thorilan
06-20-02, 12:47 AM
ok back ground since you wont be able to read it in japanese.
this was done in 1996 . so this guy was a pro back then .
this person is using flow regulated liquid nitrogen with a secondary layer to contain the excess vapor that would come off of this increadable cooler. the funny thing is this system would not cost very much for the lenght of time that 1 bottle could cool your system
her is the link ... just look at the pics and you will probly laugh cry and bow in reverance
http://www.cds.co.jp/ff/pc.htm
tell me what you think so i can talk about this in class with my students
Silversinksam
06-20-02, 01:28 AM
How about calling in sick tomorrow and discuss it with us.
I'm assuming you can translate the article for us :rolleyes:
Roman79
06-20-02, 01:29 AM
That is a really good idea!
Any hard data? ie: ml/hr? To get an idea of how long a bottle would last
thorilan
06-20-02, 07:29 PM
well no real hard facts yet but a bottle should last at least 2 months . if you think about it 1 drop of liquide nitrogen at neg 100 degrees will freeze a good amount solid =)
cant call in work sick =(
but will have some more translation soon
mhayenga
06-20-02, 07:35 PM
How expensive is liquid nitrogen? Where do people get the stuff?
Just curious.
The Spyder
06-20-02, 07:37 PM
What type of cpu is that? Its OLD!
Spyder
thorilan
06-20-02, 07:43 PM
ok here is a rough translation after running this through the english translater website another student showed me. i cut and pasted this . btw will have some errors or diferent names for tech parts.
THIS IS 6 YEARS OLD. so compare the pics and stuff to the pasted copy here to read what it says
http://www.cds.co.jp/ff/pc.htm
You have to lower temperature, in order to carry out high-speed operation of the CPU.
C-MOS is because propagation time becomes long in proportion to temperature.
One more and power supply voltage are raised.
In proportion to power supply voltage, frequency of operation goes up C-MOS.
However, although voltage is raised how much, it is a device before and behind rated 3.4V.
When applying 10V, it is hard and Koto who will be BUCHI 壊れる is not in imagination.
I will cool for the time being.
This is a PERUCHIE element.
(PERUCHIE is an inventor's name)
As for this model, PERUCHIE has two-step structure.
It is heat when PERUCHIE is also made into two steps so that a dry cell may be connected in series and voltage may be raised.
The amount of movements rises.
Although it is the 静特性 of this, the maximum difference of temperature goes to 80 degrees C.
This is one-step PERUCHIE of FUTSU.
Although a size does not change so much with two steps of things, a price is cheaper than one half.
It is the type here which can usually come to hand by AKIBA etc.
Since PERUCHIE is not heat exchange but heat movement, a heat dissipation side is power consumption [ of PERUCHIE ] +.
A part for 吸熱 is 排熱(ed).
This is a heavy quantity. Incidentally, in the upper two-step PERUCHIE, it adds 7V and is 21.
since it A flows -- power consumption -- 150W -- it is near
If heat is not radiated well, PERUCHIE will blow this in an instant.
See this DEKAI radiator!
depth -- about 60% of breadth -- it is .
If it sticks and experiments in PERUCHIE on this, the bottom rapidly the temperature by the side of 吸熱
り -- a lot of frost is attached
ほ to which a place goes up steadily and the temperature by the side of 排熱 cannot touch this heat sink, either
ど -- great time is not taken to get hot
Although compulsion air cooling is naturally carried out using a fan
Isn't there any sound without are noisy and being able to build in PC generally?
Then, it becomes the appearance of a water-cooled system.
A water-cooled unit is stuck on the 排熱 side of PERUCHIE.
What carried out fraise processing of the copper block be elaborate.
A length of one side is united with P54C.
Although this is made to circulate through water, amount of water and the flow velocity are the conclusive factors of heat exchange.
Although the fuel pump for cars is very highly efficient, if it lets water pass, rust will be generated, and it is in two - three days.
It stops moving.
Then, it experimented instead of water using oil or chlorofluocarbon (liquid).
The inside of 水枕 is like this.
For the right, the left is a thing after fraise processing, and the thing to which surface area expansion was aimed at through the small tube.
It is the sound of a pump although CPU temperature has saved -10-degree-C order by this 液冷 system.
It is in the state where it is noisy, the fan for cooling cooling liquid is noisy, and it is .., and can use regularly.
There is no は.
And the total power consumption of this system tends to be set also to 300W.
Most of these electric power consumed becomes heat, it serves as が音 only, and worsens indoor environment.
It is ている.
Now, it is bad.
I will use carbon dioxide!
--------------------------------------------------------------------------------
It may have to do before carbon dioxide cooling.
It is work of a thermometer first.
Usually, as for the electronic thermometer marketed, measurement minimum temperature has many things per -20 degrees C.
