Machine Check Exception - Desparately looking for help (Reward)

Trigmaster15

Honorable
Sep 27, 2012
16
0
10,510
I can't seem to get help with this anywhere else, so here goes. I will send $100 in cash to any person who can provide a solution that works for one week consistently. This is for both my own benefit and any other poor soul struggling with this.

My rig is composed of an ASUS Rampage IV Black Edition X79 motherboard, i7-4930k 3.4 GHz CPU, 16GB of G. Skill Trident X RAM 9-11-11-31 timing, EVGA GTX780 3GB HydroCopper GPU, and a Rosewill 1000W Lightning PSU. I also have 2 SSDs and 1 HDD (samsungs and a WD Black which ran fine previously) The RAM and GPU were running perfectly fine on an old ASUS mobo, and I decided to try to upgrade to get a better processor to help support with a simulation class I am taking for grad school. Let me walk you through my troubleshooting process:

The original mobo I purchased was the EVGA X79 Dark. Upon completion of the assembly, I started up the computer and got a POST no problem. Then I tried to install Windows 8.1 with a OEM CD. When attempting to do this, the loading screen would hang and freeze indefinitely every single time. Called Microsoft tech support and spent almost 4 hours on the phone with them. After countless hours of troubleshooting, reviewers and I agreed to try 2 fixes - replace mobo and get a new windows full version CD.

This leads to the current mobo, the ASUS rampage IV B.E. After everything was put back together for the second time, got a POST no problem at all, and I tried to install windows 8.1 again, using the full version 64 bit disc. Same exact thing, and I receive a Machine_Check_Exception error every single time. Tried all the troubleshooting methods - pulled out all components possible, left in only one stick of RAM, tried a new set of RAM, did memtest, reseated comps, checked all connections, swapped SSD with the other SSD, tried the HDD instead, tried no hard drive at all, etc etc etc etc. Then in an amazing stroke of luck, I decided to try to flash the BIOS, and Wahlah! It worked! Booted up all the way and I was able to install windows and get to the desktop. For one time - and then on the next reboot, it instantly went back to BSOD, even on the new bios update. I have posted on forum after forum, called computer repair stores, etc. The best advice I have been given is to try to replace the CPU, and try a different PSU. Is there any possible thing you guys can think of that I might have missed here? Or what allowed it to actually work ONCE after flashing? Please let me know, and I'm serious about the reward. Sincere thanks for any help whatsoever.
 
edit: your power supply has enough power but it is not the best brand, you might want to monitor the voltages it put out for fluctuations. Some of the motherboard/ overclocker guys will know which program would work best but here are some to choose from:
http://www.mydigitallife.info/check-and-monitor-computer-cpu-and-gpu-temperature-voltage-or-fan-speed/

post a copy of your memory dump, the it will contain details of your processor memory state. The file is pretty large, you should put it on a cloud server with public access (skydrive works)

generally a problem like this will be in the various voltages applied to the CPU, You will need to get all the voltages that the motherboard is setting up from your BIOS. Screen shots of the various voltage settings helps.

This is not a windows problem, it would be a hardware setting issue. You would want to run memtest86 booted from a CD to confirm your hardware works as expected

the crash dump will contain a structure that indicates why the CPU called the bugcheck. Most often it will because the internal memory controller detected a cpu cache errror. This is often directly related to VccPLL voltages being out of spec (too high) and cache error can be a side effect when people attempt to overclock their system but do not turn off some of the other CPU functions that can cause voltage fluctuations to the CPU,( speed step, turbo fucntions, c states, hyper threading)
if the voltage was set high and you had power spikes it actually damage the memory controller/cache inside the cpu. When this happens you will get repeated bugchecks in windows even after you set the correct voltage.
but each bugcheck will indicate the error and have the same cpu memory bank fail. if it is not a bad cpu, and just a incorrect phased locked loop voltage then you will get different banks failing on different bugchecks.

the command to check the data structure in the windows debugger is
!errrec (with the address of the structure as given in the bugcheck data )
pretty easy to check if you want to put your memory dump on a cloud server I will check it for you.


Note: normally you don't want to go to microsoft for a issue like this, they will give you generic hardware advice. They are a software company and will isolate your problem down to the hardware and punt the problem. You should isolate the problem away from windows using memtest86.exe on a boot disk, then have your hardware vendor help you figure out the problem. (they will often already know the answer, the problem will always work its way back to them in the end, via returned hardware that they end up looking at.)



