[SOLVED] All Browsers Crash - Mail Crashes

[A BOM]

Last summer I started to encounter 'Ah Snap Status_Access_Violations' while using Google Chrome. Microsoft Edge gave similar errors. These instances were rare, but the frequency of instances steadily increased. Refreshing pages multiple times would sometimes allow the page to be viewed, and sometimes a new error, Status_Breakpoint, would be encountered. I tried multiple paths to resolution with nothing working (no extensions installed, verified latest version of edge, repair, reset Edge, DISM, SFC, new user admin account created, and many more steps). It became apparent that a reinstall of Windows 10 was warranted and likely the easiest path to resolution.

I performed a complete clean reinstall (SSD formatted prior installation). The problem persisted immediately with the new installation on Edge. I installed other browsers and none functioned correctly either. Edge, Chrome, and CCleaner browsers, all chromium based, encounter the same errors. Firefox will not even launch without crashing (never actually loaded a page).

Since the clean install the mail app has been also crashing. I had an issue originally creating a mail account (yahoo account) that was resolved by forcing the mail app to update from the Microsoft store. Since successfully creating an account the app crashes randomly.

I have tried all of the same troubleshooting measures listed above and found no windows errors, integrity violations. I have tesed ram, CPU, performed chkdisk, verified all drivers are up to date, and windows updates are installed. There is minimal software installed on this machine as it is was a fresh install, but it handles CPU as well as GPU intensive games well. Steam (gaming application) too has chromium based store and library features and those too are getting consistent errors.

I have upgraded my windows 10 home version to pro in order to use group policy editor and forcing edge to revert to older versions with each version experiencing the same errors / crashes.

I have tested the machine with the graphics card removed using onboard graphics with the same results.

I performed another clean install on a new hard drive (new, M.2) . This time I created a new Microsoft account as well. Same errors.

I have installed OE motherboard drivers (up to date versions form MSI website), used windows recommended drivers, drivers from intel driver tool with no effect on browser errors.

I have been through network settings multiple times as well. Other machines on this network are not experiencing any issues.

Browser problems persist in safe mode as well.

Machine Build:
MSI Z390A-Pro mb (original mb, have since replaced with MSI MPG Z390 Gaming Edge AC)
16 Gig DDR4 3300 RAM (running at 2667)
I7 9700k processor (no over clock, except intel boost, 3.6 ghz base up to 4.8 ghz, note that I have tested with boost turned off in bios)
RTX 2070 graphics card
Samsung M.2 NVME 970 evo plus

The time frame for these issues coincides with the 20H1 update for Windows 10. Soon after, I also recall some driver updates at that time, though I cannot remember for certain the timing on the issues and updates.

I'm not sure what steps to take next. Any guidance or suggestions would be appreciated.

 

[Ralston18]

Look in Reliability History and Event Viewer.

Do you see any error codes, warnings, or even informational events that occur just before or at the time of the mail crashes?

Or any other crash for that matter?

PSU: make, model, wattage, age, condition (original, new, refurbished, used)? History of heavy use for gaming, video editing, or even mining?

Disk drive(s): make, model, capacity, how full?

 

[mdd1963]

test system with just one one RAM module installed, alternately testing each for a day or so....(might want to see what RAM voltage is reported as supplied within BIOS, default if 1.10V, just to see if perhaps yours is a tad low leading to instability)

Or borrow a pair of sticks (or even one) from someone else....

I'd also do a fresh install from a newly created Win 10 install USB, using the latest WIn10 build, and, using a different USB flash drive and stick with just a USB 2.0 port.(use a different port than usual, just in case there is some induced corruption/dropped bits on whatever port you'd normally use)

 

[A BOM]

Look in Reliability History and Event Viewer.

Do you see any error codes, warnings, or even informational events that occur just before or at the time of the mail crashes?

Or any other crash for that matter?

PSU: make, model, wattage, age, condition (original, new, refurbished, used)? History of heavy use for gaming, video editing, or even mining?

Disk drive(s): make, model, capacity, how full?

There are warnings for distributedCOM and a couple errors for kernel-eventtracing. There is never anything related to browsers. No events that appear significant in relation to browser or mail issues.

PSU - (will update with specs asap) Original, been in use for 2 years. No mining or video editing, but this machine does game. Never seen any issues supply power dips.

