Personal computing discussed

Moderators: renee, mac_h8r1, Nemesis

 
CampinCarl
Graphmaster Gerbil
Topic Author
Posts: 1363
Joined: Mon Jul 04, 2005 9:53 pm

Dying older machine

Fri Apr 06, 2012 9:19 pm

Well, I'm pretty sure tonight was the last gasp, but I thought I might come here to see if anyone has any ideas. First, the specs. It's the computer I built many, many moons ago from recommendations here on TR.

    Original CPU: Athlon 64 3000+
    Upgraded CPU: Athlon 64 x2 3800+
    GPU: XFX 8600GTS
    RAM: 2x512, 1x1GB Corsair PC-3200
    Motherboard: Gigabyte K8NF-9
    HDD: 80 GB Western Digital SATA150

It's been running Windows 7 x86 and Ubuntu 10.04 x86 lately. So here's what it's been doing: Will boot and run just fine and dandy, totally stable through days of Memtest, Prime95 torture testing, playing video games, etc. Then, if I turn off the system or hibernate it, many times it will refuse to boot. A lot of times, I have been able to revive it by removing the power plugs and data cables from all devices and removing the graphics card. The other day, however, I mistakenly put the computer into hibernate mode instead of sleep mode from Windows 7. Any attempts to revive it have failed. I have used multiple variations of the PSU and video card from my current computer, as well as the old A64 3000+ CPU. It refused to boot. Game over, man, it's game over! (For the motherboard, I think). Anyone have any ideas?
Gigabyte AB350M Gaming-3 | R7 1700X | 2x8 GB Corsair Vengeance DDR4-3200 (@DDR4-2933)| Samsung 960 Evo 1TB SSD | Gigabyte GTX1080 | Win 10 Pro x86-64
 
mnecaise
Gerbil First Class
Posts: 139
Joined: Thu Oct 30, 2008 1:17 pm
Location: Hampton Roads, Virginia

Re: Dying older machine

Fri Apr 06, 2012 10:54 pm

As I understand it, you have tried swapping power supplies and video cards from another computer... That's a good first step in isolating the problem.

replace the battery. reseat everything, including memory and processor. leave the drives disconnected (for now). make sure the power supply is starting and outputing the correct voltages when you try to start the computer (assumes you have a voltmeter). If you have a known good power supply and it won't power on* or the PSU starts and you have good voltages but you don't get a BIOS boot screen, it's new MB time.

*the power supply is switched by a soft power on/off circuit on the motherboard. The power supply provides a 5V standby voltage to power this circuit. You can have a perfectly good power supply that "won't start" because the motherboard soft power control circuit is dead (as in mostly), deader or really really dead (as in smoking crater where the IC used to be). If the power supply works in another computer (as you imply you have tried) then I would tend to call it "good" and place the blame somewhere else, although I do check the voltages before declaring a power supply "good".

For what it's worth, if you find the motherboard has kicked the bucket, toss the power supply too. It's my experience that motherboard failures are either:
1. thermal
2. bad power
 
ShadowEyez
Gerbil XP
Posts: 348
Joined: Wed Dec 03, 2003 12:31 pm

Re: Dying older machine

Fri Apr 06, 2012 11:39 pm

From reading your post it looks like the problem is your mother board. If the memory runs memtest, prime95 doesn't crash your system and you can use your gxf card for games, and the system sometimes refuses to boot, it's likely a motherboard problem. You could also try swapping out your power supply first and see if that works, but I've seen problems like this in the past that end up being the motherboard.

If it is the board, you can likley buy a new one for not a lot but then you'd have to swap all the part out, and in the end you'd still have an older, though still functional, system.
The finest tools are forged from the hottest fires
 
ShadowEyez
Gerbil XP
Posts: 348
Joined: Wed Dec 03, 2003 12:31 pm

Re: Dying older machine

Fri Apr 06, 2012 11:47 pm

From reading your post it looks like the problem is your mother board. If the memory runs memtest, prime95 doesn't crash your system and you can use your gxf card for games, and the system sometimes refuses to boot, it's likely a motherboard problem. You could also try swapping out your power supply first and see if that works, but I've seen problems like this in the past that end up being the motherboard.

