New build stuck in indefininte boot cycle (2 seconds on, 5 seconds off)

Toggle sidebar Toggle sidebar
W

WolfDale7

Distinguished
Jul 29, 2011
28
0
18,530
Hi there, this has been bugging me for a week, and I feel Ive pretty muuch tried everything short of RMAing the board. I really would like to avoid dealing with Gigabyte's RMA team as I live in Canada and I've heard lots of terrible stories about them.

Anyway, since I did try everything, I had no other option than to start an RMA request--and so I did. I'm currently waiting for them to get back to me.

Until then, I'm hoping that you guys can give me some input on this. Maybe there's something I haven't yet tried!

This first happened when I was installing Windows 8. The PC booted without a problem when I breadboarded it, so I assumed it was all ready for installation into the case.

After installing everything in the case, I had no issues turning on and off the PC. At this point, I hadn't installed or attempted to install any Operating System.

When I stuck my USB in for the windows 8 installation, everything was working fine--until I was prompted to enter the Windows 8 license key. After I entered it and hit "next", the PC shut down and started to cycle indefinitely. I'm not even getting the post screen, and the mobo speaker isn't letting any beeps before the restart.

I let it reboot 3 or 4 times thinking that this was part of the Windows 8 installation process, When I did notice that this was a boot cycle, I proceeded to try some troubleshooting techniques.

One thing to note is that when I let the PC rest for 6+ hrs, it's able to boot for about 30 minutes, before re-entering the boot cycle.

Here's what I tried:
- "Breadboarded" the system (completed an installation outside the case)
- Installed the ram sticks in multiple combinations (one stick in DIMM slot 1, a different DIMM slot 1, a stick in DIMM slot 3, sticks in DIMM slots 1 and 3, etc.)
- Installed a different brand of RAM sticks (first, Crucial; second, Mushkin)
- Tried booting without the Graphics card (GTX 780 Ti)
- Ensured 12v ATX was plugged in tight
- Unplugged and replugged all cables from Power supply (EVGA SuperNova 850W G2)
- Tried a second Power Supply (Corsair RM850W)
- Reseated the CPU heatsink
- Reseated the CPU and the CPU heatsink
- Cleared CMOS
- Used a different CMOS battery
- Ensured all stand-offs are in place
- Ensured all memory modules were fully inserted
- Ensured CPU was installed correctly
- No bent pins on the motherboard visible.
- Ensured CPU fan plugged in and functioning

When the computer turned on, I checked to verify that the BIOS version is version F4 (ensures compatibility with Ivy Bridge - E processors). CPU temps seemed to be normal (all cores <38 degrees Celsius). PSU voltages are not fluctuating.

This has led me to surmise that the problem is resulting from a defective motherboard. This is a brand new motherboard being installed for the first time.


System Specs:
Motherboard: Gigabyte GA-X79-UP4
CPU: Intel 4930K
CPU Cooler: Corsair H100i
GPU: ZOTAC Reference GTX 780 Ti
Ram: Crucial DDR3 1866 CAS 8 8*4gb (32gb)
PSU: EVGA Supernova 850W G2

Does anyone have any insight as to what might cause this? It seems to be occuring with different part configurations... Has to be the Motherboard or CPU..

I mean, CPU temps seem normal, PCB temps looked okay, the motherboard recognized the ram in the bios... What's left? If it's the board, I'll continue with the RMA.

Thanks guys =]

 
Solution
L


Then I think you have a bad motherboard, I had a cycling issue once I ruled out the drives, and RMAed the MB the issue stopped, now I have a new issue, don't you just love RMAs?
Sort by date Sort by votes
W

WolfDale7

Distinguished
Jul 29, 2011
28
0
18,530
@JackNaylor I kept the temps page up for about 20 minutes, the temps went up and down between 35-38/core. Nothing too alarming there.

I'll run Memtest x86 later tonight, but I doubt it'll help. I tried 12 different sticks in slot 1 and 3, also tried various combinations.


@littleleo, I'm not even posting, won't post without any hard drives plugged in (tried 2 SSDs + 1 HDD; 1 SSD; the other SSD; the HDD by itself; none). No combination worked :(
 
Upvote 0 Downvote
W

WolfDale7

Distinguished
Jul 29, 2011
28
0
18,530
@JackNaylorPE, I just ran memtest x86. I was able to complete only one pass. There was no error with the single module placed in dimm slot 1. The test was able to complete only one pass however, so not statistically enough to give us information on whether the stick is healthy or not, but it should be enough to keep the pc powered.

The pc is cycling again, and I would have to wait till morning before I can run a second test.

Any other ideas?
 
Upvote 0 Downvote
W

WolfDale7

Distinguished
Jul 29, 2011
28
0
18,530
Yea, I was assuming errors should pop up right away, or at least within the 25 minutes it was running.

Is there any chance that the cpu is DOA, or is it possibly just the mobo?
 
Upvote 0 Downvote
JackNaylorPE

JackNaylorPE

Titan
BANNED
Oct 30, 2004
32,255
14
99,960
I wuda been on the phone w/ Giga TS on the 1st day asking them to guide you thru the troubleshooting procedure.

And yes, Gigabyte RAM does suck..... so does Asus ..... so does MSI .... so does everyone else ..... worse I have had to date was EVGA .... 18 months, 20 calls and 5 RMAs for one problem w/ factory OC GFX card (wouldn't run at factory OC).
 
