• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

LN2 Issues

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

Dawgdoc

Member
Joined
Feb 20, 2007
Setup as follows:

Foxconn Mars mobo
Q6600 G0
2 X 1 GB of a variety of ram interchanged (Mushkin PC8500, Mushkin PC8500 extreme performance, Corsair XMS2 PC6400)
8400 GS (8800GTX is drying from silicone spray and Im only benching CPU tonight)
Corsair HX620W

Turn the rig on, POSTs fine (dry pot at the moment).
Go into BIOS and check CPU temp. All looks good. Hovering about 30C.
Dump some LN2 into the pot for the first time!! Temps move alot slower with the LN2 to start I guess you gotta load the pot? Im watching CPU temp in BIOS and wow.....it reading negative! First time Ive seen that....-30, -35, -40....I stop watching.

Set my volts/fsb/yada yada.....F10 out...no POST :(

Clear CMOS with the nifty little button on the mobo. No POST :(

Clear CMOS manually....POST ok...Good :)

Over the next 30-40 mins, I have a variety of issues with no POSTING and having all kinds of issues. I think its my KVM switch so I pull it. Things seem to go well for a lil while. Next.....something aint right with my vid card? (I dunno)...I switch out for a Radeon 1650 pro. Once again things go well for a little while...then no POST no clear CMOS issues again.

I put back the 8400GS instead of the Radeon...POST fine initially but then same problems...

As of right now, the latest issue which is preventing me from doing ANYTHING is as follows.....

POST ok, as the POST message starts it hangs on the words "Memory Test". Ive tried swapping to all different memory sticks and same thing happens.

Any clues on whats going on? Is this what a cold bug is? It seemed to be the same no matter if I am at -190C or -90C.....Did I freeze my DIMM slots or something?

Letting everything warm up since I have no other choice right now, and will try again in a little while but....Any clues?

BTW, after that first boot my BIOS would NOT read negative temps anymore. Stopped at 0C. Weird huh?

Now that I think about it...I assume I am done for the night? During the hour I was working on it a good amount of frost built up over the CPU pot and started on the resistors on the mobo too. When that melts....Im in trouble eh?

Sounds like Im done for the night :(

Im completely baffled.

BTW, board was working flawlessly (everything was working well) 2 nights ago when I last benched it on dice with a different CPU (X3220 G0)
 
Last edited:
Someone on another forum said it sounds like I have a wet socket.

Agree/Disagree?

Other forums 'folks' also said I should DEFINITELY be using dielectric grease IN the socket and basically to slather the CPU with it pinside.

Agree/Disagree?

Here is my current socket insulation setup:

1. Nail Polish all around the socket back and front

2. Air duct insulation back and front (grey stuff from Kingpins posts on insulating a mobo from XS)

3. Rubber gasket on bottom of board and then backplate.

4. Top of board Paper towel square cutout over socket, and then a square cutout of armaflex over socket.

5. CPU pot (Dragon EVO 2.4) on top of that with the mounting kit, armaflex tubular insulation above and below the mounting bracket, and paper towels around the whole CPU unit.

Here is a pic of what I was told I should do with the dielectric grease

 
More info.....

At the last issue of hanging on the POST message......


When I cleared CMOS I would apparantly POST fine and can get into BIOS. After resetting BIOS functions is when it would hang on the "Memory Test" line of the POST message.

I have no idea what was happening during the POST when I cleared CMOS,since part of clearing CMOS autosets the full screen logo instead of the POST message. I would spam delete and it would go right into BIOS.

Hope that bit of info helps someone to provide more detailed feedback? I dunno.....Im gonna let it thaw a few more mins and then Im gonna break it down and check the socket and board and make sure everything looks dry.
 
It seems like you're having some mixture of cold bug and condensation issues.

Are you probing temps at the base of the container? Ignore BIOS temps entirely.

The frost could've been caused either by inadequate motherboard insulation (I don't mean the socket, and skip gooping it) or simply by you spending too much time at low temps with the rig powered down. When you've got a container well in the -100 range like it sounds like you did, leaving the rig off leaves the container unloaded, and none of the components generating heat on the motherboard, which will lead to it frosting up.

Powering down on DI and LN2 is in general a no-no. You want to always at least be in BIOS, and if you can't post, leave the system hung at a black screen until you can.

I bet your mobo and CPU are fine though.

Start up with an empty container, pull down to around -80 base temps in BIOS, and after that, enter Windows to ensure that the container is being loaded well. Pull temps down further in Windows and ramp speeds up using clockgen or setfsb.

You should never see any frost on the motherboard, and especially not with the Evo, since its design is quite focused on the CPU, and does not have a wide base like other containers.
 
TY Gautam :)

I pulled the Evo off. Socket is as dry as could possibly be.

So what do you think this problem is then?

Do q6600s coldbug? I was at -190C at 1 point. Had the same probs at -190C and -90C alike.

I dont even know what a coldbug is, other than a cpu that bugs out in some way, shape or form at cold temps :)

When I was attempting to bench, I would notice that when I poured LN2 into the pot, I found that it would turn gasseous and pour N2 smoke all over the board. Really no way I could stop it. It was WAY more than with dice. With dice I would get some smoke if the pot got up to -10C or so and I added ice, or if I ran low on acetone and added more i would smoke, or when I first started up....Other than those 3 things NO smoke (and I mean NO smoke) for an 6-8 hour session. 45 mins of LN2......TONS of smoke...constant smoke.

