Question Cause for Memory Failure When All 4 DIMM Slots Used?

[Mr. Farknocker]

I could use some words of wisdom and advise from all of you.

I’m having RAM problems with the DIMM slots of my ASUS P8Z68-z pro/gen3 MB with I5-2500K CPU with Windows 7 Ultimate.

I purchased and installed a used ASUS MB which came with 2 8Gb Crucial Ballistix Sport sticks for a total of 16 Gb (see description below). The PC ran for a year without any problems until I started to add RAM sticks to the remaining two slots. When 4 identical RAM sticks used, the system crashes and i get a memory dump before the system restarts. What's weird is that the MB runs stable with three sticks added in any combination (i.e., two sticks in A1/B1 or A2/B2 with the third stick inserted into any one of the two open slots).

The stick that came with the MB are Crucial Ballistix Sport 8Gb; 1600 MhzDDR3 PC3-12800 (BLS8G3D1609DS1S00), 9-9-9-24 1.5v. I purchased two identical used sticks to increase my total RAM to 32 Gb. Unfortunately, the MB only runs stable with 3 sticks for total of 24 Gb.

In attempting to troubleshoot the problem I performed the following tests and found no errors:

• Memory tested each stick using Windows Memory Diagnostic Tool (MDT) and HCI Design’s MemTest in one DIMM slot;
• tested each DIMM slot by performing a MDT and MemTest one stick (using the same stick for each DIMM slot)
• I tested the RAM using a two sticks in either A1/B1 or A2/B2 configuration with MDT;
• Reset CMOS and BIOS with latest BIOS installed.
• Ran Intel Processor Diagnostic Tool

As stated, the MB runs stable with 2 sticks in either the A1/B1 or A2/B2 configuration and also runs stable with a third stick placed in either open slot. However, the moment I put the 4th stick in, the memory fails and the PC crashes by performing a system dump before restarting.

I did experience the exact same failure earlier when I attempted to add 2 sticks of 4Gb (i.e., 2x4Gb) G.Skill RIP which has the same specs as the Crucials (including latency and voltage) and thought it was a quirk and left it running with three stick until I was able to get my hands on the two Crucial sticks of 8Gb each. Unfortunately, make and capacity seem to make no difference as I have experienced the same problem as I did with the G.Skill.

I have never attempted to OC the memory. I attempted to OC the MB without success and was successful in OC the Graphics Card. The MB/RAM and GPU was not OC when I ran the aforementioned memory tests.

Since the contacts on the RAM sticks appear to be intact, the only thing I can think of is:

• One or more pins on the CPU socket is bent and the CPU clam needs to be either loosened or reset; or
• The CPU Memory Controller is defective, cannot manage the DIMM if more than 3 sticks are used at a time and cannot be repaired.

Does anyone have any suggestions?

 

[thx1138v2]

It's probably due to an incompatibility introduced by using 2 different modules purchased at different times. Since it's DDR3 I'm guessing it's sort of an old system. Vendors sometimes change their chips and/or controllers on their modules. They might have different densities. I'm guessing it is running in single channel mode when there are 3 modules installed but it tries to go to two dual channel ports when all 4 are installed and that is what is failing.

Mixing DRAM from different runs is always problematic. Try buying a four piece "kit". They should run OK together. If not, you've got a socket problem.

 

[Unolocogringo]

This topic comes up so much a member made this post to expain the problems of mixing ram.

How To - Troubleshooting problems with PC memory (RAM) and XMP profile configurations

In the beginning, there was......BIOS Before you go ANY further, go to the manufacturer product page for your specific motherboard model AND revision. Revision is an important aspect because for any given motherboard there may be more than one version of that model which will be identified...

forums.tomshardware.com

 

[Mr. Farknocker]

Just thought I'd report back as to recent developments in my quest to get all for DIMM slots working. I removed the CPU to check to see whether the problem may be caused by bent pins. To my glee, I found a rather large piece of lint on one of the pins closest to the edge of the socket. I removed the piece of lint and re-seated the cpu and powered her up again. Unfortunately, the system was still unstable and I suffered a few more BSODs. I was about to search for another MB or possibly a newer system when I decided to up the voltage just to see if it would help. I upped the voltage by .005 which resulted in lengthening the periods between BSOD's but the problem was not entirely eliminated. I upped it again by another 0.005 and that seem to have cured my problem. I've been running the PC daily for the past week for hours at a time and I have not experienced one BSOD episode yet. I plan to run several memory tests during this week to determine if there are still errors and will report back with the results.

 

[Nemesia]

So It was not enough voltage to the RAM was causing instabilities. What voltage are they now?

 

[Mr. Farknocker]

So It was not enough voltage to the RAM was causing instabilities. What voltage are they now?

That's my conclusion so far since its been running smoothly. I'll have to check the actual voltage when I get home tonight. My MB doesn't exactly allow me to up the voltage in .005 increments.

 

[Mr. Farknocker]

So It was not enough voltage to the RAM was causing instabilities. What voltage are they now?

The voltage is currently set at 1.51250 but appears to be running at 1.534. As I previously mentioned, the MB won't allow me to increase the voltage in increments of .005. I type in the number (e.g., 1.505) and the number is set to a different number after hitting the return key. I have no experience in fooling around with RAM so I'm not familiar with whether this is normal or not. Also, I have no knowledge of whether the actual voltage is supposed to be running that high. Is this normal?

 

[Mr. Farknocker]

For anyone interested, I ran HCI Memtest on my system all day yesterday (about 15 hours) while I was at work and at 1600% it showed no errors.

 

[supercellxmiku]

For anyone interested, I ran HCI Memtest on my system all day yesterday (about 15 hours) while I was at work and at 1600% it showed no errors.

Hi,
I just want to share a similar experience here, i bought an extra Ram kit to fill up the remaining 2 dimm slots. However its a 32gb kit(16x2) 3600mhz, my old kit 16gb 8x2 3600mhz and it wouldnt even post at 3000mhz at first. I did windows memory diagnostic and no issues. I tried swapping the ram slots and pc would somehow post at 3000mhz but i was aiming 3200mhz . I tried leaving the d.o.c.p as auto and speed at 3200mhz and tweak the timing to 19 22 22 22 and leave the rest as auto. IT FINALLY WOULD POST!

Here are my specs:
Ryzen 3400g
Asus x570-e
1tb mvme
corsair vengeance rgb pro SL 3600mhz 16gb and 32kit