Upvote 0 Downvote
W

WolfDale7

Distinguished
Jul 29, 2011
28
0
18,530
That's surprising to hear--your troubles with EVGA--if you registered your product on the EVGA site, you should be elegible for premium support.
Both times I went through Evga, they have been nothing but effecient.
I actually called in because my psu has coil whine (not considered a defect). They were willing to ship me out a replacement on the condition I give them my credit card info for collateral. At which point I am obligated to forward them my old psu via a prepaid shipping box within 48hrs. From initial contact to ticket creation, it took about 5 minutes.

I live in Canada and gigabyte support only has a California number. I'm forced to pay long distance fees to get in touch. I spoke to them 3 times already, and also sent an RMA request 3 days ago. I have yet to hear back. They also expect me to ship a product that left their factory defective on my own expense...

I did go through a few troubleshooting steps with gigabyte on day one. They told me to do various things with the ram, update my bios firmware, clear cmos, etc. So far, no luck.
They might even ding me with duties when having the warrantied item shipped back..
 
Upvote 0 Downvote
JackNaylorPE

JackNaylorPE

Titan
BANNED
Oct 30, 2004
32,255
14
99,960
Each of the 20 calls was Grundhog Day" like, very polite, basically saying the same things over and over again. Each call I had to give them the equipment list, go into the BIOS and read various settings, and go thru each of the exact same conversations, suggested reasons it could be doing that which was followed by how I had already ruled that out.

-They blamed the PSU .... jonnyguru 100 rated
-They blamed the MoBo .... Asus Rampage Extreme
-They blamed the RAM ....Mushkin Redline
-They blamed the CPU OC .... wasn't OC'd
-They blamed the GPU OC .... not only wasn't it OC'd above the factory overclock, it was at the reference card core speed.
We went thru the original card and 4 replacements all with the same results. None would run at the factory OC, the first one was closest to it.
-The argued that since we had now gone thru 5 separate cards, it must be something else. Yet they could not explain why, when I was building another box, I was able to "borrow" 2 Asus Cards in and overclock them a whopping 30% Their factory OC on the FTW was lil less than 10% as I recall) and they ran perfectly stable.
They finally sent a reference next gen card which was a terrible overclocker but it was faster than the one we sent back 5 times.

 
Upvote 0 Downvote
W

WolfDale7

Distinguished
Jul 29, 2011
28
0
18,530
That is quite odd, but logically, yes, if you tried 5 DIFFERENT units (same model, but legitametly different), its got to be something else.

Really, it could be as simple as incompatibility or as serious as a design flaw, but I sympathize with your hesitation with EVGA should I have gone through a similar experience.

Anyway, let's keep this topic-related. Need to see what I should do next. Probably will have to calgigabyte if they don't get back to me by tomorrow an hour before closing.
 
Upvote 0 Downvote
JackNaylorPE

JackNaylorPE

Titan
BANNED
Oct 30, 2004
32,255
14
99,960
While everybody else was using custom PCBs and beefier VRMS, EVGA stuck with the reference PCBs and stock VRMs on the SC and some of the FTW series. As a result, the EVGA took it on the chin on the 570 series as many overclockers with reference PCbs from several brands fried their VRMs, while the vendors with more phases on the VRMs did just fine. My guess is the factory OC on the cards was just a bit optimistic at the voltage EVGA had them designed to run at.


Most manufacturer's have extended hours, I have called Asus (US) several times in evening hours, early morning or even weekends .... Best time since they in California is right after they open as they open for east coast time when the rest of the time zones are still sleeping .
 
Upvote 0 Downvote
L

littleleo

Splendid
May 8, 2009
4,097
17
23,465


Then I think you have a bad motherboard, I had a cycling issue once I ruled out the drives, and RMAed the MB the issue stopped, now I have a new issue, don't you just love RMAs?
 
Upvote 0 Downvote
Solution
W

WolfDale7

Distinguished
Jul 29, 2011
28
0
18,530
[strike]Memtest x86 is reporting the ram frequency at 800mhz with XMP. Is there any chance that this might be causing it?

Unfortunately I just went through 20 mins of "ON" time. Now I'll have to wait for tmr morning to adjust the frequency, but maybe a start? Is it at all possible that this is due to a glitch in how the mobo is reading all my ram sticks? Probably due to the bios firmware version?
[/strike]
egr3va.jpg


vdd7iq.jpg


2eydwqu.jpg


[strike]
Could it be that memtest is reporting a single data rate (DDR=Double Data Rate), so 800mhz is actually 1600mhz?
[/strike]

Yea, dual channel, memtest is registering half. I'll update the bios firmware as a hail Mary... Hopefully it doesn't crash during install!
 
Upvote 0 Downvote
W

WolfDale7

Distinguished
Jul 29, 2011
28
0
18,530
Well, the advanced replacement just came in from gigabyte yesterday afternoon. Spent the night putting it all back together again and I'm glad to say that the issue was resolved.

Could have been anything on the motherboard. I don't see bent pins, but it could very well be a memory issue, bent board (nothing I can see), etc.

Windows 8 has been installed and has ran well for 3+ hrs. Will do more rigorous testing as time trees up.


Thanks again everyone! Appreciate your help.
 
Upvote 0 Downvote
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom