Help with diagnosing dead PC

M

MM556

Member
This machine has been working perfectly on a daily basis for over a decade. It hasn't moved or had any hardware, software additions to it recently. Friday it booted up fine. Saturday it started to boot, then instantly stopped. A few seconds later it started to boot again, and again instantly stopped. So basically stuck in a loop.
I have stripped everything off, removed motherboard from case. Now just motherboard, single stick of RAM, and CPU and cooler. This behavior still persists.
With just the 24 pin ATX connector attached all seems OK, but this fault loop starts as soon as I attached the 8 pin CPU connector.

I have tested the PSU with one of them cheap PSU testers.
I have removed, reseated CPU, cleaned and reinstalled CPU cooler.
I can't see any problems with the motherboard: no bulging caps, no signs of burning or anything like that.

What else can I do to diagnose the problem?
Could it still be a PSU issue (I don't have a spare to test with)?

Any thoughts appreciated. This machine has several SAS HBAs installed, connected to four 8-Disk SAS RAID5 Arrays which contain my archive.
That's all backed up of course, but the last thing on earth I want to do is restore 96TB of data...

Thanks!

Mark
Case
COOLERMASTER HAF-X FULL TOWER GAMING CASE
Promotional Item
FREE HOS: KAIJO DIABLO BUNDLE with GTX 950 & 960 GPUs!
Processor (CPU)
Intel® Core™ i7 Six Core Processor i7-5930K (3.5GHz) 15MB Cache
Motherboard
ASUS® X99-S: ATX, HSW-E CPU, USB 3.0, SATA 6 GB/s
Memory (RAM)
64GB SAMSUNG DDR4 2133MHz (8 x 8GB)
Graphics Card
4GB NVIDIA GEFORCE GTX 980 - DVI, HDMI, 3 x DP - GeForce GTX VR Ready!
1st M.2 SSD Drive
Plextor PX-G128M6e 128GB M.2 SSD (up to 770MB/sR | 625MB/sW)
1st Storage Drive
120GB KINGSTON V300 2.5" SSD, SATA 6 Gb (450MB/R, 450MB/W)
DVD/BLU-RAY Drive
16x BLU-RAY WRITER DRIVE, 16x DVD ±R/±RW & SOFTWARE
Power Supply
CORSAIR 850W RM SERIES™ MODULAR 80 PLUS® GOLD, ULTRA QUIET
Power Cable
1 x 1.5 Metre UK Power Cable (Kettle Lead)
Processor Cooling
Super Quiet 22dBA Triple Copper Heatpipe Intel CPU Cooler
Thermal Paste
ARCTIC MX-4 EXTREME THERMAL CONDUCTIVITY COMPOUND APPLICATION
Sound Card
ONBOARD 8 CHANNEL (7.1) HIGH DEF AUDIO (AS STANDARD)
USB/Thunderbolt Options
MIN. 2 x USB 3.0 & 6 x USB 2.0 PORTS @ BACK PANEL + MIN. 2 FRONT PORTS
Operating System
NO OPERATING SYSTEM REQUIRED
Windows Recovery Media
NO RECOVERY MEDIA REQUIRED
Office Software
NO OFFICE SOFTWARE
Anti-Virus
NO ANTI-VIRUS SOFTWARE
Browser
Firefox™
Keyboard & Mouse
LOGITECH® K120 USB KEYBOARD
Keyboard & Mouse
LOGITECH® OPTICAL USB MOUSE
Warranty
3 Year Standard Warranty (6 Month Collect & Return, 1 Year Parts, 3 Year Labour)
Delivery
STANDARD INSURED DELIVERY TO UK MAINLAND (MON-FRI)
Build Time
Standard Build - Approximately 7 to 9 working days
Price: £0.00 including VAT and Delivery
 

Attachments

  • IMG_5941.JPEG
    IMG_5941.JPEG
    462.1 KB · Views: 14
T

TonyCarter

VALUED CONTRIBUTOR
How far into the boot process does it get? Does it even get as far as completing POST and then showing the F2/Del to get to the BIOS screen?