If it is the board, you can likley buy a new one for not a lot but then you'd have to swap all the part out, and in the end you'd still have an older, though still functional, system.
The finest tools are forged from the hottest fires
 
JustAnEngineer
Gerbil God
Posts: 19673
Joined: Sat Jan 26, 2002 7:00 pm
Location: The Heart of Dixie

Re: Dying older machine

Sat Apr 07, 2012 2:52 am

I once had a case power switch go bad. That was a simpler fix than a new motherboard. You could try swapping the reset and power switches for a quick test.
· R7-5800X, Liquid Freezer II 280, RoG Strix X570-E, 64GiB PC4-28800, Suprim Liquid RTX4090, 2TB SX8200Pro +4TB S860 +NAS, Define 7 Compact, Super Flower SF-1000F14TP, S3220DGF +32UD99, FC900R OE, DeathAdder2
 
CampinCarl
Graphmaster Gerbil
Topic Author
Posts: 1363
Joined: Mon Jul 04, 2005 9:53 pm

Re: Dying older machine

Sat Apr 07, 2012 8:35 am

Okay, so, I took JAE's advice this morning. However, the result is roughly the same. Fans spin up, but I don't get an 'on' computer. No beeps for having no video card or RAM, etc.. Out of curiosity, I swapped the switches back and received the same result. So I don't think it's necessarily a broken switch. :(

Then I decided to plug everything back in, plug in the Keyboard and mouse, monitor, etc. and turn it on. Same result. Fans spin up, but nothing comes out. No beep for successful boot that would suggest a problem with the video card displaying an image to the monitor.
Gigabyte AB350M Gaming-3 | R7 1700X | 2x8 GB Corsair Vengeance DDR4-3200 (@DDR4-2933)| Samsung 960 Evo 1TB SSD | Gigabyte GTX1080 | Win 10 Pro x86-64
 
Synchromesh
Graphmaster Gerbil
Posts: 1001
Joined: Tue Apr 12, 2005 10:56 am
Location: The Land of Beetles.
Contact:

Re: Dying older machine

Sat Apr 07, 2012 10:36 am

CampinCarl wrote:
Okay, so, I took JAE's advice this morning. However, the result is roughly the same. Fans spin up, but I don't get an 'on' computer. No beeps for having no video card or RAM, etc.. Out of curiosity, I swapped the switches back and received the same result. So I don't think it's necessarily a broken switch. :(

Then I decided to plug everything back in, plug in the Keyboard and mouse, monitor, etc. and turn it on. Same result. Fans spin up, but nothing comes out. No beep for successful boot that would suggest a problem with the video card displaying an image to the monitor.


If the fans spin up, clearly your motherboard gets the signal from the power button. If you know what you're doing you can even manually short the power pins with a screwdriver or a pair of pliers. That's the same thing as power button press. I would reset CMOS just to be sure but most likely your mobo has gone to a better world.

One thing you can check for physical damage on the board are the capacitors. All of them have to have perfectly flat tops. If any tops are swollen or puffy - there is your problem. I've seen several Dell systems from that era with the same issue.
 
Aranarth
Graphmaster Gerbil
Posts: 1435
Joined: Tue Jan 17, 2006 6:56 am
Location: Big Rapids, Mich. (Est Time Zone)
Contact:

Re: Dying older machine

Sat Apr 07, 2012 10:53 am

I second Synchromesh.

Swap the Cmos battery (you can get em at walgreens).
Clear the cmos etc.

I had a similar issue and it turned out to be a corrupted cmos profile.
Leaving the machines clear cmos jumper in place for a day allowed the machine to boot with the "safe settings" profile and allowed me to get the machine back up and running.

