Memtest86+ Errors on New Kingston RAM

Status
Not open for further replies.
I'm not sure if this is the right forum section as I'm new here, so apologies if I'm posting in the wrong place.

A few days ago I noticed that I had severe errors in memtest86+

I was running 3gb of unbranded ram on Asus A8N32-SLI DELUXE. RAM was 3x1GB DDR-1 DELTA (says 94v on back of them) all three same make and model.

Always had problems with BSOD just that this was the first I've had time to read about the cause and actually investigate it.

I tested all three sticks of ram, individually using memtest86+ and testing them in different memory slots. ALL OF THEM reported errors on test 6, no matter what slot they were in. We're talking 1200 errors minimum.

Well, I thought it must be due to cheap RAM, ordered Kingston which arrived this morning, the exact same RAM model number as recommended on the mobo manual ("KVR400X64C3A/1G).

It's not good news though! I plugged in the two new sticks to the two brown slots and booted up in Memtest86 v.211 mode.

As usual, Test #6 returns 800 or so red errors and test #7 returns another 400 errors... I usually switch off before test #8 but I've noticed it's stuck at test #7 Test 3% Pass 64% and wont budge further.

My BIOS settings are defaulted as well.

Does anyone here have any idea what can be causing this? My PC crashes sometimes 15 times a day :-(

Been looking in BIOS and can't see how or what to set the RAM voltages too etc as I an no expert in BIOS settings. Any help would be hugely appreciated

SCREENSHOT MEMTEST:
www dot limelites.co.uk/memtest.jpg

SCREENSHOT BIOS MEMORY SETTINGS:
www dot limelites.co.uk/memory_settings_bios.jpg

RAM DATASHEET:
www dot valueram.com/datasheets/KVR400X64C3A_1G.pdf

As you can see from the datasheet, it doesn't say what settings I need to use for things like "MemClock Mode" "Memclock Value" "MCT Training Mode" "CAS Latency (CL)" "TRAS" "TRP" "TRCD" "TRRD" "TRC" "TRFC" "TRWT" "MCT Extra Timing Mode" "TREF" "TWCL" "R/W Queue Bypass Count" "Bypass Max" "Idle Cycle Limit" "Dynamic Idle Cycle Centre" "DDR Driving Strength" "User Config Mode" "Read Preamble" "Asyc Latency" "Burst Length" "Hardware Memory Hole"

The only one of these mentioned in datasheet is TRFC and datasheet says 70ns (min) whereas BIOS doesn't have this setting, it has 9T to 23T

Help me please??
 
The only few things that I'm aware of that could potentially mimic faulty RAM is defective CPU cache, motherboard failure/defective DIMM slots and maybe somehow dirty power from a cheap PSU.

I think the first course of action would be to run some tests on your RAM in an unrelated system that is known to be 100% working. If it comes up clean, you can begin investigating a potential hardware failure in your system beyond the RAM.
 
It can't be the RAM as I just removed three sticks, thinking it was the RAM and replaced it with two brand new Kingston sticks... Memtest shows the exact same errors... No way it can be all five sticks of RAM.

I'm thinking it must be something to do with BIOS settings?
 
If you're that certain it isn't the RAM, as I said, it's almost definitely related to your CPU cache, motherboard or PSU.

The least likely of the three is your CPU cache, and so you might want to consider either swapping out your motherboard or PSU and seeing how the dust settles. It shouldn't be too hard to get your hands on an old system with a compatible motherboard for your CPU/RAM, and the same goes for a PSU.

Eliminating the CPU cache is as easy as disabling it in the BIOS (assuming you have the feature), and then running Memtest again.
 
Pretty sure it's not the PSU and I'll certainly test the CPU cache next time the system restarts..... Is there another way to 'test' a mobo without having to replace it ?
 
PSU failure is as a whole is much more common than motherboard failure - but I'm not sure which would be more likely in this particular situation.

As far as testing your motherboard, I can't think of an easier method that will be equally as accurate in results as just swapping it out. Maybe someone else will chime in with something...
 
Don't know how PSU failure could produce memtest errors, so I'd eliminate that. My bets still hinge with incorrect BIOS settings. I've been told that I have an "enthusiast" motherboard (A8N32-SLI DELUXE) which means, "it requires more tweaking to get it to run both fast and stable"..... Therein lies my problem, I'm a website designer not a technician.... Not clued up at all on BIOS settings and such like :-(

I'm really trying to find out the best settings in BIOS to match my RAM but can't seem to figure it out at all.

There are so many settings in BIOS that I simply don't understand. I've tried messing with them but choosing the wrong settings results in a system that wont even boot and I have to remove the CMOS battery to reset BIOS in order for it to boot at all. I don't like messing with the BIOS settings as I don't know what I'm doing in there.... What settings would you change first off to try and gain stability through compatible mobo/RAM/BIOS settings?
 
Regardless of whether or not your motherboard is an "enthusiast" motherboard and features a slew of settings for hardware manipulation, it would likely only provide the ability to modify your RAM's frequency and voltage settings; however, the only time you really have to play with that is when you buy RAM that can support higher frequencies and is marketed as such - but is sold with SPD data that runs at a decreased clock and/or voltage, or if you just flat out want to overclock.

Now, granted that may or may not be an option, you shouldn't have to modify any settings to run your RAM without errors being reported by Memtest. I would set your BIOS to stock/default settings and take things from there, but if you're so adamant about it being BIOS settings affecting your RAM - take a look at the RAM manufacturer's documentation.

RAM, as with any other component, requires clean power to operate in an ideally functional state. If your PSU cannot provide clean power, then, yes, I would almost certainly guess that a crap-box PSU could mimic errors in a memory test, just as it can cause data corruption on a HDD. In a debate about how likely your PSU is to be causing your issues, I would argue not very high - but being easier to swap out than a motherboard, it's not a bad idea to spend 15 minutes trying it.
 
Status
Not open for further replies.
Back