 

Trigmaster15

Honorable
Sep 27, 2012
16
0
10,510
 
edit: you might be able to boot into BIOS, wait 5 or 10 seconds then hit the reset button to reset your cpu and boot
and see if your system works correctly for that boot.

I was thinking about why it would work only for one boot. I think your power supply is a liar.
The power supply is supposed to stabilize its current and regulate its power signals, after they are stable it is supposed to send a power_ok signal to the motherboard, this signal is used by the CPU to tell it the power is as good as it is going to get and to start its initilization. Cheaper powers supplies save some money and don't put the circuit in to check the power output and just fake the signal and hard code the power_ok to always be true.

I think you booted in to bios, updated the bios, this caused the BIOS to call a reset to reinitialize the motherboard electronics. By this time the power supply was already stable and the CPU could function correctly.
I would think that you could reset the BIOS back to defaults each time you boot and have the same effect because
it only takes a few seconds for the power supply to stabilize.

- you should google on how to save and locate your crash dump.
it is in different locations on different systems . c:\windows\minidump\memory.dump
or it can be other names, it is a setting you can select in control panel

- note: VccPLL should be 1.7 volts for your chip as the default unless you have a reason to change it.
you will want to check the setting in your BIOS to confirm.




 

joesklar

Honorable
Sep 17, 2013
358
0
10,960
I would suggest using a KNOWN GOOD Power Supply for troubleshooting. I was building a very similar spec machine for a friend and had the same issue whereby the system would bluescreen, but the work perfectly after clearing CMOS.

Try a new PSU, then post your results.

Best of luck bud.
 
I do think your best option for success it to save your reward and put it to a new quality power supply.
a year or two ago seasonic made all the best power supplies (even ones sold under different brand names were seasonic power supplies) I have not gotten one for a while, I used to look for the correct wattage and for more than one rail. You might start another thread and ask people which ones are actually good to get today or google for reviews. even www.newegg.com will have customer reviews



 

Trigmaster15

Honorable
Sep 27, 2012
16
0
10,510
Guys, I can't thank you enough for all the responses back. As soon as I get back from work today, I'll start working on each one of these and post back the results. Thanks again, I really appreciate it.
 
I just remembered the last power supply I got was for my wifes machine
I gave her a choice between the seasonic or a nzxt and she was very happy with the nzxt
It was a older model of this series i think
http://www.newegg.com/Product/Product.aspx?Item=N82E16817116020

very good, modular and very quiet. to the point she hears my computer and can not here hers.


 

Trigmaster15

Honorable
Sep 27, 2012
16
0
10,510
Quick update - I was trying some easy fixes, and I reset the CMOS on the motherboard. The computer booted up instantly with no problem, 2 times in a row. Then I tried flashing the motherboard, back to square 1, BSOD. Tried resetting CMSO again, still BSOD. Tried flashing again, BSOD. I really wanted to try with my old PSU, but these new motherboards need 2 plugs for the CPU - this PSU only has one. Will need to wait until I can test another supply. Thanks again for all the help.
 

J Dallas

Reputable
May 29, 2014
1
0
4,510
So did changing the power supply work? I'm having exactly the same problem. I'm traveling abroad and my laptop not working is a nightmare. Without my normal tools with me (because I'm an idiot and did not bring them) I wasn't able to open it up and do all the cleaning and connection testing I would normally do so I found some IT guy from California who said for 250 Euros he'd do all the cleaning and testing and reinstall windows and make sure everything was working well but I got my computer back and the second day I had it back the same problems resumed - Check_Machine_Error BSOD, sometimes booting all the way up and working just fine once or twice and then either not making it all the way through the boot process or booting and immediately freezing etc.... When I said the laptop still has the same problem he told me it was just fine when he gave it to me so if I want him to look at it again I have to pay another 125 Euros (because clearly my installing Firefox and AIM made my system unstable... excuse me? You gave me back a "perfectly working computer" on which I just can't install anything? No, I don't think so!) ! I am so mad I don't even know what I'm going to do..... It'll be a while before I even have enough money to buy another Power Supply but if that fixes the problem that's what I'll have to do... but before I throw away more money that I don't have... did it fix the problem?