (two units) the OP has:
http://www.tech-faq.com/mtbf.shtml

Here is a more accurate definition of MTBF:
http://www.relex.com.../prediction.asp

Quote

What is MTBF? There are many forms of the MTBF definition. In general, MTBF (Mean Time Between Failures) is the mean value of the lengths of time between consecutive failures, under stated conditions, for a stated period in the life of a functional unit. A more simplified MTBF definition for Reliability Predictions can be stated as the average time (usually expressed in hours) that a component works without failure.

"Computer-life" is nearer to Reliability than to MTBF, this might be of interest:
http://www.vicorpowe...ty/Rel_MTBF.pdf

But still it is a probability, and from an engineering or mathematical point of view, "3 to 4" years is a non-number, it's a "vague range", and applying it generically to ANY PC, regardless of:

• how many hours per day
  
• how many days per week
  
• how many weeks per year

it is on, regardless of:

• make/model/technology used
  
• type of applications running on it
  
• whether it is always on or swithed ON/OFF often
  
• quality of AC power ("straight", filtered, UPS)
  
• quality of environment (conditioned room/outdoor shed/cellar/your bedroom)
  
• maintenance performed (dust cleaning/checking fans/checking thermal paste where applicable)

and of course:

• luck

is simply inaccurate.

In other words the "3 to 4" years does NOT represent "computer-life", it is simply puntoMX's report on his personal experience, or that of his friends, or that of the places where he worked, of course it is worth consideration , but not in the least represents a one-size-fits-all value, at the most it could be called "rule of thumb" or "common sense" .

jaclaz

[cluberti]

Also being an engineer and 20 years in the business, I know how it works as well. And yes, you can get more than 5 years out of most good equipment, but we've all had hardware (sometimes even from "good" companies) that flakes out within 12 months.

[Ponch]

Naki, on Dec 20 2008, 11:44 PM, said:

Hmm, both boards are DFI, both have issues...
The currently used mobo boots into Win2K and Win98 fine, but boots only one of about 10 times into WinXP, the other times it freezes on the logo screen. Once it boots, it seems to work fine.
(My mother uses Win98, thus avoiding the WinXP issue.)
There is also a problem with the ATA HDDs DMA (random, seems to work for long periods fine) and also problem with built-in LAN (so using PCI LAN card).

I guess you mean Windows XP logo, (not DFI logo), in which case, the problem (on that one) migt be software or HDD hardware related.
Are all OSs on the same physical disk ?
I'm now entering the pedantic discussion about MTBF. In my experience, there are parts that are bad and fail quickly and parts that are good and go far longer than 8 years, so that "no warranty that they will wotk well" is to be taken both ways, there is no more warranty (except "commercial") that a computer will work well in it's first years either.
Take 100 lamps, wait for the 50th to blow, it says nothing at all about how long the remaining lamps will stay on.

[jaclaz]

Ponch, on Dec 22 2008, 10:57 AM, said:

In my experience, there are parts that are bad and fail quickly and parts that are good and go far longer than 8 years, so that "no warranty that they will wotk well" is to be taken both ways, there is no more warranty (except "commercial") that a computer will work well in it's first years either.

Yep , often first year covers part of the "downhill" section of the typical bath tub curve.

jaclaz

[puntoMX]

jaclaz, on Dec 21 2008, 12:53 PM, said:

In other words the "3 to 4" years does NOT represent "computer-life", it is simply puntoMX's report on his personal experience, or that of his friends, or that of the places where he worked, of course it is worth consideration , but not in the least represents a one-size-fits-all value, at the most it could be called "rule of thumb" or "common sense" .

Okay, so what do you really want?; A complete under build report of each component used on every computer part? That I can´t give you that but when I say 3 to 4 years than I´m talking about "general" life of a computer under "normal" use, without the screen and input devices. In accountancy the value is set to 0 (zero) in 3 years so, not only to be "obsolete" but also because of the MTBF AND CTO.

The numbers 3 to 4 years are taken from my own experience, as this is my profession for more than 12 years now, and my hobby for 15 years before that.

Remember that there are chemicals used that EVEN degrade when using or not using the equipment, and yes, some fail even out of there package... And there are more points...