RAM gets too hot to touch, will this damage it?

Just Another Guy

Distinguished
Mar 10, 2012
123
0
18,690
Hi, just tried running memtest and the RAM got to the stage where you can only touch it for half a second before you're forced to take your finger off (think microwaving food and not physically being able to pick the bowl up after)

The RAM isn't overclocked, but it's triple channel running at 1600mhz and I have all 6 slots used

I was running around 6 tests at 2gb usage each, and after 10 - 15 minutes or so, each one got an error within 20 seconds, making me believe it's overheating.

Based on what I've said, do you think this would most likely be the problem? And I haven't just permanently damaged each stick right? :p
 
Solution
One easy way to check if heat is the issue is to run whatever benchmark tests you are using with the side panel off. Then if the error still occurs, heat may well not be the cause. There are two factors here when dealing with heat. First is to cool the component that is getting hot and second is to remove the heat from the case. An example is many of today's video cards have great cooling but discharge the heat inside the case. They depend on good case airflow to remove the heated air. If you need to cool the memory a bit more here is a handy item (http://www.amazon.com/Antec-Spot-Cool-SpotCool-System/dp/B000I5KSNQ). As to the voltage, Corsair memory should not have an issue with it but it will run cooler with even a small reduction in...

Dogsnake

Distinguished
I doubt you damaged the memory. The real question is why is it running so hot. Since you gave no specs or model information I suggest you got to the bios and set the RAM voltage by hand. The board may be upping the voltage causing the heat. Get http://www.cpuid.com/softwares/cpu-z.html and verify the speed of the memory. Are you O.C. the cpu? Please post your component specs and models if you need further hep.
 
If you can only touch the memory for about 0.5 of a second that would imply a temperature of around 50 to 60 degrees Celsius. This is within the operating temperature limits. However you should check the BIOS to see the memory voltage is correct.
 

Just Another Guy

Distinguished
Mar 10, 2012
123
0
18,690


So... the mother board is an ASUS one (P6X58D-E), and these are the values CPU-Z gives about the ram:
DDR3, triple channel
NB Frequency - 3200mhz
DRAM Frequency - 800mhz
FSB: 2:8
Timing: 9-9-9-28

The cpu was overclocked by the shop we got it from so runs very stable, it's an i7 930 (2.4ghz) running at 4ghz
And I doubt it'd make a difference, but just in case, the graphics card is a GTX 680, also slightly overclocked

Cheers :)
 

^+1
This is exactly right. You should also be aware that motherboard manufacturers usually recommend additional cooling for ram when all slots are occupied (a ram fan - two in your case)
 

Just Another Guy

Distinguished
Mar 10, 2012
123
0
18,690
The CPU usually runs between 45-55c depending on the weather, which I know isn't great, but it doesn't go past 95c if I render something out so it'll do :p



Ah right thanks, didn't realise they actually suggested it
My case came with two side fans, but we removed them due to not fitting with the watercooling fan. I've just taped them to various other points in the case where they'll actually fit, so hopefully I've got better airflow now haha :)

Anyways, glad to know the ram will only temporarily stop functioning correctly if it gets too hot (which has only ever been during multiple memtests so it should be fine normally), cheers for the help guys :)
 

Dogsnake

Distinguished
I have the P6X58d with a 920 at 4ghz and run all 6 slots populated at the same speed (12gig). Check the bios to be sure the ram voltage is set to 1.6-1.65v. This board needs to have the voltage set manually to the correct value. If it is set to auto it may well be over volting the ram. My cpu idles 38-41 depending on time of year and room air temp. Under full load it get up to 78-80 (all in C). Air cooled btw (Noctua NH14).
 

Dogsnake

Distinguished
The chipset for this board and the tri chn. memory it uses specs. 1.6v. You are referring to later series requirements for different hardware. In any case you should set the voltage to what the manufacturer recommends. This what I have 2 sets of in my 920 system on the same MB (http://www.newegg.com/Product/Product.aspx?Item=N82E16820145242). I use it as an example. What is the brand and model of the memory we are talking about here. It is helpful when the questioner provides specs. in the request for help.
 