Disk Drives - Currently 3 installed. Samsung Evo M.2 1Tbyte, 500+G free, OS installed here. 250G SSD, was original location of OS, currently 100G Free. HDD, 1TB, 700G Free.

 

[A BOM]

test system with just one one RAM module installed, alternately testing each for a day or so....(might want to see what RAM voltage is reported as supplied within BIOS, default if 1.10V, just to see if perhaps yours is a tad low leading to instability)

Or borrow a pair of sticks (or even one) from someone else....

I'd also do a fresh install from a newly created Win 10 install USB, using the latest WIn10 build, and, using a different USB flash drive and stick with just a USB 2.0 port.(use a different port than usual, just in case there is some induced corruption/dropped bits on whatever port you'd normally use)

This build originally had 32 GB ram (4 x 8G). When problems started to occur I removed two sticks (remaining two are in proper locations) and ran ram test (memtest86, 4 full passes) for approximately 14 hours. Zero faults detected. I have also tested with one ram module as well as other two currently not installed modules. No difference in performance.

The last reinstall I performed was on a brand new USB, fresh install file, and used a different usb port. Used port on back of machine on MB instead of cabled USB port on front of case as with previous installations.

 

[A BOM]

PSU - thermaltake smart, 600 watt

 

[mdd1963]

Not many parts remaining left other than CPU, mainboard, or trying another network path (if using wireless, try wired, or, vice versa, just to rule out some sort of weird network induced corruption?)

If mainboard were at fault, you might try tweaking up RAM voltage .05V and CPU core voltage by .05 over default, just to see if it helps. (This would never be necessary normally, but, you are facing strange symptoms, and, no one wants to buy another mainboard and/or CPU, and, whichever way you guessed, it'd probably be the other anyway...

GOod luck!

 

[Ralston18]

My thought is there are two problems to consider:

1) the 600 watt PSU does not provide enough wattage to support the build. For example, the GPU's recommended PSU wattage is 550 watts.

https://www.asus.com/Motherboards-Components/Graphics-Cards/Dual/DUAL-RTX2070-8G/techspec/

[Do verify that I identified the correct GPU.]

Most likely that the GPU does not demand all that wattage on its' own but if the system cumulative wattage demand is nearing 600 watt (especially at peak power requirements) then there will be problems.

2) The PSU itself being two years old and heavily used means that the PSU may be nearing its' built in EOL (End of Life). PSU may no longer be capable of responding with the necessary power and/or rapid changes thereof.

You did not provide the rating etc. but if the PSU is a lower end product its' demise may come sooner rather than later.

Try to obtain another known working, higher wattage PSU for testing.

Remember do not mix and match in PSU cables from other PSUs.

 

[A BOM]

My thought is there are two problems to consider:

1) the 600 watt PSU does not provide enough wattage to support the build. For example, the GPU's recommended PSU wattage is 550 watts.

https://www.asus.com/Motherboards-Components/Graphics-Cards/Dual/DUAL-RTX2070-8G/techspec/

[Do verify that I identified the correct GPU.]

Most likely that the GPU does not demand all that wattage on its' own but if the system cumulative wattage demand is nearing 600 watt (especially at peak power requirements) then there will be problems.

2) The PSU itself being two years old and heavily used means that the PSU may be nearing its' built in EOL (End of Life). PSU may no longer be capable of responding with the necessary power and/or rapid changes thereof.

You did not provide the rating etc. but if the PSU is a lower end product its' demise may come sooner rather than later.

Try to obtain another known working, higher wattage PSU for testing.

Remember do not mix and match in PSU cables from other PSUs.

80+ gold. Since i tested with the graphics card removed, i have hesitation that it could be the power supply. The power requirements systemwide under those conditions would be ~ 250 watts. I have looked at the voltage outputs under load (cpu, ram, 12v, gpu) and they appear to be normal.

I agree that MSI recommends 650 for a machine running these cards so a power supply upgrade couldn’t hurt. A power supply calculator when fed my specs estimates 450 watts.

Also considering the browser errors happen loaded or unloaded, with edge being the only program running, a PSU problem seems less likely. I can close edge, load a game on insane graphic settings, push the cpu and gpu to +70% with any regular issues.

