Fail to Post - plus clear CMOS puzzle?? 3

[eponymous]

Gigabyte GA-6VX7-4x
p3 SL5DX (866/256/133/1.7v)
xp pro sp2

Fails to post / no response from monitor
Mobo LED / CPU & PSU fans work
Switched with spare and known working PSU - symptons same

Tested with only PSU / RAM / CPU connected

Tried clearing CMOS using mobo jumpers

Puzzle: now will only power up if jumper in CMOS clear position and not when returned to Default position? [still fails to Post]

Does anyone have any experience of this please and whether it helps pinpoint a fix?

Should it be possible to power up in both jumper positions?

Were I to try a straightforward clear by simply removing the battery could this resolve anything that may have gone awry when using the jumper?

Any help would be most appreciated.

Many thanks,
Eponymous

 

[wolluf]

most boards will not power up when CMOS jumper in clear CMOS position - but some do.

question - you say it WILL power up in this case - do you mean it will actually work properly (ie, boot operating system?)

Removing the battery is usually less effective than using the clear CMOS jumper - but no harm trying.

While clearing the CMOS will sometimes revitalise a 'dead' board, I find its often just a stay of execution - the board still dies shortly after.

Have you inspected the board for the usual suspects (leaking capacitors?)

I'd suspect your board is dead/dying.

 

[wahnula]

Removing the battery might help, take it out,hold the power button for 30 seconds, place the jumper in "clear", then return it to normal. Replace the battery and reboot. I'm not sure about running with CMOS jumper shorted, I would return it to its proper state. If there's no activity, something (I'm not sure what!) is up.

The sentiment from the 'Net is all over the board, from "it will kill it for sure" to "nothing happened when I did it". Google the string "CMOS jumper shorted". It would not hurt to replace the battery, but that only holds settings when the mains power is disconnected.

Double-check all the power leads, like the 4-pin CPU header. Looking at the board's specs, make sure you have a 2X-4X AGP card. If it's an 8X card it might not work. Also, it supports ECC RAM, uncommon in a desktop board. Are you using ECC RAM?

Try a different graphics card (known working) if you have one on hand. Any beeps from the mainboard during all of this? Try booting with no RAM, the system should protest with a flurry of beeps. Post back with updates.

Tony

Users helping Users...

 

[eponymous]

Wolluf,

Thanks for this. Have had a closer look at the manual and how the pins are identified on the board. The order of pins 1,2,3 are a bit contradictory. From the diagram in the manual, I may well have the jumper on default - whilst according to the board the jumper would seem to be on clear. (i.e. could just be misleading labelling.)

You may well be right regarding the board being on its last legs and shall make a visual check in daylight.

However, is there also a sequence of checks for testing failure to Post that you would also recommend, please?

Many thanks,
Eponymous1212

 

[eponymous]

Wahnula,

Thanks for this. Shall explore CMOS references in more depth (please see my message above re possible labelling problem).

re: power up mean just Mobo LED / CPU & PSU fans working (no post / no monitor response)

- tried swapping graphic cards ... no joy
- removed ram ... no beeps
- worked fine for years so can safely assume parts original/compatible.

it's not looking good is it? If it is saveable would like to have a go, if poss?

Plus, thinking back, may have been an emerging/intermittent problem. Started with no power up. Thought it was problem with power switch and replaced. Then OK for a couple of months. Now just suddenly stopped again. Realise now, probably part of the same problem.

Many thanks,
Eponymous

 

[edfair]

If you had a logic probe you could look at a data line in one of the PCI slots or address/data lines in one of the memory slots to see if you have any lines shifting.

To post you need power good coming out of the PS, that allows the clock circuits to work if the crystal is pumping.

This can be voltage issues on the board, the clock/crystal, or a multitude of other MB issues. Or a processor chip that has fried.

Keep in mind that spending a lot of time diagnosing it is good education but you can find equivalent machines so cheap that your labor fixing one is worth probably $50 divided by the hours you work on it.

Ed Fair
Give the wrong symptoms, get the wrong solutions.

 

[wahnula]

Another thing to look at would be bad capacitors (

http://www.badcaps.net).

On a board that age that would be a real possibility. I see it's a Socket 370 (Pentium III) and wonder if it is worth the effort to repair it or even spend a great deal of time troubleshooting it. Sometimes it's better to cut your losses and look to newer hardware.

With the cost of MB/CPU/RAM these days, you could build a decent system with all three for less than $100.

There's still modern boards that support AGP like this but I would look at a board with onboard graphics and a PCI-e slot for future expandability.

Tony

Users helping Users...

 

[eponymous]

Edfair,

Thanks for this and I do see your point vis-a-vis cost-effectiveness.

Perhaps it works differently with different groups.

To a fee-based consultant with many clients and where time is at a premium, what you propose is both crucial and cost-effective.

To the computer hobbyist, where time is free; the enterprise very small scale (possibly a single computer); costs need to be at minimum and equipment justify sufficient levels of use. Here knowledge is the value that is at a premium - and two very different value systems are at play.