On the other hand its too bad it suddenly fixed itself like that. I almost got the wife's ok to upgrade to a core i5 :D
Main machine: Core I7 -2600K @ 4.0Ghz / 16 gig ram / Radeon RX 580 8gb / 500gb toshiba ssd / 5tb hd
Old machine: Core 2 quad Q6600 @ 3ghz / 8 gig ram / Radeon 7870 / 240 gb PNY ssd / 1tb HD
 
CampinCarl
Graphmaster Gerbil
Topic Author
Posts: 1363
Joined: Mon Jul 04, 2005 9:53 pm

Re: Dying older machine

Sat Apr 07, 2012 10:03 pm

The caps are fine; that was one of the first things I checked for. None leaking, no unreasonable bulging going on. I'll try the trick with a new CMOS battery tomorrow, didn't have time to stop at the store today. Trying to tire out a 3 year old chocolate lab is tiring!
Gigabyte AB350M Gaming-3 | R7 1700X | 2x8 GB Corsair Vengeance DDR4-3200 (@DDR4-2933)| Samsung 960 Evo 1TB SSD | Gigabyte GTX1080 | Win 10 Pro x86-64
 
CampinCarl
Graphmaster Gerbil
Topic Author
Posts: 1363
Joined: Mon Jul 04, 2005 9:53 pm

Re: Dying older machine

Sun Apr 15, 2012 6:31 pm

So I finally got around to getting an extra CR2032 and resetting the CMOS with the new battery. Still no go. I think it's dead :(
Gigabyte AB350M Gaming-3 | R7 1700X | 2x8 GB Corsair Vengeance DDR4-3200 (@DDR4-2933)| Samsung 960 Evo 1TB SSD | Gigabyte GTX1080 | Win 10 Pro x86-64
 
Dposcorp
Minister of Gerbil Affairs
Posts: 2771
Joined: Thu Dec 27, 2001 7:00 pm
Location: Detroit, Michigan

Re: Dying older machine

Sun Apr 15, 2012 7:10 pm

I would guess the board most likely, however, in a case like this, my last ditch effort is to simulate a fresh build.

1)Board out of the case sitting on a cardboard motherboard box or similar,
2)Remove and re-install CPU and HSF
3)Remove and re-install ram. (start with 1 stick in the first recommended slot per the manual) Have spare ram.
4)Onboard video only or low power video card.
5)Clear CMOS and u can even leave the battery out.
6)Install power supply, and have a spare handy to swap.

See if u can boot to BIOS using a screwdriver on the pwr pins; if not, try it with different ram / different ram slot, and also different PS.

If none of that works, most likely is is the board, but only way to test is to have a spare board.

However, once in a while the above worked for me, and I was never able to track down the exact problem.
(Dirty ram slot, maybe the CPU somehow shift in the socket, power supply connectors needed to be removed and re-inserted, something in the case shorted out, etc...

Works comes to worse, time to sell AS-IS and move on.


Good luck and let us know.
 
CampinCarl
Graphmaster Gerbil
Topic Author
Posts: 1363
Joined: Mon Jul 04, 2005 9:53 pm

Re: Dying older machine

Sun Apr 15, 2012 8:47 pm

Dposcorp: Actually, when I swapped in the A64 3000+ for the X2 3800+, I did a full rebuild at that time. Didn't improve anything :( I might try the screwdriver to PWR pins trick though. I kind of doubt it'll help since I already swapped the buttons to check for broken switches, but it's worth a shot.

Thanks everyone :)
Gigabyte AB350M Gaming-3 | R7 1700X | 2x8 GB Corsair Vengeance DDR4-3200 (@DDR4-2933)| Samsung 960 Evo 1TB SSD | Gigabyte GTX1080 | Win 10 Pro x86-64
 
Waco
Maximum Gerbil
Posts: 4850
Joined: Tue Jan 20, 2009 4:14 pm
Location: Los Alamos, NM

Re: Dying older machine

Sun Apr 15, 2012 9:56 pm

I'm agreed that it's likely a dead mobo...but to be sure, you could try doing a completely cold boot. Remove the PSU power cord, remove the CMOS battery, wait 5 minutes, then try to boot it up.

The fact that it did this when you set it to hibernate sounds like coincidence.
Victory requires no explanation. Defeat allows none.
 
dragmor
Grand Gerbil Poohbah
Posts: 3644
Joined: Mon Sep 23, 2002 7:24 pm
Location: Oz

Re: Dying older machine

Sun Apr 15, 2012 10:38 pm

I've had this happen with machines when the P4 power connector was loose / plugged in backwards / corroded, etc.

