My "FF XIV has crashed again" jpurnal.

[Dylfin]

I can't get in now either, it keeps crashing at loading character.

I have this shit from open beta. And this happens with new characters too after tutorial. But more worse is that only 9 days of subscription left and I didn't play even a day.

[catch22atplay]

Buy a new PSU as that one seems to be putting out 7 volts instead of 12 volts on the 12+ volt line. Might want to check it in the bios as the program you're using may not be working correctly. A volt meter would be nice to confirm volts as well. Check at idle and most especially under load.

If it is putting out 7 volts instead of 12 volts it's possible you have a +5v line crossed with a +12v line. Hm actually it would be the negative side putting out 5 volts and the positive side putting out 12 volts, which equals +7 volts. Gotta luv DC. This may be internal of the PSU or external of the PSU. In either case it is not normal and should be looked at before you fry something. If you're using any adaptor cables please check them. Ensure you're not using a proprietary PSU that is not compatible with your system. And ensure you ARE using a proprietary PSU for those type of systems. xeon systems for example if memory serves. Also if memory serves there is a 4, 6 or 8 pin plug that connects to the motherboard. The proprietary and non proprietary do NOT have the same voltages on this plug. This may be the whole issue right here.

Something definitely isn't right here.

[skitatic]

Now this

"Faulting application name: ffxivlauncher.exe, version: 1.0.0.0, time stamp: 0x522527fd
Faulting module name: ffxivlauncher.exe, version: 1.0.0.0, time stamp: 0x522527fd
Exception code: 0xc0000417
Fault offset: 0x00085a65
Faulting process id: 0x1988
Faulting application start time: 0x01ceac7e84737b08
Faulting application path: M:\SquareEnix\FINAL FANTASY XIV - A Realm Reborn\boot\ffxivlauncher.exe
Faulting module path: M:\SquareEnix\FINAL FANTASY XIV - A Realm Reborn\boot\ffxivlauncher.exe
Report Id: c7aac667-1871-11e3-8bb4-08606e88004c"


It will simply just crash the launcher. FFS. Apparently it knows when I have put in the wrong password, but the right one. BANG CRASH! *sigh*

[skitatic]

Buy a new PSU as that one seems to be putting out 7 volts instead of 12 volts on the 12+ volt line. Might want to check it in the bios as the program you're using may not be working correctly. A volt meter would be nice to confirm volts as well. Check at idle and most especially under load.

If it is putting out 7 volts instead of 12 volts it's possible you have a +5v line crossed with a +12v line. Hm actually it would be the negative side putting out 5 volts and the positive side putting out 12 volts, which equals +7 volts. Gotta luv DC. This may be internal of the PSU or external of the PSU. In either case it is not normal and should be looked at before you fry something. If you're using any adaptor cables please check them. Ensure you're not using a proprietary PSU that is not compatible with your system. And ensure you ARE using a proprietary PSU for those type of systems. xeon systems for example if memory serves. Also if memory serves there is a 4, 6 or 8 pin plug that connects to the motherboard. The proprietary and non proprietary do NOT have the same voltages on this plug. This may be the whole issue right here.

Something definitely isn't right here.

Would you know of a good volt meter?

[catch22atplay]

I use Flukes myself and they're 300 bones each. A cheap one from radioshack or wherever for 20 bucks is just fine for quick testing. List your mobo model as well as your psu model please.

Also as it's a workstation it uses error code correcting ram. But i believe some workstations or chipsets systems you can run normal ram as an option. I would think there'd be an option in the bios to enable ecc or disable ecc. Just another thing to look into. I've never had an ecc ram system so just taking a shot in the dark here. And i believe the proprietary PSU's are typically for Dell systems. Not sure if any others have the weird pin out. Just thought i'd mention it as seeing 7 volts leads me to think something fishy is going on here.

And when you check the 12 volts with the volt meter check the psu first. Read the label and see how many rails it has. Yours may have 3 12 volts rails in which 1 of those 3 are bad. So you'd have to check most every yellow 12 volt line that plugs into the mobo. If there's 3 rails it'll say +12v 20amp-----+12v 20amp-----+12v 20amp for example. Basically it'll list it 3 times where as a single rail psu like mine only lists +12 volts once.

Again check the volts from your bios/cmos. This will show the correct volts and eliminate the windows mobo monitor program as possibly being at fault. Do that before spending money on a volt meter.

[skitatic]

I use Flukes myself and they're 300 bones each. A cheap one from radioshack or wherever for 20 bucks is just fine for quick testing. List your mobo model as well as your psu model please.

