Insulating for Cascade vs. SS Phase?

Hi Witchdoctor!

That looks identical to what I do for ss. I'll try to fill any gaps around the cpu then add the thin layer of eraser across the top of the holdown.

Still not entirely clear on the "Use electrical tape or even masking tape for sealing the CPU to the plastic socket" guidance, though.

You know the goal (seal the socket) but have to make it happen. That can be accomplished with a tic tac toe board patter with the center square touching the edges of the IHS sealing the socket it sits in.

A picture is worth a thousands words, no doubt... but there is no chance you will be my pictionary partner...

Yep...know the goal, will see if I can make it happen (would stilll love to see a pic, though).

Yeah...am an excellent partner for scrabble, trivial pursuit, crosswords, cards, etc but am always the last one picked for pictionary hehe

Here's a before and after pic. I used 3/4" gray electrical tape, just pull off the length you need and cut it then cut that piece down the middle and the width should be good.

Nice! Thanks MetalRacer!

For some reason, thought any tape was fine? Used painters tape as it was sitting on the desk then filled all the gaps with eraser. Then covered all the exposed metal with eraser (i slipped the holdown lever in between the 2 layers of armaflex).

Think this setup's good to go (still cutting gaskets so those will come shortly)?

Update: Unless ya'll feel it's good to go, am going to use electrical tape instead

I've used painters tape before without any problems.

Whew. Cool. Thanks MetalRacer!

So we good to go?

Runway 11 , cleared for take off ......


Nice board MR, love the OCF line of boards, I rock it's little brother

trying to get my hands on a z97OCF

egg was sold out last time I looked

Excellente! Thanks Witchdoctor!

Wa-a-a-y behind where I'd hoped to be at this point. Ran into trouble getting the clamshell installed and still have 10 neoprene gaskets to cut (plus tape/eraser on the 2 other mobos). Also still no luck on getting the - button on one of the vgas unstuck :/

Thank god benching is so much darn fun otherwise it'd just be kind of pitaliscious

Like those ocf's, too. Gotta send my z77 in for repair but it's a low priority atm.

Don't sweat the details bro,

What ever is not ready we can finish on the 19th

No rush, we bench what we bench, it is all about the fun

figure we will only have 230 liters to burn

Woohoo!!!

Testing's proceeding! First test (just bare, no mobo/cpu) went well...stable -114C.

Finished wrapping the clamshell (used armaflex) and taping it up. Ditched the stack o' gaskets idea (for now anyway) and made my first 'real' test last night.

Ran the mve with a celery (wanted to use the cheaper cpu jic). Idle temp was -98C (did not load test yet). Basically just let it run for about an hour (I also confirmed my ide, quad-boot is working properly). Everything seemed great...no condensation or anything but the heatsinks around the cpu got REALLY cold (do I need to do something about that?).

Anyhoo...totally psyched! Next up is the rive/3930k (including load testing), then mve/2500k. If I have time before heading off for vacation (and the Witchdoctor's the following weekend), figure I'll load/test the REX and try to bin out my q6600's so I can determine which one's the speeder (one of them passed wp1024 in 269.64 @ 4.8GHz with LN2).

Hard core bro, looks like you are in business ........

rive/3930k testing was not entirely delicious yesterday

Memory was whacky from the first boot (so I immediately cringed...possible bent pin). Showed both sticks in bios (and usually in real bench) but not in cpu-z. Messed around with it for a few hours....all the sticks are fine but I was really feeling the bent pin thing.

Cascade sits at -98 to -99C at idle. Ran some load tests (p95 with cpu @5.3GHZ, 1.55v)...temp hit it's max load temp of -93C in about 20 minutes. Gave real bench a go at 5.2GHz (without any real tweaking to ram, etc)...got an opencl error but final score of 103K so that wasn't too horrible.