Is the boot SSD working fine?
 
M

MM556

Member
Thanks Tony,
Sorry, I should have been clearer.

As soon as one powers on the machine, it stops and resets. The CPU fan starts to spin up, and then everything powers down.
The only sound is the relay of the PSU clicking on and off.
It doesn't even get as far as starting the POST.
 
T

TonyCarter

VALUED CONTRIBUTOR
Sounds more hardware than software then. Which is going to be difficult to diagnose over the internet.

The RAID should be fine if it's just a failed CPU/motherboard. Are there any 'motherboard' error beeps/LEDs/codes on that motherboard which would tell you where in the process it is failing?

Found a error code list for that chipset...any of those shown on the motherboard?

Manual is here...

...and pages 1-21 to 1-23 show where the LED / Q-Codes can be found...
X99-S Error Code Locations.jpg
 
M

MM556

Member
@TonyCarter thank you so, so much for your time.
Really appreciated.

Sadly it doesn't even boot up long enough to give me an LED error code. No helpful beeps.
It's like it's shorting out immediately and restarting, or something similar.
As you say, maybe just impossible to sort out over the internet :rolleyes:

Thanks again

Mark
 
T

TonyCarter

VALUED CONTRIBUTOR
If it's not even getting that far, then it's motherboard or PSU - although you could try resetting/replacing the CMOS battery as a last test before going kamikaze on the system.
 
M

MM556

Member
Thanks, Tony.

I did replace the CMOS battery early in the process, and did a reset with the motherboard's hardware reset switch.
So how do I figure out if it's the motherboard or the PSU when I don't have a spare PSU... I suppose I can order one from Amazon and 'return' it after testing :)
 
SpyderTracks

SpyderTracks

We love you Ukraine
MM556 said:
Thanks, Tony.

I did replace the CMOS battery early in the process, and did a reset with the motherboard's hardware reset switch.
So how do I figure out if it's the motherboard or the PSU when I don't have a spare PSU... I suppose I can order one from Amazon and 'return' it after testing :)
Click to expand...
Use a PSU tester is the simple answer

This one is very highly rated by thermaltake

www.scan.co.uk

Thermaltake Dr. Power III Power Supply & Diagnostics Tester upto ATX3.1

Buy from Scan - Thermaltake Dr. Power III Power Supply Tester, LCD Display, PCIe 12+4pin Detection, Flexible Testing Modes
www.scan.co.uk www.scan.co.uk
 
M

MM556

Member
SpyderTracks said:
Use a PSU tester is the simple answer

This one is very highly rated by thermaltake

www.scan.co.uk

Thermaltake Dr. Power III Power Supply & Diagnostics Tester upto ATX3.1

Buy from Scan - Thermaltake Dr. Power III Power Supply Tester, LCD Display, PCIe 12+4pin Detection, Flexible Testing Modes
www.scan.co.uk www.scan.co.uk
Click to expand...
Thanks, yes, I used one of these... but there seems to be an opinion on the internet that you can't really test a PSU unless it's under load, and none of these testers really provide a load, as I understand it.
 

Attachments

  • IMG_5941.JPEG
    IMG_5941.JPEG
    462.1 KB · Views: 6
SpyderTracks

SpyderTracks

We love you Ukraine
MM556 said:
Thanks, yes, I used one of these... but there seems to be an opinion on the internet that you can't really test a PSU unless it's under load, and none of these testers really provide a load, as I understand it.
Click to expand...
That may be true if it was an issue specifically presenting under load, that doesn't apply for you, this is purely powering the system

You can't use a cheap PSU tester, they're fake basically, this one is industry recognised and will tell you if any of the PSU outputs aren't reaching their specific Rail minimums.
 
M

MM556

Member
OK thank you. That's good to know. I'll get on to Scan for that one. I do have an original Dr Power from about 1995 :) but it doesn't test the 8 pin 12V connector for the CPU.
 