You don't get any error beeps, as the board shuts it self down as a saftey before it initialises.
SZ87R6/i5 4560 stock/24GB 2333mhz/840 Evo 250GB/Seagate 2TB/ASUS 760GTX/Dell 2711
Rainbows lie in corded knots
While thunder wakes the sleeping crocs.
 
Waco
Maximum Gerbil
Posts: 4850
Joined: Tue Jan 20, 2009 4:14 pm
Location: Los Alamos, NM

Re: Dying older machine

Sun Apr 15, 2012 10:41 pm

I love that people call it a "P4 power connector" just because Pentium 4s were the first CPUs to suck so much power that they required a second connection. :lol:
Victory requires no explanation. Defeat allows none.
 
ShadowEyez
Gerbil XP
Posts: 348
Joined: Wed Dec 03, 2003 12:31 pm

Re: Dying older machine

Sun Apr 15, 2012 10:48 pm

You swaped CPU's, full rebuild, even changed the CMOS battery - again, very likely the mobo. Do you have another AM2 mobo to try?
The finest tools are forged from the hottest fires
 
Waco
Maximum Gerbil
Posts: 4850
Joined: Tue Jan 20, 2009 4:14 pm
Location: Los Alamos, NM

Re: Dying older machine

Sun Apr 15, 2012 10:57 pm

It's a 939 mobo. I just recently sold all of mine off otherwise I'd send him one for the cost of shipping. :(
Victory requires no explanation. Defeat allows none.
 
JustAnEngineer
Gerbil God
Posts: 19673
Joined: Sat Jan 26, 2002 7:00 pm
Location: The Heart of Dixie

Re: Dying older machine

Mon Apr 16, 2012 4:18 am

The good news is that you're just in time for an Ivy Bridge upgrade.
· R7-5800X, Liquid Freezer II 280, RoG Strix X570-E, 64GiB PC4-28800, Suprim Liquid RTX4090, 2TB SX8200Pro +4TB S860 +NAS, Define 7 Compact, Super Flower SF-1000F14TP, S3220DGF +32UD99, FC900R OE, DeathAdder2
 
ryko
Gerbil Team Leader
Posts: 256
Joined: Tue Feb 27, 2007 3:58 pm
Location: new york

Re: Dying older machine

Mon Apr 16, 2012 8:46 am

I agree with everyone else in that it is most likely the mobo. However I did want to chime in and say that I was never able to successfully get a s939 mobo to behave properly in sleep mode or hibernation. I thought it was a design flaw as I remember it being pretty prevalent across many forums. Most people just disabled any type of sleep and called it a day. I also remember that maybe coolnquiet had something to do with it. If enabled it would mess up your sleep mode. so no power saving enabled in bios and no sleep or hibernation....just oc the crap out of it and live with increased energy costs!

My s939 would go into sleep and then never come out....fans spinning but no display activity. Very similar to what you report. The only thing that worked for me was to pull out the psu power cord and wait 5 minutes with the clear CMOS jumper in the clear position. Then switch the jumper back and power up the pc. Any less or no CMOS clear would net me a no boot with fans spinning but no display.

Good luck! And maybe last ditch effort would be to try a different brand gpu. So if amd(ati) try nvidia or vice versa.
 
srg86
Gerbil Team Leader
Posts: 262
Joined: Tue Apr 25, 2006 7:57 am
Location: Madison, WI

Re: Dying older machine

Mon Apr 16, 2012 9:34 am

ryko wrote:
I agree with everyone else in that it is most likely the mobo. However I did want to chime in and say that I was never able to successfully get a s939 mobo to behave properly in sleep mode or hibernation. I thought it was a design flaw as I remember it being pretty prevalent across many forums. Most people just disabled any type of sleep and called it a day. I also remember that maybe coolnquiet had something to do with it. If enabled it would mess up your sleep mode. so no power saving enabled in bios and no sleep or hibernation....just oc the crap out of it and live with increased energy costs!

My s939 would go into sleep and then never come out....fans spinning but no display activity. Very similar to what you report. The only thing that worked for me was to pull out the psu power cord and wait 5 minutes with the clear CMOS jumper in the clear position. Then switch the jumper back and power up the pc. Any less or no CMOS clear would net me a no boot with fans spinning but no display.

Good luck! And maybe last ditch effort would be to try a different brand gpu. So if amd(ati) try nvidia or vice versa.