Also as it's a workstation it uses error code correcting ram. But i believe some workstations or chipsets systems you can run normal ram as an option. I would think there'd be an option in the bios to enable ecc or disable ecc. Just another thing to look into. I've never had an ecc ram system so just taking a shot in the dark here. And i believe the proprietary PSU's are typically for Dell systems. Not sure if any others have the weird pin out. Just thought i'd mention it as seeing 7 volts leads me to think something fishy is going on here.

Thanks for the help so far dude. Mobo and PSU are a ASUS P8Z77-V LK Motherboard and Corsair TX-650 V2 Power Supply

[catch22atplay]

Cool that PSU is built by Seasonics and not CWT. And it's a single rail design. Both of which i prefer, design and manufacturer. A very good PSU. If it only put out 7 volts like your mobo monitor shows it's supposed to shut off. So i'm leaning towards the utility not showing the correct voltage. And the 8 pin plug is the standard EATX plug from the PSU to the mobo. So no issues there. We may be spinning our wheels here until you check the voltage from cmos.

[skitatic]

Cool that PSU is built by Seasonics and not CWT. And it's a single rail design. Both of which i prefer, design and manufacturer. A very good PSU. If it only put out 7 volts like your mobo monitor shows it's supposed to shut off. So i'm leaning towards the utility not showing the correct voltage. And the 8 pin plug is the standard EATX plug from the PSU to the mobo. So no issues there. We may be spinning our wheels here until you check the voltage from cmos.

Hey mate, this is what the bios is telling me which I think is what you are after. It seems to be showing 12V

http://img.photobucket.com/albums/v4...atic/bios.jpeg

[Rhalix]

It's your card.
I run a GTX 660 Ti OC and I had this exact problem. Exactly. I did everything you've done and more. I tried to undo the overclock. I was fine for a couple days and then it started happening again. I did A LOT of research and came to find that this game simply hates the GTX 660 (and a few other cards). The GTX 660 is actually a pretty faulty card in general and requires being under clocked to run this game at a stable state. The fans don't run like they should and the card fails to report correct temps. Your card could be secretly over heating. I reapplied TIM to my card to attempt to fix this issue with no luck. My next step to a perma fix is flashing fresh bios to the card. If no luck then I will purchase a new card. Good luck to you.

This is my temp fix while playing. I have been solid ever since. (Now I just have to figure out why the client randomly stop responding -.-)
(In afterburner)
I currently have my Voltage at +26, Power at +120, and core/mem at -20. My fan incline is 35/50%, 40/80% and 60/100%
(In client)
Borderless Window mode
Shadows at max/norm settings
Mob shadows off
NPC shadows off
Battle effects for player maxed, party maxed, all else simple.
Only show player names when in combat.

By doing this temp fix I've only lost roughly 4 FPS. off my constant 75+ FPS lag free with no GPU crashes for weeks.
I hope this temp fix works for you. I know it's a real pain in the ass trouble shooting this one.

[skitatic]

It's your card.
I run a GTX 660 Ti OC and I had this exact problem. Exactly. I did everything you've done and more. I tried to undo the overclock. I was fine for a couple days and then it started happening again. I did A LOT of research and came to find that this game simply hates the GTX 660 (and a few other cards). The GTX 660 is actually a pretty faulty card in general and requires being under clocked to run this game at a stable state. The fans don't run like they should and the card fails to report correct temps. Your card could be secretly over heating. I reapplied TIM to my card to attempt to fix this issue with no luck. My next step to a perma fix is flashing fresh bios to the card. If no luck then I will purchase a new card. Good luck to you.

This is my temp fix while playing. I have been solid ever since. (Now I just have to figure out why the client randomly stop responding -.-)
(In afterburner)
I currently have my Voltage at +26, Power at +120, and core/mem at -20. My fan incline is 35/50%, 40/80% and 60/100%
(In client)
Borderless Window mode
Shadows at max/norm settings
Mob shadows off
NPC shadows off
Battle effects for player maxed, party maxed, all else simple.
Only show player names when in combat.

By doing this temp fix I've only lost roughly 4 FPS. off my constant 75+ FPS lag free with no GPU crashes for weeks.
I hope this temp fix works for you. I know it's a real pain in the ass trouble shooting this one.

Hmm, ok thanks for the input mate. Its just odd that I have the exact same card as a friend of mine, who has not experienced any issues at all. I will however try what you've said and see how we go. Unfortunately can't at the moment as the servers are down for maintenance.