M

MM556

Member
SpyderTracks

SpyderTracks

We love you Ukraine
MM556 said:
Thank you. I tested the PSU with my multimeter and all the voltage readings are correct for all the pins on the 24 pin ATX connector, and on the 8pin connector to the CPU.
Can I now assume that the PSU is fully functioning, and that it's an issue with the motherboard/cpu?
Cheers
Mark
Click to expand...
If it was the CPU, you'd still get all lights on the motherboard, it just wouldn't boot.

If there's no lights on the motherboard at all, it's likely the motherboard.

But it's not worth fixing, it's well past it's prime anyway.
 
Scott

Scott

Behold The Ford Mondeo
Moderator
In your shoes I would remove all connections that aren't required for a boot, remove the CPU and remove the RAM. Switch it on and see what happens. You should get a RAM/CPU warning. It won't post or anything like that but it should give error lights.

If it reboots and goes through the process again then it's definitely the motherboard. You've ruled out the CMOS etc so it would need to be the board itself.

If it doesn't go through that process. Switch it off, stick in the CPU (carefully) with the cooler on and attached and give it another spin. You should get an LED fault on RAM. If that happens, put in one stick of RAM and see how it goes.

If you get post then it's likely one of the RAM sticks. These faults normally present after post but there are times when serious hardware faults will just kill the boot altogether. IME it's typically CPU failures outside of the motherboard that will present like this.

As suggested though, I wouldn't spend money to get parity. Upgrade the board, CPU & RAM for a modern system.
 
M

MM556

Member
Scott said:
In your shoes I would remove all connections that aren't required for a boot, remove the CPU and remove the RAM. Switch it on and see what happens. You should get a RAM/CPU warning. It won't post or anything like that but it should give error lights.

If it reboots and goes through the process again then it's definitely the motherboard. You've ruled out the CMOS etc so it would need to be the board itself.

If it doesn't go through that process. Switch it off, stick in the CPU (carefully) with the cooler on and attached and give it another spin. You should get an LED fault on RAM. If that happens, put in one stick of RAM and see how it goes.

If you get post then it's likely one of the RAM sticks. These faults normally present after post but there are times when serious hardware faults will just kill the boot altogether. IME it's typically CPU failures outside of the motherboard that will present like this.

As suggested though, I wouldn't spend money to get parity. Upgrade the board, CPU & RAM for a modern system.
Click to expand...
Thanks Scott. I tried your suggestion. Switching on without CPU and RAM results in same things: it powers up, powers down immediately, and repeats. So yes, definitely the motherboard wouldn't you say?

This has been my fileserver, and it's hard to find a current motherboard that I can afford that has the full length PCIe slots I need for my SAS HBAs, Thunderbolt, and 10Gbe ethernet. I'll drop sales a line and see what they suggest.

Thanks again

Mark
 
Last edited:
SpyderTracks

SpyderTracks

We love you Ukraine
MM556 said:
Thanks Scott. I tried your suggestion. Switching on without CPU and RAM results in same things: it powers up, powers down immediately, and repeats. So yes, definitely the motherboard wouldn't you say?

This has been my fileserver, and it's hard to find a current motherboard that I can afford that has the full length PCIe slots I need for my SAS HBAs, Thunderbolt, and 10Gbe ethernet. I'll drop sales a line and see what they suggest.

Thanks again

Mark
Click to expand...
A lot of current motherboards have 10Gb Ethernet on the board, although with PCS offerings it would need to be a workstation, but at the consumer end, the Asus Pro-Art X870e is one example with 10Gb ethernet, shame PCS don't offer that as they did offer the X670e version until it sold out, hopefully it will come back.

All modern motherboards include thunderbolt as standard (USB 4 incorporates Thunderbolt 3, but a lot of modern boards will have dedicated thunderbolt ports on top).

So that's definitely 1 slot you can free up for the thunderbolt card (if that's what you were requiring a PCIe slot for), and then a board like the Crosshair Hero X870e would allow for 2 full sized slots at X16 PCIe 5
 
Last edited:
You must log in or register to reply here.
Top