Noob with condesation

[blindfitter]

Hi Guys,

I am embarking on the sub zero route, fell at first hurdle.

Have one LD’s SS, will leave the P5b and 6700 on water for now, dug out dfi expert and opty 150 as guinea pigs for my learning curve.

Read and re read forums showing how to prepare the board.

Sprayed the back of board with Clear Lacquer – got DC4 grease into socket and around the surrounding area – fitted cpu grease the sides of cpu.

Next, cut up the foam for the back of MB, attached and bolt tight cut up front with hole for cpu fitted snug, filled any small gaps I could see with DC4. Mounted cooler to MB all fine so far.

In a recent thread, it was suggested to run the cooler with no volts on MB and check for frost (for 2hours) I did this and at around 20 minutes could see frost developing on the capacitors further inspection showed condensation on other important components.

This is surely not right? Or does the heat from the MB when running stop this?

Have now stripped it down, no condensation in the CPU area.

Currently drying MB out.

Please advise guys

Cheers Blindfitter

[tim-]

you should not see frost in 20minutes, you should do it again and do it right, proberbly you just need to screw it togetter little harder to make sure the foam seals against the mb..

[Avman]

It's likely that those components will heat up enough during powered use that the condensation won't form. Just to be safe, I would apply an extra layer of conformal coating to everything within about 3 inches of the socket. Make sure to cover/tape any slot/socket openings before spraying the CC. Alternatively, you could paint on some Liquid electrical tape (available at Home Depot etc..) or even nail polish to cover up any surfaces that might be exposed to condensation.

[Avman]

you should not see frost in 20minutes, you should do it again and do it right, proberbly you just need to screw it togetter little harder to make sure the foam seals against the mb..

I believe the caps and components he's referring to are outside of the foam. With the board not running while he's testing this it is certainly reasonable to see some frost.

[rogard]

Id use the preparation from here:
http://www.xtremesystems.org/forums/showthread.php?t=105834

and the insulation from here:
http://www.xtremesystems.org/forums/showthread.php?t=105942

Remember that it has to be as air tight as you can make it.

(sounds like its not tight enough or you dont have any insulation)

[blindfitter]

I believe the caps and components he's referring to are outside of the foam. With the board not running while he's testing this it is certainly reasonable to see some frost.

that is right, and those were the guides I read and followed,

the only item I didn't do was nail polish top of board around the cpu area, before applying dc4.

will try again tomorrow and see if we can improve before firing up the board.

a thought, does the cold temp travel pcb? to componants outside the insulated area?

cheers and thanks for comments

blindfitter

[[XC] gomeler]

You can get away with surprisingly little insulation if your evaporator is being fully loaded to it's capacity, therefore raising the temperatures. With everything powered down there isn't a source of energy to be removed so of course the motherboard is going to frost up. Motherboard components, particularly the PWM area, do generate a bit of heat and depending on the insulation do a decent job of either staying in above-dew-point temps or causing condensation that does form to melt, frying the board. I'd suggest running the system and keeping an eye on if, flucuate between load and idle, and then if you are worried let everything warm up and tear your insulation down and inspect.

[blindfitter]

a question on insulation material,

is this: the same as armaflex

Expanded Neoprene Sponge Rubber Sheet

Expanded neoprene sponge rubber with or without an self adhesive backing. Ideal for use as cushioning and sealing for control panels, etc. Similar in appearance to mouse mat material or wetsuit material. CLOSED CELL

availble: 3mm, 6mm and 10mm


PROPERTY
TYPICAL VALUES

Base Polymer
EPDM/Neoprene

Density (kg/Cubic Mtr) Approx
90

Colour
Black

SAE Specification 18-R
SCE42

ASTM Specification D-1056-67,68
SCE42

ASTM Specification D-1056-78
RE42

ASTM Specification D-1056-85,91
2A2

Ozone Resistance 72 Hrs @ 50 PPHM ASTM D-1171
0

Temperature Resistance Low˚C
-30˚C

High & Intermittent
93˚C - 65˚C

Compression Deflection 25% kpa varies with thickness
35-63

Shore 00 Durometer Approx Average
50

Water Absorbtion By Weight Max ASTM D1056
5% (10% Allowed)

Heat Aging (7 Days @ 70˚C) CD% Max Change
30

Elongation % min
150

Flame Resistance UL94HBF*HF-1*V-O
Pass

1056 Suffix M MIL-R-6130C
Pass

GM6090MB4,A
Pass

Flame Resistance, FMVSS302” Min Thickness
2 mm

[Avman]

that's the stuff

[blindfitter]

thanks guys,

will get some on order, looks like I'm going to be using it.

thanks for the quick responses.

No doubt I'll be back with questions as my learning curve progresses.

