chilly1 It would be amazing, but i don't think cpu will work at -150C or -200C it's too low for such electronic
Printable View
chilly1 It would be amazing, but i don't think cpu will work at -150C or -200C it's too low for such electronic
Chilly wrote:
Hello there ... even though I am pretty new to phase change stuff ... the quote kinda hits on something that I have though about for a while now ...Quote:
I would be great if we could replace the chip with a manafactured one that had channeling for refrigerant and ports and lines to attach. So that we could cool the circuits directly no heat spreader and no conductors between the refrigerant and the silicon..
Even though the CPU does have a heat spreader, no channels for direct contact cooling, etc ....
Wouldn't we get better results starting to cool the CPU from both sides??? This might sound and/or be awfully stupid, but it's something that has been spinning in my head for quiet some time time now ... On the other hand it could be a lot easier to protect this capsule, and thus the cpu from condensation ... :confused:
I guess that in order for this to work the CPU would have to be encapsulated in its own block ... pins from the CPU would have to be extended and the capsule would plug right into the socket ...
What do you think? Is this the stupidest idea ever, or might it be a "beginning"??? :stick:
What would be absolutely wonderful would be a 1/2 inch hole through the mobo, directly behind the CPU. We could then cool the back of the chip as well as the front.
I know that a capsule would be a pain in the ass to design and build ... However, I am sure that somebody here is be capable of accomplishing such a task.
The capsule would be constructed using high performance insulating materials like microlen or epdm ...
However ... at the very bottom we could use a non conductive material, which will transfer heat, and go prommy style, using a heating element which will heat up the extended pins, thus preventing any condensation.
Thanks Gary for not having declared me in saint ... :)
Check this out
I seriously dought it will work on every motherboard, but it worked on that one.
Guys... w8 for BBUL technology where the die is available from both sides and you have smth like a slot celery, anyone remembers slot celeron sandwiches ? :D
OOO dual head -150 chiller total freeze out... grease the slot and seal it to the board... souns like a winner...
IBM did some testing on processors and their findings are the optimal temps are around -178C . The -120 C failures we saw last year were with liquid nitrogen and were colled quickly therefore I think the temperature gradient within the chip destroyed the electrical connections to the silicon... "shrinkage"Quote:
Originally posted by root
chilly1 It would be amazing, but i don't think cpu will work at -150C or -200C it's too low for such electronic
Heres a article: http://anandtech.com/cpu/showdoc.html?i=1542
Most electronic components operate better the colder they get and have been shown to operate at <10K. The exceptions are components which require a chemical reaction to operate such as batteries and capacitors.
I think it most be the mechanical stress that stops those cpus operating. Thermal shock, you can prevent thermal shock in 2 ways.
1. Reduce the temperature slowly
2. Reduce ambient.
The best solution is probably both. If you can maintain ambient temps to -30C you could have an evap temp 0f -150C and it would have the equivalent thermal shock as evap temps of -100C(and we know these dont shock the cpus too much)
Want to get to -178C then reduce ambient to -60c.
How cold can you make a closed loop recirculating nitrogen atmosphere Chilly1 :D .
A double skinned chamber would work for the mobo, external connections wouldnt be a problem as the inner atmosphere would only need to be maintained at 1 or 2 psi to ensure no air ingress, so sealing wouldn't need to be that extreme.
Wish I had the practical skills to build these systems:(
Regards
John.
As always, there are three important questions involved in this:Quote:
How cold can you make a closed loop recirculating nitrogen atmosphere Chilly1 .
What is the heat load?
How cold do you want it?
Will that be Visa, MC, or Discover Card?
Do you take coupons Gary?:D
regards
John.
How would capacitors handle the vaccume?
what do they use in space?
Heat pipes on the dark side with fins to radiate into space, Conduction and infrared radiation work quite well when you take into considering that on the dark side of the ship it is near absolute zero..Quote:
Originally posted by Redwolf
what do they use in space?
Capacitors because they are filled with a volitile oil would need to be made of stronger materials than their current construction, In a vacume they would distort and may even burst... When a non-electrolytic cap is shorted or overvolted it bursts and expells the oil plates and paper everywhere...Quote:
Originally posted by someone
How would capacitors handle the vaccume?
I can make a closed loop nitrogen chiller recirculating with the refrigerating equipment I have run at 70 kelvin at 40 watts. zThis uses a hydrogen cryo pump a vacume chamber and a cryo torr head, The cryo torr has a 2 stage stirling engine powered by 115v and the cryo pump a hydrogen absorber to remove the heat of compression from the stirling engine... Capacitors do not use a chemical reaction to operate. Mearly the proxmity of the electrons to the lack of electrons wether you consider flow from to or to from ...Quote:
Originally posted by pythagoras
Most electronic components operate better the colder they get and have been shown to operate at <10K. The exceptions are components which require a chemical reaction to operate such as batteries and capacitors.
I think it most be the mechanical stress that stops those cpus operating. Thermal shock, you can prevent thermal shock in 2 ways.
1. Reduce the temperature slowly
2. Reduce ambient.
The best solution is probably both. If you can maintain ambient temps to -30C you could have an evap temp 0f -150C and it would have the equivalent thermal shock as evap temps of -100C(and we know these dont shock the cpus too much)
Want to get to -178C then reduce ambient to -60c.
How cold can you make a closed loop recirculating nitrogen atmosphere Chilly1 :D .
A double skinned chamber would work for the mobo, external connections wouldnt be a problem as the inner atmosphere would only need to be maintained at 1 or 2 psi to ensure no air ingress, so sealing wouldn't need to be that extreme.
Wish I had the practical skills to build these systems:(
Regards
John.
Ah Absorption. There you go.. buy yourself one of those over priced NG Fuel Cells and use some of the extra heat for a Helium Heat Pipe system.
Well the vacume pump I am still looking for and a fullsized chamber would be nice I currently have one for a single 12 pin ic. The helium heat pipe equipment hasnt come up on Ebay yet but the cryo torr stuff has ...the vacum pump I want was 12,000.
lol
The ultimate goal is to freeze one cubit foot of air.
Ah you need to get creative and start making your own I think. Thought you had a line on someone with a CNC?
I do but he still charges by the hour, the router table I can get to use won't do over 2 inch thick....,,,
LOL Gary :DQuote:
Originally posted by Gary Lloyd
As always, there are three important questions involved in this:
What is the heat load?
How cold do you want it?
Will that be Visa, MC, or Discover Card?
Why dont we just tape a computer to the back side of a spacecraft, it would be cheaper :shrug:
J/K
LoL
...nice...
That doesn't disagree with what I said! High temperature makes conduction band promotions easier, but hurts mobility. The effect this has depends on the material.Quote:
Originally posted by zabomb4163
[B]No...
http://www.allaboutcircuits.com/vol_1/chpt_12/6.html
http://sub.allaboutcircuits.com/images/10226.png
"meaning that resistance increases with increasing temperature"
most materials increase in resistance with increase in temperature. certain substances such as germanium and silicon decrease in resistance with increase in temperature.