Hello guys,
After days of troubleshooting, even though memtest and testmem5 tests come up with no errors, a lot of forums think there is an issue with my RAM, which can be related to my chipset drivers.
I updated all of the drivers to their latest, but I still get BSODs randomly.
Here is my rig:
Ryzen 5 3600
ASUS TUF B450M Plus Gaming
Corsair Vengeance LPX 16GB (2x8GB) DDR4 DRAM 3200MHz C16
EVGA RTX 2070 SuperAntec EA750G Pro 750W Gold
Win 10 Pro 2004
The RAM is listed under the QVL list from AMD, but it still seems to have issues when I ran it at 3200mhz.
So basically, I tested stock/default settings, tested DOCP only, tested Auto with tweaks, just auto and a bunch of other voltages and clocks, and it still seems unstable. I'm not sure if it's somehow related to the latest chipset drivers, but I'd sure love to check with AMD themselves.
Please help me resolve my BSODs.
The problem you are experiencing has nothing to do with the chipset drivers. If your RAMs are working properly at 3000Mhz but causing problems at 3200Mhz, there is a quality problem with the RAMs. For 3200Mhz, you need to manual RAM stabilization in BIOS memory timing settings. Replace the first two memory timing values to 18 from 16.
Make sure you put the Dual RAMs in the correct DIMMs. Select the XMP 3200Mhz profile with the BIOS. If you're having trouble, try to set the memory timing settings.
Thanks, but the behavior of my RAM is inconsistent.
When I activate DOCP and run it as is, it works fine but I got a BSOD at some point. Today, after using stock settings (2133mhz, no oc) I got a BSOD again but all memory test don't show any errors in the memory.
I tried going back to DOCP, no luck, now the PC won't even boot into Windows and when it does, the memory is at 2133mhz and not 3200, even though it is set to 3200 and the docp profile is activated in BIOS.
Zip up the latest minidump and upload it so we can have a look at the error.
C:\Windows\Minidump
You have to get the BSOD file from this path. And share it if you want.
Or share the BSOD error messages you got.
That is the file from the path you mentioned.
Yeah the files from the folder mstfbsrn980 listed.
The full memory.dmp file is too big to download.
If you are going to OC your processor or RAMs, you must increase the core voltage for the CPU or change the timing settings for the RAMs. You may need to change the CPU voltage value and the memory timing settings of the OC system over time. Because with OC, system parts age faster and lose their old consistency.
Your RAMs may be corrupted. Or the parts that provide stable voltage to your processor may not be applying the old stable voltage value. Before making OC to the processor, you have to read core voltage with a test program and make OC by going above the value you read. In other words, there is no OC without making any changes, only by increasing the processor or RAM speed. In addition, OC potential may vary according to the part number of each processor and RAM and the quality of your product according to the standard.
Configure in manual mode in the bios memory with these values
1st test the XMP profile but forcing the RAM voltage to 1.35v
If this fails try these values
Cas latency (CL) 16
RASto CAS 18 (TRCD)
Ras Preload (TRP) 18
Cycle time (After) 36
# Bank cycle time (trc) 74 #
Comand rate (CR) 1T
The value marked with # is the only one that changes and is the one that matters the most.
I have the same memories and with these values they work perfectly using the 2 options, voltage and values.
Thanks, I'm in the middle of a long process of troubleshooting all of my hardware so I will use those values once I'm "allowed" to OC again, it's stock for now.