As an individual hobbyist, your comment reminds me, as with investing money, there is a point at which investing time becomes non-productive. Similarly, equipment may be necessary to assess the situation sufficiently early. The criteria by which that choice of equipment is made may vary quite differently from a choice made in a commercial setting.

Re the computer in question, I think I have to accept it is sufficiently gone to warrant nofurther time or money.

Many thanks,
Eponymous

 

[edfair]

I'm at both places. There have been lots of times I would spend hours trying to understand what is going on with a problem board. This was primarily back in ancient history when you could replace components, but I still do it sometimes. It helped that I spent 24 years working on equipment that required diagnostics to find the exact part that was malfunctioning before I started working on the micros.

Once you are to the point of understanding the operation you'll find a need for diagnostic tools. You buy whatever you have to have for one problem and generally find you never need it ever again. Should be a law of "Murphy"

 

[eponymous]

Wahnula,

Thanks for this. I have to accept it has reached the point of no return. The computer in question has a sentimental attachment. I inherited it from a fellow computer hobbyist. As with anything inherited, one tends to want to preserve it.

My decision now is whether or not to pick up another mobo.

Presume the key combatability issue is the cpu. It has a coppermine (p3 SL5DX (866/256/133/1.7v). Should I find myself at a computer fair - and needing to make a quick purchase - are there any general rules for checking compatibility, please?

Many thanks,
Eponymous

 

[edfair]

You are very unlikely to find a M/B that is going to be enough compatible to stop XP from upchucking.

Were you a customer I would suggest that you find a leasing company that has P4 stuff off lease and pick up something like a Dell GX280 for under $200 and transfer data from your old drive to the replacement.

But you can find other that is closer to what you want on ebay for cheaper.

 

[wahnula]

...pick up something like a Dell GX280 for under $200

I am an advocate against throwing new money at old hardware. Take a look at my recommendations in thread602-1483721 for assembling a budget PC that'll STOMP the P4 box.

I am a PC hobbyist and my workplace satisfies my lust for building systems and cobbling together systems from old hardware, but at some point I got fed up with the lack of support, drivers, incompatibility, and general wonky behavior of old hardware. Plus that micro-fine dust of old gear...YUCK!

You can do a repair install of XP with the old HDD connected to the new hardware to get your data off it, or buy a new HDD and connect the old as a slave to shuttle the data over to the new HDD. Either way you're ahead and will have a decent machine when you're done. You can re-activate XP over the phone (you're only using it on ONE machine) so you should be OK there, but don't expect to re-use your old case & PSU.

Tony

Users helping Users...

 

[eponymous]

Edfair and Wahnula,

I appreciate and indded agree with your suggestions. I realise that mine is a somwhat unusual situation. I already run higher spec machines. The computer in question came from a late friend and so has strong sentimental value making it difficult for me for it to die. Hence my considering rebuilding it as an act of preservation. However, with your help, have realised this is not a viable option. Many thanks.

Regards,
Eponymous

 

[edfair]

It comes down to the circumstances that we deal with and the comfort level we have with older equipment. I've taken multiple customers to used equipment but it is with manufacturers that support their hardware by keeping drivers available. For the price of new stuff I can get an operating machine and a backup machine and have a substitution ready as fast as moving a hard drive from one machine to another.

Except for one of my customers and the Sunday before Memorial Day the process has worked well. The backup machine lost the OS drive and the data drive on the second boot as I was preparing for data transfer and I had to do a total rebuild onsite to the extent of 7 hours. The usual transfer from primary to backup is 25 to 30 minutes.

 

[wahnula]

eponymous,

Sorry about your friend. We don't see many "sentimental value" situations here, most PCs are not that personal. But I can understand that one is. And I mistakenly thought yours was a Dell, you can most likely re-use the case should you decide to do a rebuild.

And you need to identify exactly what about the PC reminds you of your friend; is it the case, their data, their name in the Welcome screen?

Should you decide to rebuild, think about the old farmer who swore he had the same ax for 50 years. He'd changed the handle seven times, and the blade four, but by golly it was the same ax!

At what point does it stop becoming your friend's PC and start becoming something else? Only you know the answer to that one.

Tony

Users helping Users...

 

[eponymous]

Tony,

Thank you for your very thoughtful and helpful response, which I shall certainly ponder on. Main memories are of computer fairs and of learning by working on computers together. This one was particularly problematic. I like your idea of keeping the case for a rebuild and retaining the welcome name. Thanks again.

Kind regards,
Denis

 

[Prydonian]

Ok. This thread is an example of why I love being part of this forum. We have people here who take the time to try to figure out a way to do the immpractical, the improbable and in some cases, the immpossible, knowing full well that it doesn't always come down to dollars and cents, that sometimes it is the challenge, or in this instance, an emotional investment.

Kudos to all involved! The world would be a better place with attitudes like those expressed here.

Denis: I fully understand your position. As long as we have something to remember people by, they will always be with us.

Tony: I loved your farmer and the ax story! I will write that one down and it will definitely be used again!

Mike, The IT Guy.

If you can read this, thank a teacher. If you can read this in English, thank a veteran!