However, the target temperature of the cooling system using the evaporation heat of carbon dioxide is -60 degrees C.
I want to also raise the response of a thermometer.
In per second one measurement, it is unsatisfactory.
Then, the thermometer was made using digital panel meter.
Pt, i.e., a platinum sensor, is used for a 測温 element.
Since resistance [ as opposed to temperature in the advantage of a platinum temperature sensor ] was decided, proofreading is very.
It is an easy point.
- If it proofreads at a point (60 degrees C and 0 degree C), the maximum error of the middle point will be settled in ±1 degree C.
Circuit composition is generated to the constant current circuit which supplies current to a platinum sensor, and the both ends of a platinum sensor.
OPEA for amplifying to the voltage for which digital panel meter needs the voltage carried out
It is MPU.
On the substrate, the circuit was constructed small and mounted at the reverse-side side of digital panel meter.
A power supply attaches the standard power supply connector for PC to take from PC.
The 0.65mm HORUMARU line was used for the line to a sensor in order to let between the legs of CPU pass.
Another subject.
It is a measure against dew condensation.
EKUSUTENDA suffers troubles and various kinds of trials are performed.
The most general one hardens and carries out the circumference of CPU with silicone rubber, a hot bond, etc.
まう thing.
However, then, maintenance nature is bad.
Then, it tried to have experimented in vacuum heat insulation.
namely, -- by the principle of a thermos, a vacuum layer will be prepared between cold CPUs and open air -- also saying
It is の.
The common name which this created from acrylics "a CPU house"
The acrylics board with a thickness of 3mm was used.
Since processability of acrylics is good, for the use said like this, it is tight.
The bottom fixes between mother boards with a double-sided tape and little silicone.
The upper surface fixes the lid of acrylics nature on a double-sided tape.
Since an inside is decompressed, a mother board and a lid adhere exactly.
The CPU house of a photograph is completed through that trial production how many times.
Although the thin acrylics board (2mm) was used at first, atmospheric pressure was lost and it has destroyed.
It is の.
a pump -- the so-called "pump of a goldfish" -- 親分み -- the thing of the structure [ た / structure ] -- using it -50
It has decompressed to cm/Hg.
.. which a result has and dew condensation does not generate at all lets a CPU socket pass, and it is a mother board. [ good ]
The very thing will begin to get cold and ice will crawl to Mother.
Then, coating material was used for the mother board.
Alcohol is used as a solvent, and turned coating up, silicone is applied, and this is れ.
ば剥 is easy also at the time of す.
It will be hard coming to get dry and notes should flow out, if a lot of coating is given at once.
Moreover, it is the point of going up to a CPU socket and causing poor contact.
However, it is this much easy and, otherwise, the thing finished in KIREI cannot be considered.
--------------------------------------------------------------------------------
The accident is an accessory at a clock rise.
Until now, one Mother, two I/O cards, and two HDD have been sent to the next world.
An I/O card will be the cause of a power supply (probably).
This card is another power supply about 5V system power supply, although ±12V are needed as an object for RS-232C.
(-- if it supplies not from the power supply for AT but from) -- the standup of 5V system and ±12V system -- そろわ
It is lost.
Probably it would also cause the latch rise owing to.
COM 1/2 has died.
About HDD, it cannot say not at all. (って声 which "is the cause of a clock rise" is sonority.)
る ..
Even if it is using the machine thing ordinarily, a life is because it visits.
How to break can do a lot of poor sectors, and stops receiving even a physical format.
IDE is weak although the thing of a SCSI interface will seemingly be comparatively strong.
For KATTARUI って people, starting from FD is using cheap HDD for a test.
We recommend you.
--------------------------------------------------------------------------------
On Mother, since a picture of the picture of force was not able to be taken, it is movement on a table.
To the block in a CPU house, it is a meter ring valve (兼 expansion valve).
It let it pass and carbon dioxide was supplied.
A meter ring valve is for industry and bears 50 kg/cm2.
A corrosive one-proof is made of good SUS.
A hose is stainless steel RESUME about the surroundings of the Teflon tube currently used for the racing car etc.
What was covered by SSHU.
This is OK also by two or more 150 kg/cm pressure.
Having scattered around is the so-called dry ice.
Under normal temperature 常圧, liquefaction carbon dioxide is not with a liquid. It solid-izes immediately.
Appearance! This temperature.
Even the temperature of CPU can save -60 degrees C or less.
The temperature of the copper block [ itself ] is lowered so far.
Although this is used by air release for the experiment, it carries out a part to dew condensation prevention in fact, and it is CPU.
It returns to a house and also is emitting in the pipe outside the window.
if it emits indoors -- the head -- since it becomes painful
The figure covered with dry ice
If dry ice is stuck for the small tube inside a copper block, naturally it will not get cold.
A meter ring valve adjusts the moderate amount of gas.
