Blue screen occasionally after adding ram.

dj41490

Honorable
Feb 11, 2013
18
0
10,510
Hi, I had 2 x 4gb 1600mhz, cl9 (9-9-9-27), 1.65V, DDR3 hyperx kingston rams, model number KHX16C9B1BK2/8 https://www.kingston.com/datasheets/KHX16C9B1BK2_8.pdf. http://www.pcgameware.co.uk/reviews/memory-ram/kingston-hyperx-black-8gb-1600mhz-memory-review/. I recently bought 2 more used rams which are 2 x 4gb 1600mhz, cl9 (9-9-9-24), 1.65V, DDR3 hyperx kingston limited edition rams, model number KHX1600C9D3X1K2/8GX http://www.hardwarecanucks.com/news/memory/kingston-launching-limited-edition-ddr31600-hyperx-memory/. I put them in like 2 dual channels, manually set one channel to 9-9-9-24, 1.65V, and the other 9-9-9-27, 1.65v. Tried setting both channel to 9-9-9-27 as well. Ran memtest86 once the entire night and passed with 0 errors. However, from time to time I get a blue screen for hardware failure, and I'm not quite sure what the problem is. Did I make any mistakes? Any suggestions?

Specs of my computer:
CPU: i5-4670k overclocked to 4.2 Ghz
MOBO: Gigabyte GA-Z87X-D3H-CF (supports up to 32gb ram)
OS: Win 7 Ultimate
GPU: R9 290
SSD: Crucial M500 240GB
16gb DDR3 Ram
 
Since memtest didn't find anything the RAM seems fine. However you could try testing the RAM sticks in various RAM slots. Also I think its best to set all the sticks at the same latency timings. As a test try to relax the latency timings a bit more. Those settings are only for 2 stick operation not 4 and since you can't add more RAM voltage, you have to relax the timings.

I also have some notes about your system.

1)The overclock may not be stable. If you did it before you added the extra RAM you need to start from the beginning (stress tests, tweaking the various settings). The moment you added that RAM your systems configuration changed. The CPU RAM controller gets stressed more and you have to play with various system voltages.

2)That RAM isn't a good match for your system. 1.65V is too much for a Haswell memory controller, 1.6 is the official limit, beyond that you are risking permanent CPU hardware damage. As you can see you already have crossed that limit. Some motherboard even add more voltage automatically for stability purposes, thus the actual RAM voltage may be more than 1.65V. So those 4 RAM sticks may be too much for the CPU memory controller. You should have gotten lower voltage memory for that system. Perhaps you should lower the voltage and configure the RAM at a lower clockspeed. 1.5-1.55v and 1333mhz may be more stable.
 
Ram is sold in kits for a reason.
A motherboard must manage all the ram using the same specs of voltage, cas and speed.
The internal workings are designed for the capacity of the kit.
Ram from the same vendor and part number can be made up of differing manufacturing components over time.
Some motherboards, particularly AMD can be very sensitive to this.
This is more difficult when more sticks are involved.
That is why ram vendors will NOT support ram that is not bought in one kit.
It is safer to get what you need in one kit.
But, it is a bit more expensive because of the added matching of all the sticks to insure compatibility.

That said, running memtest86 overnight with zero errors would indicate that you lucked out
You might try increasing the ram voltage in the bios a bit.
Sometimes that works.
 


Good thought but I'm afraid that it would be too dangerous to add more voltage to the haswell memory controller. He is already at 1.65V (the actual voltage may be higher). He has already exceeded the limits of his CPU and he may already have damaged the CPU memory controller.
 

dj41490

Honorable
Feb 11, 2013
18
0
10,510
Guys sorry for the late response...... I think I figured out the problem. I had another blue screen today and the error was DRIVER_IRQL_NOT_LESS_OR_EQUAL related to a system file named atikmdag.sys. Did a bit of google searching and found out it was related to my ATI drivers. I reinstalled my AMD to the latest version and haven't seen any PSOD so far......