Tradesman1

Legenda in Aeternum
Without the DRAM specs, as Dogsnake mentioned, it's hard to look at the problem, which is why I through out the IF IN FACT the sticks are spec'ed at 1.5, then 1.65 is higher than I'd want to run them - you are overvolting the sticks. Most DRAM manufacturers usually will suggest no more than 0.10 over spec without additional/extra system cooling above and beyond a pull fan in front and a push fan in the back of the case
 

Just Another Guy

Distinguished
Mar 10, 2012
123
0
18,690


So yeah, only just realised after my last two attempts earlier this week that it'll not actually post a reply if I just type a message in normally, apologies about the long silence :p

They're similar to yours, corsair xms3 ones, although I seem to remember a ddr3 value in the bios (on the same page as where the cpu overclocking was done), which was set to 1.66v, so I think they're more or less running at the recommended voltage.

Sorry about not posting specs, I didn't realise people were actually so helpful on here lol, I expected I'd just get a yes or no answer ;D


Anyways, I had to remove some side fans when I got the case due to not fitting with the cooling, although I've now decided to just tape them on the side of the case where they'll fit, so the airflow should be a little better :)
 

Just Another Guy

Distinguished
Mar 10, 2012
123
0
18,690


Would 0.01v really make that much of a difference? :p
and it's this case, i thought it looked good, just was a bit disappointed to discover that it was impossible to fit the side fans on haha
http://www.cclonline.com/product/55382/8818B/Cases/CiT-Vantage-Gaming-Case/CAS0231/
 

Dogsnake

Distinguished
One easy way to check if heat is the issue is to run whatever benchmark tests you are using with the side panel off. Then if the error still occurs, heat may well not be the cause. There are two factors here when dealing with heat. First is to cool the component that is getting hot and second is to remove the heat from the case. An example is many of today's video cards have great cooling but discharge the heat inside the case. They depend on good case airflow to remove the heated air. If you need to cool the memory a bit more here is a handy item (http://www.amazon.com/Antec-Spot-Cool-SpotCool-System/dp/B000I5KSNQ). As to the voltage, Corsair memory should not have an issue with it but it will run cooler with even a small reduction in voltage. Whoever built your system may have found that they needed to set that voltage to get a stable system at the O.C. you have. The memory controller for your 930 is tied to the base clock of the cpu (blk). In order to get a cpu bandwidth of 4ghz the blk needs to be at 200 and the multiplier set to 20. This will also yield a memory speed of 1600. There is enough cpu sample variation to require voltage tweaking to get stable systems. I would try lowering the memory voltage and see if things get any cooler. Your memory should pass memtest. If you need to go further what you will do is remove the 3 memory modules from the black slots leaving the blue in place. Run the test. If all passes then remove from the blue and replace with those that were in the black. Retest. If there is a fail, then start swapping one by one with a known good stick from the first test. You will then be able to find if one of the modules is bad. Before you do to this open the side of the case and if you have a house fan of any kind aim it into the case blowing air. Run your tests, if no errors then heat may well be the issue. I assume you have had this system for a few years. The bottom line is does it run OK? Do you get blue screens often. If it is not broken do not try to fix it :)
 
Solution

Just Another Guy

Distinguished
Mar 10, 2012
123
0
18,690
Thanks, it runs fine with a single memtest (easily over 100% coverage), but 8 at once throw errors relatively fast which is why I think it's probably due to heat, I might give the 3 sticks idea a shot sometime soon, we have a decent fan I use for drumming so that should be sufficient to see if it's heat or an actual problem :)

And I've had this nearly 3 years now, the old case which had an almost nonexistant airflow did cause a few blue screen issues when rendering (cpu would hit 100c then shut down), but with the new case, and especially with the watercooling, don't think i've had any blue screens since :)
The one confusing thing is why after the errors occured, firefox stopped working correctly, but windows remained perfectly fine :p

Anyways sorry, don't use this site much lol, totally forgot about picking solutions so I'll do yours, cheers to everyone else for the help though too :D