Weird! My s939 machine, which for now is still my main rig has never had any troubles with hibernate or sleep mode. It has cool'n'quiet enabled and way back when I used to run WinXP, it worked fine (no hibernate or sleep issues) and same with Kubuntu now, it's just fine. I've used hibernate on this machine a lot over the years as well! I will say that since maxing out this motherboard to 4GB RAM I have had the odd lock up here and there that I've never had before, but that's a different issue (RAM tests as fine).

I guess where others didn't, DFI got sleep and hibernate to work properly.
Intel Core i7 4790K, Z97, 16GB RAM, 128GB m4 SSD, 480GB M500 SSD, 500GB WD Vel, Intel HD4600, Corsair HX650, Fedora x64.
Thinkpad T460p, Intel Core i5 6440HQ, 8GB RAM, 512GB SSD, Intel HD 530 IGP, Fedora x64, Win 10 x64.
 
ryko
Gerbil Team Leader
Posts: 256
Joined: Tue Feb 27, 2007 3:58 pm
Location: new york

Re: Dying older machine

Mon Apr 16, 2012 11:44 am

Hmm maybe it is only when oc'ed (and with coolnquiet) that it screws up sleep/hibernation??? At stock it might be ok, but I have 2 s939 dfi nf4u mobos here that refuse to go into sleep. Ymmv I guess!

I do know that populating all 4 ram slots has a detrimental effect on stability though for sure! I could never get my ddr500 to it's rated speed with all 4 slots. Even ddr400 settings weren't completely stable...
 
TDIdriver
Gerbil First Class
Posts: 119
Joined: Mon Jul 11, 2011 3:49 pm
Location: Alabama

Re: Dying older machine

Mon Apr 16, 2012 11:45 am

have you tried the "DFI long bios clear"?
Q9450|GA-EP45-UD3P|8GB G.skill PC2-8800|HD7870 Myst
120GB Vertex3|300GB Velociraptor|HX650|600T
 
Walkintarget
Gerbil Team Leader
Posts: 291
Joined: Tue Jan 29, 2008 10:15 am

Re: Dying older machine

Mon Apr 16, 2012 12:24 pm

My first guess was caps as well, as I have a similar Gigabyte NF4 SLI board with swollen caps, but its still running, so I use it for a testbed PC for questionable used video cards or other hardware I buy. Talk about an odd mix - it has an AMD 3500+ Venice in it and I popped a GTX260 into the box to test the card. Talk about CPU bound !

One other thing to try would be a different mouse or keyboard, and if using PS2, try USB or vice versa. I spent a good chunk of the day yesterday testing out a shelf full of my questionable hardware, so I feel your pain on this one. I can test any hardware that you have as well if you do need a similar config to test anything that is determined to be a possible culprit.
AMD Ryzen R5 2600x, Asus Prime x470 Pro, Sapphire Nitro Fury 4GB, HP EX920 512GB NVMe SSD, 16GB Corsair Vengeance RGB Pro, 4TB WD Red HDD, Thermaltake Level 10 GT Snow Ed, Win 10 Pro
 
thedude
Gerbil In Training
Posts: 2
Joined: Tue May 29, 2012 12:19 am

Re: Dying older machine

Tue May 29, 2012 2:04 am

Wow. You said you had an older machine, but it looks like you're a good generation ahead of me.
I'm typing this on an AthlonXP/nforce box I built about a decade ago now, I guess. Still works great.
Got 2 more just like it. I guess it will come in handy when you need to salvage parts, I guess.

Anyway, these are some good guesses as to what's going on (yeah, if I had to guess, I say it was probably the motherboard).
But, why guess when decent multimeters are so cheap? Especially if you have a Harbor Freight where you live. That way, you can test the power supply rails and see if they're within spec.
http://www.harborfreight.com/catalogsea ... multimeter
Or, if you build/repair enough PCs, it might be worth it to get a power supply tester. You can pick one of those up for less than $20 as well. Beats the hell out of buying replacement PSUs or mobos unnecessarily. :wink:
http://www.google.com/search?q=power+supply+tester.

Who is online

Users browsing this forum: No registered users and 1 guest
GZIP: On