After about 6 hours the mobo started acting real squirrely...all kinds of random freezes and eventually no posty....figured something's wet so shut everything down. When I freed the board, there were small drops/puddles of water (actually...it was pretty much soaked) around 1 side of the cpu area and I could see where there had been a small gap between the insulating gasket and evap 'collar'. Pulled up the insulation/eraser, got the cpu out and yep...a couple bent pins

Moving back to the mve but now with the 2500k. Changed up how I do the eraser/insulation (eraser first then the 2 layers of armaflex...will post pics later) so that should be better. Going to see if I can fix the bent pins on the rive (*fingers crossed*) today so I can give it another try on Monday or Tuesday evening.

Ah well...definitely a learning experience! Will keep trying :/

Working it out ....

the socket 2011 is an insufferable pain in the *** when it comes to seating of chips. When seat them them generally I like to push down on the heat spreader while locking the mechanisms in place, otherwise I fear that the chip is going to lift due to pressure on the edges in strange points.

I wish they would get rid of the LGA assembley and go to something like A + B upper and lower pin set. course i have multiple reasons for likeing this idea besides just the surface area for pins.

I tend to allways have problems with condensation around the outsides of the evap head, even when you think you have done a good enough job of insultating, a long run time is still gonna have some exposure of a cold surface to the air.

I have allways wondered if it would be possible to run an upside down setup for this. Have so that what drips off will not end up on the board. But I am not sure how that will work with the evaps heads and then you still have the back of the processor socket to worry about.

Has anyone tried the neverwet spray?
I have some sitting here. Maybe I will spray some on and see what happens.

Not yet, haven't seen anybody else try it either.
I'd be interested in seeing if you can make electrical contact through it. If you can, that changes everything.





Also:


You're aiming for an airtight seal if at all possible.

Thanks everyone!

Yesterday with the rive/2500k did not go well but remembered that the cpu will not run if colder than -70C. Let the cascade run for about an hour anyway...still getting a little condensation around the heatsink area.

Got the pins straightened out on the mve so started re-doing that setup. Added more eraser inside the cpu holdown. Didn't redo the tape per Bobnova *groan* (did it before you posted) but really sealed all around the cpu with eraser (inside the holdown) then erasered like normal once the holdown was shut.

TsunamiJuan said:

When seat them them generally I like to push down on the heat spreader while locking the mechanisms in place, otherwise I fear that the chip is going to lift due to pressure on the edges in strange points.

I tend to allways have problems with condensation around the outsides of the evap head, even when you think you have done a good enough job of insulating, a long run time is still gonna have some exposure of a cold surface to the air.

I have allways wondered if it would be possible to run an upside down setup for this. Have so that what drips off will not end up on the board. But I am not sure how that will work with the evaps heads and then you still have the back of the processor socket to worry about.

Click to expand...

Yeah...I've learned to hold the cpu down from prior bad experiences :/

Going to try to add some more insulation around the evap head/clamshell...may go back to my original stacked gaskets idea. Will also cover from the holdown to as far as possible with shop towel.

Upside down....hmmmm

xsuperbgx said:

Has anyone tried the neverwet spray?
I have some sitting here. Maybe I will spray some on and see what happens.

Click to expand...

Not sure what I'd do with it...board's already conformaled.

Bobnova said:

Not yet, haven't seen anybody else try it either.
I'd be interested in seeing if you can make electrical contact through it. If you can, that changes everything.

Also:
View attachment 145752

You're aiming for an airtight seal if at all possible.

Click to expand...

Worked on the setup before your post. Hoping the eraser inside the holdown did the trick but we'll see

Think I maybe, possibly, perhaps got it now. :/

Gave the mve/3930k another shot last night. Got all the pins straightened and the board/cpu seem to be working properly. woohoo!!

Added insulation at the end of the evap line (wrapped armaflex around it) and went with the stacked gaskets idea. Pics are up above but cut 4 gaskets of 1/2" heavy neoprene. Since the clamshell is exactly 2", I added roughly 1/3" of rubber gaskets so that the entire structure compresses down onto the mobo (hopefully sealing off any air).

Didn't have much time but ran the setup for about 2 hours. Temps on the surface of the mobo (and particularly the heatsinks around the cpu socket) are much warmer (?) than before. Tonight will see how the overclocking goes...being Sandy am kinda happy that it's not cold bugging out. Wondering/hoping if I'll be able to improve on the prior owner's overclock.

Sounds like your on lock and load mode FTW