• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.

Last Try Before I Trash It (ASUS X58 Sabertooth)

Joined
Feb 21, 2008
Messages
6,873 (1.08/day)
Location
S.E. Virginia
System Name Barb's Domain
Processor i9 10850k 5.1GHz all cores
Motherboard MSI MPG Z490 GAMING EDGE WIFI
Cooling Deep Cool Assassin III
Memory 2*16gig Corsair LPX DDR4 3200
Video Card(s) RTX 4080 FE
Storage 500gb Samsung 980 Pro M2 SSD, 500GB WD Blue SATA SSD, 2TB Seagate Hybrid SSHD
Display(s) Dell - S3222DGM 32" 2k Curved/ASUS VP28UQG 28" 4K (ran at 2k), Sanyo 75" 4k TV
Case SilverStone Fortress FT04
Audio Device(s) Bose Companion II speakers, Corsair - HS70 PRO headphones
Power Supply Corsair RM850x (2021)
Mouse Logitech G502
Keyboard Logitech Orion Spectrum G910
VR HMD Oculus Quest 2
Software Windows 11 Pro 64 bit
Benchmark Scores https://www.3dmark.com/spy/34962882
I was given a X58 Sabertooth motherboard. I was told it didn't work, but upon very careful inspection, I can find no kind of damage that would suggest anything is fried. So last night, I decided to finally try to get it working. I assembled the basics for a system on my (wood) dinner table and turned on the power to the PSU. Green LED lit up on MB. I tried to boot the system up (using a screwdriver to contact the power switch points). First try, I got a little flash from two red LEDS on the MB, heat sink fan kicked a little, then nothing. Green LED stayed lit. Second try yielded zero results. Turned the power supply off and back on. Green LED turned off and on with PSU. Tried to boot system, got the red LED flash and fan kick again, and that's all. Tried to boot the system again, nothing.

Components-
CPU- i7 960
MB- ASUS X58 Sabertooth
Cooling- Cooler Master Hyper 212
RAM- 8 Gb (2*4 GB) Corsair Vengeance LPX DDR3 1600
GPU- EVGA GTX 780Ti SC
PSU- Cooler Master Silent Pro M 1000
HD- Seagate 7200.12 500 GB

So what are the chances that the motherboard is actually done for with zero visible damage? Personally, I've never seen it before. But then again, I'm not the type of person that has worked on/with thousands of MBs in my lifetime. As far as I know, everything else does work, but I will be testing them today in assembling the AMD PII 940 system for testing I have laying around.
 
Mine did suddenly die the same way. Exactly X58 tooth. It were not any power traces. I suspect a internal trace.
 
How picky is the X58 Sabertooth when it comes to RAM? The actual RAM sticks I'm trying to use are not on the QVL list for the MB, but the MB does support Corsair DDR3 1600 memory.

My RAM-
ram.jpg


Saberttoth QVL-
ramqvl.jpg
 
The fact that it switches on and off immediately suggests a short somewhere.

Unplug the 8-pin EPS and see if it stays on for a bit longer. If so, your VRM is fried.
 
How picky is the X58 Sabertooth when it comes to RAM? The actual RAM sticks I'm trying to use are not on the QVL list for the MB, but the MB does support Corsair DDR3 1600 memory.

It was rather unpicky... it even booted with ECC ram paired with Xeon. Worked... it looked it almost got a heart attack and that's it.

It looks more than went flat than short... voltages were fine, I force powered it later on to check it... call it autopsy results. Clockgens all were ticking, just the LPC didn't do anything. I've noticed it actually on ebay that they have suffered from it a lot. It may be a design flaw.
 
Unplug the 8-pin EPS and see if it stays on for a bit longer. If so, your VRM is fried.

I'll give this a shot real quick. I am thinking the board is dead and not save-able.


EDIT: Holy Hell, we have signs of life. I partially re-assembled the basics. This time I only had 1 stick of RAM, no GPU, no HD, and only the 24 pin power wire connected. Board booted up. I connected the 8 pin EPS wire, and I got it to boot up. I'm going to fully assemble it and see if I actually get a picture on the monitor and go from there.

EDIT #2: Ok, so now the system boots up, but I'm not getting a signal to the monitor. I tried two different GPUs, a GTX 780Ti and a GTX 285. i guess next is to test the monitor on my 3930k system to make sure it's still in working condition. It is old and has been stored in the corner for quite some time now.

EDIT #3: Monitor works perfectly fine on my 3930k system
 
Last edited:
I'm sure you checked, but any bent pins on the socket that might prevent a connection?
 
Yeah, check the pins and for obvious PCB damage. If there are neither of those present, your IOH (North Bridge) might be past its best by time.
 
If you don't ever get a usable system out of it, don't trash it. Sell it as is for parts with a description of that clearly states in the listing title "Parts or Repair" and in the description in bold "As Is, Sold For Parts or Repair, Does Not Work". You will likely get $20-50 out of it on eBay.
 
The red leds are diagnostics; which ones flash will tell you something.

The one for the PCI-e slot is under it, you can see it by reflection, tho.

But one of the other slots should work if the main one is dead.

There's a cpu led, which means the cpu isn't clocking; and a mem led, that means the memory isn't working, next to the MEM_OK button, near the front panel.

The MeM_OK button should find a memory speed that will run, it may take multiple reboots.

I'd clear the cmos, and start from scratch.

You will need the latest BIOS to run a XEON, and it should support 24GB of memory.

