Boot issues

[Tunnelcat]

I've just started running into an issue with my gaming rig. I press the power button, the boot process starts, the POST beep occurs but then it hangs at the boot screen. Even though it's displaying the BIOS screen, you can't get into it. Everything is frozen. The motherboard is all lit up, the fans are running and even the 2 USB keyboards are lit up but no response from the system. I tried the reset button, but it boots right back into the same situation. So if I fully power it down with the main power button and then turn it back on, it finally boots into Windows. It did it 2 days ago, booted fine yesterday, but then it did it today. Any ideas on where to start troubleshooting? I already ran an OS drive scan in Windows and it found no errors. It's only a 2 year old rig with an EVGA 1200W power supply, running on an ASUS Rampage IV Extreme Omega motherboard with an Nvidia GTX 2070 graphics card and 32 GB's of RAM The main OS drive is an NVMe 250 GB Samsung drive that stays cool while running and there are two 1T Samsung SSD's for game storage. There are also a couple of Optical drives and a WD 2T spinner HDD for storage. I haven't messed around in the BIOS either.

[Krom]

The simple ones would be low BIOS battery (check for clock drift or loss of BIOS settings) or lose connections (power, ram, video card).
Less simple: memory failure (run windows extended memory diagnostic).
PSU failure is possible, extended periods of low loading seems to cause more damage than high loading and that PSU is easily double the size that hardware needs.
Even less simple: motherboard giving out (could be anywhere).

[Tunnelcat]

I ran the Windows Memory Diagnostic and no issues were found. Memory good. Power connections are fine. I forgot that this motherboard has a small led screen that will display error codes when booting. Next time it happens, I will see if the MB senses where there's an issue. I'll also check the Q-LED lights to see where things may have hung up as well. I'll have to get back with you on this because it only does it from a cold boot, not from a warm reboot.

By the way, what size power supply would be optimal for this Intel i9 system? I oversized it just to future proof things for the possible later installation of a faster graphics card, especially if I were to go the SLI route. Of course, you can't find a newer graphics card to buy even if you wanted to, so that's not something that's going to happen soon.

[Krom]

550-650W would have been fine. I have a 9900k and a GTX1080, if I run furmark and prime95 AVX at the same time I can barely talk the system up to 400w, even the 850 I have is overkill.

Yeah tell me about it, been interested in getting a 3080 for months (worse because I need a specific card to water cool). Trying to buy a GPU right now isn't just a dumpster fire, its a volcano erupting under the whole freaking landfill.

[Vander]

I have a client that did a trade in with EVGA in early December. He had a 2080TI that was purchased a few months before, and wanted a 3090. There was no availability, so he ended up calling EVGA. They had him send in the 2080TI, and they sent him a 3090 for $1000 off. The whole process took about a week, which was about half as long as it took to procure the 2080TI in the first place.

[Tunnelcat]

550-650W would have been fine. I have a 9900k and a GTX1080, if I run furmark and prime95 AVX at the same time I can barely talk the system up to 400w, even the 850 I have is overkill.

Yeah tell me about it, been interested in getting a 3080 for months (worse because I need a specific card to water cool). Trying to buy a GPU right now isn't just a dumpster fire, its a volcano erupting under the whole freaking landfill.

I wonder how much the Covid pandemic is affecting this? I've also heard that even Ford is having trouble obtaining electronic parts for their cars and had to shut down some assembly lines because of the shortage. I do have a modular Corsair 850 Watt PS in my old Win 7 computer if it does turn out to be a power supply issue.

[fliptw]

this reminds me of what was happening with my old 486: the culprit was the power circuit it was plugged into, maybe there was too much load on the hole circuit, I can't remember, running a power cable to another circuit solved that.

[Krom]

I wonder how much the Covid pandemic is affecting this? I've also heard that even Ford is having trouble obtaining electronic parts for their cars and had to shut down some assembly lines because of the shortage. I do have a modular Corsair 850 Watt PS in my old Win 7 computer if it does turn out to be a power supply issue.

It is having a big impact, lots of people who can work from home aren't going out and spending money on entertainment, so they are starved for entertainment and they have more money than they normally would so $700 for a GPU is not at all a lift for them. The demand is beyond anything the industry has ever seen and shows no signs of letting up.

[Tunnelcat]

this reminds me of what was happening with my old 486: the culprit was the power circuit it was plugged into, maybe there was too much load on the hole circuit, I can't remember, running a power cable to another circuit solved that.

The circuit it's plugged into is beefy and has 12 gauge wire with a 20 amp plug. We specified it when we built the house. It IS plugged into an APC 1500 UPS before the wall socket however and we did have a major couple of power burps before it went out for 6 hours a few days ago. I had to get up at 1 AM just to turn off the 2 UPS units I'm using since I didn't know how long the power was going to be out. They were fired back up in the morning. I doubt that was long enough to empty the CMOS battery. The BIOS time was correct too. The UPS batteries are 4 years old but still hold a charge. Could the UPS be causing issues?

