Question PC won't boot

May 26, 2019
11
0
10
0
Hi Guy's

This my first post so I hope the community can help me, apologies for the long story but I've been down a few avenues already.

PC specs, built October 2018
Gigabyte X470 Ultra Gaming mobo
R7 2700X
GTX1080
EVGA 650W G3 and EVGA 850W G2L
32Gb Corsair Vengence LPX DDR4 3200 CMK16GX4M2B3200C16 (not on the qvl list but had worked for 6 months)
970 Evo Pro (500Gb), 860 Evo (500Gb), 3 HDDs and a DVD drive
Corsair H115i Pro

So my reason for posting, I have an issue with my PC. It's been running fine for 6 months, since I built it; one night a couple of months back whilst watching YouTube the system froze, only allowing me to restart the PC with a hard reset.
I rebooted the PC only to get continuous issues trying to boot to Win10. It would not repair itself, even though it attempted to and did not offer any recovery options.
I started trying to reinstall Windows but could not reach the end of the install process before the system crashed.

The PC will still POST and allow me to enter the BIOS. Ver F30.

I tried installing Windows 10 onto an m.2 nvme, Ssd and Mechanical Hard Drive but all options have failed.

Thinking this might have been a motherboard issue, I sent it back to the shop I bought it from for RMA but it came back a couple weeks later with no fault found.

I then took the system to a local Repair shop, they checked the graphics card and memory and tried a 750W PSU, still no change.
I need to check if they did anything to verify the processor functions correctly because at present I think this could be the culprit?
If the processor is the problem, would it still allow me to start the Win 10 install and run Memtest86?

The repair shop thought the 750W PSU they tried may still be struggling but didn't have a larger one to try, so based on their feedback I put a brand new 850W EVGA in it but still had the same problem.

The PC will let me choose a boot device, either USB or DVD, it will start to try and install Windows but once the circle of dots start spinning the system shuts down again shortly afterwards.

It doesn't seem to want to boot in UEFI mode from the boot menu but will try to boot from the non UEFI options.
I'm not sure if the copies of Windows I have are too old or if the usb 3.0 drives are causing an issue? I read somewhere that UBS 3.0 drives could be problematic and the USB 2.0s weren't.

I have tried various permutations in the BIOS with CSM on/off and to be fair it's all becoming a blurr now as nothing has been successful to this point.
My main problem is not having any spare parts to swap in/out as I'm sure this problem shouldn't be as difficult to diagnose as it has been.

Any help welcomed, thanks in advance.

Chris
 

DavidM012

Reputable
Feb 4, 2016
486
40
4,840
21
Psu Tier List evga G3, tier b budget, evga g2 Tier a+ high end g2L looks like it's got the hi quality japanese capacitors, should be fine.

Well the board's been cleared, clean install of windows didn't work, try disconnecting the nvme drive from the board when reinstall windows on the 860evo ssd p'raps.

The cpu could run windows setup maybe since it's not the stressiest task but, windows 10 setup might all crash similarly to what's going on with your regular install, it's the same os. Memtest similarly wouldn't necessarily crash the cpu.

If your usb drives were working enough to install windows setup media on them, probably not the problem. Any usb drive could be problematic I mean I've had one or two that worked one day and not the next, that were usb 2 while my usb 3 drive is lexar usb 3 caddy for micro sd drives, so I can switch out the micro sd card for several varieties, I like my lexar micro sd because it's 150mb/s for a fast install, no problems for years. These days you can get some really fast ssd speed like usb 3.1 drives, like 425mb/s.

I'd be thinking along the lines of maybe the cpu was getting too much voltage at some point so if you can read the vCore from the bios, and what your llc settings might have been, and an indication of temps. the cpu was running at, what if any overclock/boost was applied.

Can you see the pump rpm in the bios? Is the cooler pump still working? A failing pump could cause an overheat. Check it's secure with good contact and thermal paste.

Can try memtest86 but doesn't sound like the mem. is a problem if the system booted first time and it all ran well for 6 months.

Heading in the direction of cpu, can also try switch gpu with a different one.

Doesn't want to boot in uefi mode sounds glitchy, it would be unfortunate to have a combination of a glitchy system with glitchy usb drives, and maybe the nvme drive is also glitchy, or the cpu or gpu is glitchy. ugh.

Try setup with one drive installed at a time, and with a different gpu if poss. Also set bios to defaults.
 
Last edited:
May 26, 2019
11
0
10
0
Hi DavidMO12,

Thanks for your quick reply.
I currently have the mobo on a wooden block with only the necessary components connected.

The nvme has been removed and I just have the DVD and 860 ssd connected.

The repair shop said they tested the gpu in another machine and it worked fine, so should be ok.

Vcore ~1.4V
3.7 Ghz no oc
Memory running at xmp profile 3200 MHz
29dC temp in bios
Cpu fan is now back to orig wraith, 1750 rpm at present to eliminate the cooler.

I just tried to install the win10 iso on DVD drive, windows logo appeared but it didn't get to the circle of dots before it crashed again.

I am just creating a rufus windows install inner format to see if that makes any difference...
 

DavidM012

Reputable
Feb 4, 2016
486
40
4,840
21
vCore looks a bit high at 1.4 for it maybe that is an indication of something at default it should be :

https://www.google.com/search?q=2700x+default+vcore&rlz=1C1CHBD_en-GBGB847GB847&oq=2700x+default+vcore&aqs=chrome..69i57j0.4991j1j7&sourceid=chrome&ie=UTF-8

Starting to look like a voltage problem maybe with board power regulator tho the 2700x should be able to survive 1.4 volts it should also be able to do 4.2ghz at that vCore. Something looks a bit skew-whiff there.