Those boards have solid caps, and mil spec parts; they shouldn't age out like most stuff.

It looks like 1402 is the latest bios, but there's a button on the back to load a bios from a usb stick, and it's supposed to work without a cpu.
 
I'm sure you checked, but any bent pins on the socket that might prevent a connection?
Yeah, check the pins and for obvious PCB damage. If there are neither of those present, your IOH (North Bridge) might be past its best by time.

I have not seen any damage, but below is some pictures of the socket and board it's self. Please look them over and see if I may have missed anything.



124693
124694
124695
124696
124697



IS the RAM in the correct lots when you failed ?

A1, D1 and C1 ... 2nd , 4th and 6th from CPU


When I was able to get the board to show life, I only had one stick installed in the closest to the CPU slot.


Previous owner didn't say anything about a BIOs flash. I don't think he has the confidence to try one either.

The red leds are diagnostics; which ones flash will tell you something.

The one for the PCI-e slot is under it, you can see it by reflection, tho.

But one of the other slots should work if the main one is dead.

There's a cpu led, which means the cpu isn't clocking; and a mem led, that means the memory isn't working, next to the MEM_OK button, near the front panel.

The MeM_OK button should find a memory speed that will run, it may take multiple reboots.

I'd clear the cmos, and start from scratch.

You will need the latest BIOS to run a XEON, and it should support 24GB of memory.

Those boards have solid caps, and mil spec parts; they shouldn't age out like most stuff.

It looks like 1402 is the latest bios, but there's a button on the back to load a bios from a usb stick, and it's supposed to work without a cpu.

The board did have two red LEDs that stayed lit when I got the board to boot up. CMOs has been cleared. I'm trying to run a i7 960. Pictures of the board above.
 
i7 960 was a latter edition cpu. What bios is on it now? (you have omitted the detail) You may need a flash just to recognize that cpu.

4+ pins look slightly out of alignment.

Inked003_LI.jpg
 
Last edited:
It looks like 1402 is the latest bios, but there's a button on the back to load a bios from a usb stick, and it's supposed to work without a cpu.

I don't think so, not on the X58 Sabertooth anyhow.
 
I don't think so, not on the X58 Sabertooth anyhow.

You're right; it's only on the later ones.

The Red LEDs being lit is a problem; my x79 stops when one of these is lit.

The manual linked above lists where they are, and what they do.

I'm running an x58 P6TD Deluxe, with an x5670 Xeon, and it's only about 10% slower than my main system, at 4.6GHz.

Those Xeons run Very cool; even in Handbrake, it never went over 70C, and that's a 6 core processor.

It's a shame these won't go over 24GB of memory, even with the Xeon.
 
I have tried with an I7-920, but I have not tried with the Xeon; I have more 8GB DIMM's, so I'll give this a shot.

The worst that can happen is it won't boot, lol.

I will drop the latency down, maybe that was the issue before.

Actually, I think tried 8GB and 4GB DIMMS; I'll have to see how 6x8GB works. :)

6x4GB DIMMS worked for 24GB wbw, so I know those DIMMS worked in that mobo; but I just tried to add the new ones, and it wouldn't.
I'll revisit that.

THANKS! That was a nice link. :)

EDIT: looking closely at your Pix, it looks like there's some crap in DIMM B2's socket, at the right in the pic.
Looks like Cheeto dust or something.

I'd carefully clean the slot out, being Really careful not to band any pins.

Look for crap everywhere on the PCB; something like that will make a multi GHz trace not work at all.

EDIT2: Yeah, that boards filthy.

I'd dismount the CPU, remove the battery, and wash it carefully in soapy water in the sink; you can use a clean flux brush to scrub with, being careful not to remove any of the SMT components.

Let it dry overnight; if you have an air compressor, blowing the water out from under the components first is a good thing. :)

Sticking the corner of a paper towel under the edge of things will also wick it out, but it's more labor intensive.

A dishwasher and alconox is my preferred method, but I wouldn't do that in my home dishwasher, people will complain, lol.

Don't use soap other than Dawn or similar in the sink; Don't use Dawn in a diswasher, lol. (It will fill a room with suds in seconds, lol. Guess how I know?)
 
Last edited:
1) There is something inside RAM slots (dust ?).
2) LGA 1366 requires at least one module in one of the further away slot (otherwise, it can't boot).
 
So no one else noticed he took the the pictures of the motherboard laying on the most static prone carpet floor possible. That my friend is a big no no.
 
Check re VRM's ( unscrew the heatsinks and check for burns, smell, marks etc. also check the back of the motherboard for the same things)
then clear it all with alcohol ( even dirt with some old conductive thermal paste may be shorting staff around)
let it dry and try again ^^
if it doesnt work there is a chance some cmd pop'ed internally in that case you will have to pull out the voltometer and check the whole thing ^^ time wasting if you dont know what to look for sadly
 
if you have an old pair of headphones you can cut off the 3.5mm jack and wire them to the 2 pins on the front panel beside the reset switch pins for post beeps. a video of you powering on the board showing the diagnostic LEDs with the beeps would go a long way to figuring out the problem. other then that the easiest way to see the problem for a more advanced repair would be to use a thermal image camera and not too many people have one of those laying around. another method thats cheaper takes a little more skill is to cover the board in isopropyl alcohol and see what areas evaporate the fastest.
 
Both DIMM A2 and B2 have crap lodged in them
 
Board needs a nice wash. Perhaps even dishwasher treatment.
 
Back
Top