Krom, IF my PS is starting to have a soft failure, will it slowly get worse or eventually die suddenly altogether? I noticed online that even power supply's are in short supply. I hope my Corsair has enough plug ports if I need to cannibalize my old Win 7 system to use it for the newer rig. I'll probably end up doing that anyway since it's so old it can't really run Win 10 all that well.

[Krom]

A failing PSU would eventually just stop powering on the system entirely. Definitely keep notes on those debug error codes, write them down even, if it is the same one every time then you have a likely suspect (either that one, or the one immediately after it in sequence) but if it is different often then it starts looking more like a power issue either on or off the motherboard.

[MD-2389]

Did you look through the fan grill to see if you had any bulging capacitors? A blown cap can cause weird issues like this.

[Tunnelcat]

I can see through the whole side of the case because it's clear. Nothing is visibly showing as nasty with any capacitors on the board that I can see, but there are others hidden under heat shields I can't. What's weird is that it has booted fine for days now. No issues. If it does it again, I'll watch for those error codes on the motherboard. If it is the MB, I'm out of luck since it's long out of warranty.

[Nosferatu]

I was wondering if you have checked to see if there is a bios upgrade for you MOBO.
I actually have seen things like that caused by a bug in the bios that was fixed in later releases.
(I fixed memory issues on an old Dell Precision T5500 with just a bios upgrade)

Short of that, it kind of sounds like you could have some oxidation on the power connectors or even on the PCIe connector itself.

I like to have some of this on hand.

https://www.amazon.com/CRC-05103-Electr ... 165&sr=8-2

Its meant for engines but also works wonders on all sorts of electronics including computers.
(You can also find it in the automotive section of Walmart)

(And no I am not affiliated with them. Just a fairly happy customer. )

If it were me diagnosing that, I would remove the graphic card, use some cleaner on the PCIe slot, go back over it with canned air to make sure its dried out, then reseat the graphic card.

I would also use some of the cleaner on the power connectors, letting them dry before plugging them back on to the card.

Then I would go to the power supply end of that power cable, take it out and give it a shot of the cleaner, let that dry and plug it back in.

I have also cleaned up memory slots like this which have solved memory issues in the past.

Just throwing these out as some things that came to mind in case it helps.

EDIT: I just thought of another little something to look out for. GTX 2070's are 2 power connector cards. You didn't use a loop back power connector on the same cable did you? That would be bad. Use two separate cables to the power supply. That power supply is way overkill and should be able to handle it with no problem.

[Tunnelcat]

OK, it finally did it again. Got Code A2 Detect HDD, so it's getting stuck detecting the boot drive. or one of the drives for that matter. I currently have 3 Samsung SSD drives installed, the OS drive is a 512GB 970 Pro NVMe, the other 2 are 2TB 860 Pros and there's also a Western Digital 2TB mechanical drive for bulk storage. In fact, something sounded weird with either a mechanical or an optical drive when it tried to boot this time. I'm getting a signal to the monitor because it lights up with the logo, but I can't get into the BIOS. No, I haven't updated the BIOS yet since ASUS had a sketchy BIOS update last time around, but I'll check again. Some have tried cleaning SATA and RAM connections like Nostferatu suggested, some have cleared the CMOS and changed the battery ( I did have a power outage that lasted 6 hours, but is that long enough to drain the battery?) and some even fixed it by using a HDMI connection instead of DisplayPort on their monitor. The system will only boot normally after this error with a hard power down reboot, not a reset. Any more ideas to narrow it down a bit before I go cleaning and re-seating things and maybe updating the BIOS?

[Krom]

Make sure your drives are all healthy for one, run crystal disk info and see what the SMART monitoring says (I actually have that program run in the background all the time and it automatically rescans my disks once a day, it has caught the early warnings of a disk failure before and let me copy all the data off the offending drive relatively easily).

[Tunnelcat]

I currently have Samsung Magician monitoring the drives, but it only really works for the SSD's. If the Western Digital crapped out, it's not the end of the world. I'll run Samsung's diagnostic first since it does have Smart monitoring. Is the Smart Monitoring program you use free or a paid version? If Magician doesn't tell me anything, I'll try that out. I also found out that any ASUS BIOS for my Omega MB above 0802 (which is pretty old itself) has nasty issues with reducing CPU performance due to all the attempted fixes for the Specter and Meltdown vulnerabilities, so I'm not too excited to update it beyond that version.

[Krom]

Crystal disk info is freeware and it shows you the smart info for both mechanical disks and SSDs.

[Tunnelcat]

Crystal Disk is showing no issues so far. However, I think I accidentally figured out the culprit. Every time the issue occurred, one of the optical drives made a weird double noise during boot, then the system would lock up. The optical drive I assumed was the culprit was an older Sony unit that was new and never used. I had added it to the system not long ago because I had the space, it's older and is thus more likely to have less trouble reading older DVD's. Well, I decided to unplug it's SATA connector and try things out. Well, so far, no boot issues have reared their ugly head. Keeping fingers crossed here, even though none of the optical drives are even listed in the BIOS boot menu.