Incidentally, if it takes out also like 100 cc/m (as normal temperature gas capacity), it will get cold enough.
carbon dioxide will be stopped if PC is turned off -- as -- electromagnetism -- a valve is used
Since it is time and effort when having shut the valve one by one.
I install a large-sized (40kg is entered) cylinder in the outdoors besides a small cylinder which is in a photograph.
It is carrying out.
Although the length of piping is set to about 5m, since it is in the state which is not evaporated, there is no fear of piping getting cold.
--------------------------------------------------------------------------------
Do you cool temperature to -60 degrees C at the time of operation of CPU, and how much carry out improvement in a performance?
It is the maximum. Although based also on the performance (at the time of a clock rise) of a circumference tip, is it 50% APPU って所?
P54C-90/100 move by 150MHz at last.
A place [ P54C-166 have good 230MHz grade ].
Since this moves by about 200MHz also by air cooling, it is considered that there is no cooling effect not much.
It could be said that tuning is as difficult as the high device of a nominal clock.
Moreover, performance does not come out of making the Mother clock low and moving it by 3X.
Since the rate of bus use is high, P54C must move the exterior at high speed.
When considering use by 200MHz, they are cash and DRAM まわ from CHUN of CPU.
I want to attack り and to challenge Mother clock 100MHz.
--------------------------------------------------------------------------------
The state where it set to Mother be elaborate.
The meter ring valve is also covered from acrylics.
The place which looks white is silicone.
The white hose extended from a CPU house is an object for carbon dioxide discharge.
The black hose of the next door is a CPU house again, after returning some discharged carbon dioxide to room temperature.
It is alike, and it leads and is made for indoor air (for humidity to have been included) not to enter in a CPU house.
Can it turn out that the copper block in a CPU house has not dewed at all?
Moreover, the entrance of gas is using the place of a card slot so that the case of PC can be shut.
The fan (the line of red and blue and a main part are black) seen to the front (kana ? to understand) is an object for DRAM cooling.
SB16 and an Ether Net card mount in an ISA slot, and PW864 mounts in a PCI slot.
It is carried out.
--------------------------------------------------------------------------------
For delay decreasing, if temperature is lowered, P54C is [ C-MOS ] temperature why but as stated first.
Is it alike and can't high-speed operation be carried out [ is it in inverse proportion and ]?
The process of P54C is Bi-CMOS. That is, BAIPORA and C-MOS combine.
Although it is thought that it is used for places, such as a bipolar は bus driver, it is 低 of this bipolar 、 temperature.
With the bottom, hfe falls and SURESSHORUDO goes up.
That is, it is hard coming to carry out high-speed operation.
It is considered [ whether this hit offsets a performance and ] ?.
moreover, CPU voltage -- 3.8V-4.6 -- although there is a point good (on high-speed operation) for per V -- this にしても
pure -- if it becomes C-MOS -- voltage -- you may be proportional -- that is right
Although some expectation is carried out to P54C-200, it sees from the characteristic of -166, and 240MHz of air cooling is すれ of operation.
I will carry out ば size.
As long as it moves by a pipeline's shallow P54C so far, you may say that it is splendid.
Since there is also a talk that it becomes all C-MOS, I want to expect next P55C according to the rumor.
As for that that seems to be right and P6 with a deep pipeline (Pentium Pro), 300MHz operation is also checked.
Why don't I change to P6?
An answer is easy ". Since it is high " .. although there is also って事, the favorite has stepped on with the next term P6
It is made in Intel as it understands, if the fate to which 60/66MHz which appeared with great fanfare also followed Pentium is seen.
I do not want to fly at the initial version of elegance.
--------------------------------------------------------------------------------
Is a life shortened if high-speed operation of the CPU is carried out?
An answer is Yes.
Although a design life does not know how much it is, it may fall by about 1 figure.
Since it is a semiconductor, degradation is high speed about the inside of thin wiring, although ? may be thought in that there is almost nothing, and it is 電.
CPU to which a child is moved deteriorates gradually.
Does it cry, if the designer of CPU sees, since more than rated voltage is applied and it is moved above a rated clock?
surely .
However, there is nothing that broke as my experience (is it broken ..?).
There is nothing to the extent that it can prove clearly also about characteristic degradation.
Since other portions become doubtful previously rather than CPU breaks, it may be ..
Hmm. Babel fish cannot translate it??? odd.
Google's translation is stupmed too:
http://translate.google.com/translate?sourceid=navclient&hl=en&u=http%3A%2F%2F216%2E239%2E39%2E100%2Fsearch%3Fq%3 Dcache%3AFnR5ntj9E8AC%3Awww%2Ecds%2Dco%2Ecom%2Fff% 2Fpc%2Ehtm%2B%26hl%3Den%26ie%3DUTF8
The above translation is very unclear, that's why I am trying to use a different system.
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.