Temp prob is at the way way way bottom of the pot. Vince drilled it for me so I have to imagine its in the ideal/perfect spot :)

Regarding BIOS temps yea....I always ignore em....but I thought it was weird that it was reading a subzero temp. -40c actually!!! then after that first reboot it would report zero only.
 
Last edited:
So, is it possible that I OVERcooled the pot and the CPU because I never really got a load on it.

I booted into windows ONCE in 45+ minutes of subzero. Fair amount of BIOS, always had near maxed out everything volts in BIOS, but POSTed so infrequently I was prob mostly in stock volts because as soon as I would F10, I wouldnt POST.....then Id power down......clear CMOS...reboot.......get into BIOS under said cleared CMOS...change settings......F10....no POST....etc....repeat again.....

I dont think I EVER really applied any loads to speak of, but I COOLED THE LIVIN DAYLIGHTS OUTTA THAT POT!!!

Could that have been my problem? Overcooling and no load?
 
Sorry for the chain posting but....

When I say it hung on the Memory Test line...I mean it would stop mid-word

Me (HANG)

Mem (HANG)

Memor (HANG)

etc........never got past the word Memory
 
Sorry for the chain posting but....

When I say it hung on the Memory Test line...I mean it would stop mid-word

Me (HANG)

Mem (HANG)

Memor (HANG)

etc........never got past the word Memory
Hmm...must be bad memory....send it to me!!! :D

Honestly though, it sounds like it coldbugged. If it worked the first time (30c) and not the others...
 
I figured the socket would be dry, but you said the board's resistors weren't. -190 is way too cold. There's no way the rig would POST that cold. Typical cold bugs are between -120 to -90 on G0's. As I said, you shouldn't start really going cold until you're in Windows.

Also you're not seeing nitrogen gas. You're seeing very cold water vapor. That usually only happens when your container isn't fully pulled down.

I dont think I EVER really applied any loads to speak of, but I COOLED THE LIVIN DAYLIGHTS OUTTA THAT POT!!!
This is exactly what you shouldn't do.
 
I see. Well no screenies tonight but a learning experience.

When you say "pulling down" the pot do you mean applying a load to it?

Here are images of how I insulate my boards. Tell me if its ok.

First nail polish top and bottom all around socket area.

Then grey foam duct insulation, then paper towel, then armaflex square, then pot just as these pics show.

Ok or not ok?





 
ROFLMAO....-190...you my friend are the best, I needed a pick me up :)

Both the Q6600 "GO's" I have used coldbugged (stopped working) around -130 to -125

Don't tell anyone, but I did the -190 as well the first time out
 
The insulation looks good, actually probably overkill. If you had anything frosting through insulation like that, it means you were definitely running way too cold for having no load.
 
Resistors....yes....

At the end of the session (45 mins) the entire plastic pot "cap" that gets bolted down was frosted 100%. styrofoam half-cup at the top of pot frosted on bottom half, AND....I think....cant say for sure that I did but I think I saw small frost deposits on the resistors around the socket that are outside of the grey foam area on my pic.

The resistors between the socket and memory to be specific.

So yes, I think there was a bit of frost.

Also, is that smoke a problem? Is it excessive? I thought it was gasseous N2 but obviously what do I know hehe :)

Corey gimme your addy Ill mail the RAM over ASAP! :D
 
Last edited:
I see.

I understand now. First session was a learning one on the properties of LN2 and how to effectively use it, as opposed to a benching session. Its all good though. Always another night and I still have like umm.....75L of LN2 left Id bet :D

Dean, glad I could make your night. Come back again tomorrow the show is gonna continue. Maybe Ill blow another cap tomorrow for ya hehehe :bang head

In retrospect, and after speaking to everyone more experienced, I was WAY WAY overzealous with the LN2. It was slow to get to -80C, perhaps 5 full mins as opposed to DICE which went to -70C within 30 secs.....

I never applied a load. I was trying to boot right from the get go at 520 FSB, and my system wasnt hearing any of it. 520 on dice was a no brainer. I dang near hit 550...but I was at -150 before I ever applied anything other than stock volts and by then I assume, I was coldbugged. It wouldnt clear CMOS right from the get-go from the button on the mobo, which I used like....50,000 times per session on dice :)

Ok....tomorrow is another night...and maybe my 8800 will be dry by then too :)

Thanks again ppl.
 
Last edited:
Excellent work tonight :)

How much ln2 did you pick up if you have 75L remaining??:eek:

Dunno, my silly gas supply company sells it to me by the pound LOL.

1 dewar has 50lbs of LN2 (not counting dewar), and the other has 35 lbs. They were both supposed to be 35L dewars but since 1 holds more LN2 weight Im assuming 1 is bigger. They both appear the same size externally, but have no clue on the actual internal dimensions.

I used about 1 and 1/2 thermos-fulls tonight. Big thermos yes, but barely a scratch in the total volume I have.

Im benching a CPU and a GPU, and I had planned on doing it tonight AND tomorrow, and then re-evulating if I needed to pick up more before the weekend or if I was ok :)
 
Hmmm.... That would mean I have 48 liters.

I dont doubt your 1.7 number, but I think they probobly misweighed.

I cant possibly have 48 liters. These are 2 giant-sized dewars that are waist high on me and Im 6t tall.

Hmmmm I may have to perform a fill test after I empty them out.
 
Last edited:
Back