 

[A BOM]

Not many parts remaining left other than CPU, mainboard, or trying another network path (if using wireless, try wired, or, vice versa, just to rule out some sort of weird network induced corruption?)

If mainboard were at fault, you might try tweaking up RAM voltage .05V and CPU core voltage by .05 over default, just to see if it helps. (This would never be necessary normally, but, you are facing strange symptoms, and, no one wants to buy another mainboard and/or CPU, and, whichever way you guessed, it'd probably be the other anyway...

GOod luck!

i have replaced the motherboard. I went from a z390 Pro to a z390 gaming edge ac. No effect.

the ram voltage in bios is set to auto and is indicated at 1.212. I can easily try a higher setting and advise.

Ive ran the intel stress test on cpu with good results, but as you’ve stated… there aren’t many parts left.

What part of a cpu goes bad that only affects web browsers but allows for uninterrupted intense gaming?

 

[A BOM]

Tested ram voltages with .05 incremental adjustments up to 1.3 volts with no change.

Left as setting 1.25v.

 

[Ralston18]

Reading back there are still things you can do and tools to help.

When the computer is first booted let it sit idle for awhile. Then open either Task Manager or Resource Monitor to observe system performance. Use both tools but only one at a time.

Leave the monitor window and open one of the browsers - you may catch some change taking place before the system crashes.

Considering that crashes may get in the way also consider using Process Explorer (Microsoft, free).

https://docs.microsoft.com/en-us/sysinternals/downloads/process-explorer

Repeat the browser tests and determine if some other process appears or changes in some manner. But does not do so when games are played.

Could be some other process being triggered by the opening of a browser. Task Scheduler may help discover if that is indeed the case.

Lastly, look in the hosts file:

https://www.howtogeek.com/784196/how-to-edit-the-hosts-file-on-windows-10-or-11/

Is there anything in the host file regarding the problem websites?

 

[A BOM]

First off, i want to thank both of you for the help this far. I’ve been so close to giving up on this thing numerous times.

I will give this a go and see if anything looks fruitful. If there is any information, logs, etci can gather that may be helpful, let me know. Ive been looking at this for so long that a fresh perspective may be needed.

 

[cryoburner]

16 Gig DDR4 3300 RAM (running at 2667)

Have you ever ran the RAM at it's full rated speed via its XMP profile, and is it just currently turned down for debugging the problem? I was just wondering if you might have manually configured the RAM, and have its timings set to something that the kit might not be able to handle. Memory testing utilities might not always show a problem.

Disk Drives - Currently 3 installed. Samsung Evo M.2 1Tbyte, 500+G free, OS installed here. 250G SSD, was original location of OS, currently 100G Free. HDD, 1TB, 700G Free.

Have you tried disconnecting the two other drives and only running the drive that the OS is on? And where are the applications installed, on the OS drive, or somewhere else?

 

[A BOM]

I have run the ram at 3000. I think originally it ran with the MSI extreme memory profile at 3000 and 1.35v. For the purpose of debugging i have the xmp turned off and all ram settings at default / auto with exception to voltage at 1.25 as discussed earlier. I did try xmp earlier today when making voltage adjustments with no difference in browser errors.

I have physically disconnected other drives multiple times. When OS was originally on sata ssd, it was the only dive in the machine. When the m.2 was installed with os, it was the only drive connected (was forced to do this because windows wanted to be installed on the other two drives).

 

[Ralston18]

I do not understand:

"When the m.2 was installed with os, it was the only drive connected (was forced to do this because windows wanted to be installed on the other two drives)."

Where is the source of that "windows" OS software that is trying to dictate or force the target physical drive?

I am not aware of any circumstances when/where Windows attempts to dictate the target physical drive. Windows may identify or find any present drive(s) but then seek some confirmation that the OS install go forward on the User's selected drive.

The usual reason for disconnecting other physical drives is simply prevent some end user error or other bug from corrupting other drives.

 

[A BOM]

All 3 drives were listed in discovery. When the m.2 was selected as the target for install the install would say it could not be installed on that drive.

I disconnected the other two drives physically and install proceeded normally.

 

[A BOM]

Update: Issue successfully diagnosed and resolved. I have a faulty CPU. An opportunity to test an alternate CPU in this build presented itself and with no additional changes the problem was instantly resolved.

Thank you to everyone for their time and help!