Dunno what it might have been under a load if it was revving up to 1.5v+ or something, can't see precisely what was going on but it looks like the cpu is dodgy possibly as a result of something being wrong with the board power phases.
 
May 26, 2019
11
0
10
0
I agree it does look high, but to check I just set the clock freq to 4.2ghz and the voltage dropped to 1.25V???

The rufus install made no difference when I created it in mbr format.
 

DavidM012

Reputable
Feb 4, 2016
486
40
4,840
21
looks like some vDroop, cpu not healthy maybe but it must also have a cause so that can only be the power delivery system on the board. Maybe the bios of the board is malfunctioning in some way. Could be some various settings, I suppose you disabled pbo or whatever the boost settings are, so it's all locked in at 3.7ghz, anything that looks like it's a power control, just reduce it to minimum, wonder if the system would even boot if you locked the vCore to 1.25 at 3.7ghz?

I think the default should be even lower than 1.25v at 3.7ghz but I don't have any hands on experience with ryzen yet, so by all means, seek a second opinion perhaps others may volunteer a suggestion. Or a support ticket with gigabyte, they will be familiar with any known glitches of your board.
 

DavidM012

Reputable
Feb 4, 2016
486
40
4,840
21
May 26, 2019
11
0
10
0
I'll have to check with the company i purchased the board from, not sure what their returns policy is on an 8 mth old board that they have already rma'd with allegedly no fault found???

I wonder if they would check my processor/ram for me if I left it on the board?

Thanks for your input David
 

DavidM012

Reputable
Feb 4, 2016
486
40
4,840
21
The board has been given a thumbs down review by gamer's nexus and they didn't pick up on it. If you can get a warranty repair try go direct to gigabyte but it seems much like the actual problem is a design flaw with the line of boards so really the only option to avoid the problem recurring is a different make and model of board altogether. Of course, nobody would try and run with your money. Seem all a bit like a consumer rights issue to me really. I wouldn't be surprised if you get the runaround.

I'm pretty assiduous about reading reviews and checking forums for known issues before buying anything these days, install avoidance strategy if poss. Sometimes the sales blurb gets me hyped so now I look it up to debunk. I had read before that gigabyte vrms weren't the best but it's even worse than expected.
 
May 26, 2019
11
0
10
0
I did start to read a lot of negative reviews a couple of months after i bought it, especially related to the marketting blurb concerning the 8+3 phase vrm that is really only a hyped up 4+3 configuration. The Gaming 7 has much better VRM circuitry and is twice the price.

I was hoping the board was at fault the first time I rma'd it and would have pushed for a superior board then, but when it came back NFF my heart sank, so started looking for another component being at fault.

I'll get onto Gigabyte support to see if the mobo can be reset without having to reflash the BIOS.
 

DavidM012

Reputable
Feb 4, 2016
486
40
4,840
21
A lot of boards use doublers on the vrm but that in itself isn't a problem for ryzen boards last time I read about it, the gigabyte vrms were particularly hot on the b450 boards compared to some other b450's but they were saying the x470's had better quality components. On overclockers . Perhaps even that isn't necessarily the case.

I don't see how the board's going to revive something happened when you entered the lower vCore, hopefully it wasn't the psu giving up as well.

Was the system prebuilt or did you buy the components individually? Or prebuilt and maybe added a few drives or something I dunno, but either way I still think it's better to buy the parts separately and assemble them, than pre builts, and I don't see that a board failing 8 months in, rather than the day after purchase should be any different if it's supposed to have a 1-3 year warranty.

Totally go for something else there's nothing anyone can do about bad quality vrms on a line of mobo's. They are still selling them and so rolling the dice and the customer doesn't get any comeback? Ah, the woes of tech companies and resellers trying to shift b grade stock. Must be heart breaking.
 
May 26, 2019
11
0
10
0
I built it all from brand new parts as I was upgrading from a Bulldozer FX-8150 and had the 860 and hdds already, the 970 was new, I put a new 850W PSU in it yesterday.

I did just get the Dual BIOS to activate and the previous version F6 loaded where I noticed the Vcore was sitting at 1V. I just rebooted the system and it has reverted back to F30 where the Vcore is now 1.38V.
 

DavidM012

Reputable
Feb 4, 2016
486
40
4,840
21
Still abnormally high if the cpu is at default without any pbo and the problem is still severe enough that windows doesn't load so it doesn't really fall into the 'it's working properly' category whatever the glitch is, nothing has altered to stop it from happening again. it's still faulty and probably just going to do the same thing if you try and change any settings, more of the 'it's still not working' category and going round in circles.

See if windows loads, if it's still glitching out, get a warranty refund, dunno if you have any consumer organisation to extract a refund for you, and just get a different board.

I suppose it would be possible for some electronics whizz to identify the mosfet chips and replace them with more quality versions but never heard that anyone would do that for modern mobo's, it's all manufacturer stuff and none of it is user modifiable I mean maybe it's possible to turn an ailing line of boards into good products but it's most likely that it's all in the economically non-viable category or something. They have competition that doesn't have the same problems and they can't charge more to fix it.
 
Last edited:
May 26, 2019
11
0
10
0
Update: The CPU got RMA'd and was replaced under warranty. When I dropped it in yesterday afternoon, it all came back to life again.
 

ASK THE COMMUNITY