Still everyone started as a noob

cheers blindfitter

[wdrzal]

insulation is in 2 parts to prevent condensation. insulation must be thick enough. and vapor barrier must be complete so no air moves from inside to outside. insulation must fit around tube without air gaps under it.

1 insulation
2 Vapor barrier


some insulation(closed cell) work as both in 1 some do not.

last resort cover insulation with a thick coat of Vaseline for a vapor barrier,keep of electrics ,I never checked if it was dielectric.

[Ssilencer]

a question on insulation material,

is this: the same as armaflex

Expanded Neoprene Sponge Rubber Sheet


That's great stuff, I can't get neoprene here so I use armaflex, and now I'm using liquid electrical tape instead of the nail polish as it is a lot easier to get rid once you want to sell the mobo.
As some of the guys told, expand your insulation a little more and cover the capacitors with dielectric grease.
Also, there is no need at all to run your computer at low volts, if you want to get condensation you will, run the unit without turning on the computer and sooner or later you will get condensation, but this is as easy as in real day by day use you will never do that, so it doesn't make any sense.

[blindfitter]

Hi Guys,
Back again,

Just redone the board with expanded neoprene a lot better, funny how it gets better the more you do it.;) ;)

I’ve got the self adhesive stuff; do I put the adhesive to the board? Does it come of easy?

Will put pics tomorrow of progress, also firing it up tomorrow,

Question, as yet I don’t have K type thermometer, but do have a Fluke73, can this be used to give temp. As I measured the Ω @ room temp and then switched on the cooler and it changed, not a great deal, started at 13.4Ω and rose after 5mins to 18.2Ω just wondered if there was a conversion table, and yes I know it won’t be that accurate.


Cheers for now Dave

[tim-]

if you got a pt100 or ntc resistor you can read the Ω and then know the tempture... there is lists available for what tempture pt100/ntc have at specific Ω

regards
Tim

[[XC] Hicks121]

Fluke 73 temp probe, type k http://us.fluke.com/usen/products/AccessoryDetail.htm?cs_id=8572%28FlukeProducts%29&catalog_name=FlukeUnitedStates

[Nosfer@tu]

I did a Move of how you can Insulate with Sealstring :)

Itis in raw format and 455 mb for 5 min so Need to convert it and seed it as a torrent. :)

[teyber]

another try is:

nail polish
cork tape(black gooey stuff)
1 layer of neoprene
dialectic grease wherever

[blindfitter]

if you got a pt100 or ntc resistor you can read the Ω and then know the tempture... there is lists available for what tempture pt100/ntc have at specific Ω

regards
Tim

cheers Tim,

can you explain a bit more ,detail, sketch ?

as will get the proper job soon, but for the short your suggestion would be most helpfull.

regards

[blindfitter]

Hicks121;2223441']Fluke 73 temp probe, type k http://us.fluke.com/usen/products/AccessoryDetail.htm?cs_id=8572%28FlukeProducts%29&catalog_name=FlukeUnitedStates

thanks for the link, seems an expensive way to get temps, plus only goes to -50c,

its cheap to get dedicated tool, unless I.ve missed something

regards

[blindfitter]

Some pics of my insulation,

Have it up and running, 150 opty boots and runs memtest @ 3252mhz 1.65volt -26c, but i cant get xp to load to hard disc, so far 2 hard discs, 1 sata 1 ide, 2 dvds 1 sata 1 ide, 2 copie of xp, starts to load but fails at install about half way through, any clues, also changed ram, and have put back to stock speed to install xp

6mm expanded foam
http://img296.imageshack.us/img296/9030/dscf0189ev3.jpg
10mm expanded foam
http://img99.imageshack.us/img99/8258/dscf0190yr7.jpg

check to see contact
http://img524.imageshack.us/img524/7304/dscf0187fj3.jpg
built and ready to go
http://img183.imageshack.us/img183/3933/dscf0191th8.jpg

cheers guys

[tim-]

it looks good, should not be any problems with that setup!

regards
tim

[teyber]

make sure its properly insulated on the bottum, and try putting cork tape around all the capacitors. Also, nail polish first layer?

[tim-]

blindfitter:
I didn't see your question earlier.. I will try to explain it for you, hope you understand.

a pt100/ntc resistor is most common in digital termometers etc.

they have specific resistans at specific tempture. pt100 means platina with 100ohm @ 20dgr C and the resistanse increase with tempture.

NTC, which is an termistor (correct spell?) which works the same way but the resistans increase when the tempture gets lower.

really easy and works good on SS, isn't enough polite lower then 50dgr c or something.

the good thing with those is that they doesn't need a reference as typ K, J, T.

you simply read the resistans over the pt100/ntc component and translate it to C by using a pt100 scheme.