To chill ot NOT to chill...-140C that is.

[charlie]

So,
I've got the cryo-chiller...
But to make it work I have to charge the system and plumb on some evap's. $1075 per charge. Youuchh. So that's the cost of a Prometeia Mach II, for refrigerant.
Captain Cascade checked out the system...it's ready to roll. So that's the question today, fellow hobbyists...how much is -140C on the evap worth? Some people have talked about how too much cold can be a bad thing for a CPU, I don't understand this, LN2 is much colder....is there such thing as too cold? I think not.

C

[sandman]

Wow, 1075 is alot of money, but then it's also like what 80C colder than a prommy mach 2?

[eva2000]

that's alot especially since that's USA dollars :eek:

[isp]

$1075 that's sick! :eek:

I hope you go through with it still though, just to see all the results and pics etc :D Your money though...

You didn't pay that much for it though right? You said you got it from ebay IIRC?

[Major]

I'd shop that price out ! I'd say for that much you could get your epa license, buy all the stuff you need to do it, and have a side job reworking promies and vapo's to r404a ! lol

Do it !!! if you have the money !

You can't ever get it too cold, condensation prevention would be the only concern I can think off.

[captaincascade]

lol, thats his cost direct from the manufacturer. dck

[charlie]

The problem is that it's a proprietary blend of refrigerants...the exact quantities and ratios is unknown. Cascades are tuned for the boiling points of specific refrigerants....just loading r404a won't work. Luckily I got it off Ebay or $76...
So I guess my total cost with a couple Baker18 evaps would be about ??? $1400 ???
C

[captaincascade]

its an autocascade, i cant just guess what goes in it, i didnt engineer it. its a wierd little thing.

autocascades take very precise blends of lots of different refrigerants

[Major]

Man I'd say it would be kick freakin azz !!!!!

For the price of a Mach 2 ?? whats the question again ?? :D

[charlie]

Originally posted by Major Slaughter
Man I'd say it would be kick freakin azz !!!!!

For the price of a Mach 2 ?? whats the question again ?? :D

Well, yeah :D
It's a good unit, master wrench Captain Cascade on the job...
I'll have the unit manual on Wed. hopefully it'll explain things a bit more...I'm just curious...at those temps won't the whole motherboard from the ddr slots to the last pci slot just turn into a sheet of ice? How do you ice/water proof something like this??
C

[sandman]

just die electric grease the whole damn thing.

Or put it all in mineral oil or something non-conductive.

I'd imagine at that temp the traces on the mobo would generate condensation.

[charlie]

Hmmmmm....like a shallow pan, or tray with dielectric mineral oil, good idea, I like that!
C
here's the msds for "polyCold" proprietary refrigerant

[Ranger2992]

LOL this is too funny. You're the guy who beat me in the bidding for the cryocooler. I thought I had that thing for sure and I went to bed knowing it was gonna be mine. I wake up in the morning and I got outbid by $1. Oh well I am glad it went to somebody that can try and make it into a supercooler. Good Luck :D

[charlie]

Originally posted by Ranger2992
LOL this is too funny. You're the guy who beat me in the bidding for the cryocooler. I thought I had that thing for sure and I went to bed knowing it was gonna be mine. I wake up in the morning and I got outbid by $1. Oh well I am glad it went to somebody that can try and make it into a supercooler. Good Luck :D

lol...yup I'm the "Sniper"

[Xenogias]

yeah, you'd have to insulate the §§§§ out of it, and then probably immerse it in some kind of mineral oil, and then maybe even heat the oil in some way so it wont freeze near the evap. It's doable. Can you make it work for your vid card as well?

[xDUCK]

Originally posted by Ranger2992
LOL this is too funny. You're the guy who beat me in the bidding for the cryocooler. I thought I had that thing for sure and I went to bed knowing it was gonna be mine. I wake up in the morning and I got outbid by $1. Oh well I am glad it went to somebody that can try and make it into a supercooler. Good Luck :D

Hey there... Funny how many fish are here at XS. ;)

Btw, there's a new 2k1 and 2k3 champ in town. :D:cool:

[DaBit]

I have written some software which can solve the set of equations needed to give you a basic mixture with just about the right molar fractions of the various components.

I could run this software to give you a basic hydrocarbon mixture consisting of methane, ethane, propane and butane, which should work fine. From there you can experiment. I once ran this software against a mixture with various known molar fractions for a 80K cooler, and results were off by about 15%.

It is the result of my efforts to build an autocascade, but I decided to try regular cascade first, and then autocascade the low stage.

As initial input I need the design low side pressure, design high side pressure and target evaporation temperature.

[Popcicle]

There ya go.....

[]JR[]

WRT to condensation take dabits approach, where theres no air theres no moisture, and therefore no condensation.

Build yourself a big box and vacuum the air out of it and put your mobo and gfx inside have everything else outside. Simple.

No messing about for rma's either since theres not 400miles of seal string and a gay parades full of vasaline smeared over everything :)

]JR[

[DaBit]

I am extremely happy with my airtight mainboard housing.

[charlie]

Dabit,
Very nice work. I'd be interested in seeing more info on your airtight mainboard housing!
And as soon as I receive the manual I will send you info about the autocascade spec's for a custom refrigerant blend.
C

[charlie]

Originally posted by ]JR[

No messing about for rma's either since theres not 400miles of seal string and a gay parades full of vasaline smeared over everything :)

]JR[ [/B]

LmAO.....
:ROTF:

[st0nedpenguin]

DaBit, thanks for the link to your project, it looks like I have something to get me through my boring lunchbreak at work now. :D

And charlie, I'll definitely be keeping an eye on this project of yours, do you have any pics of the cryofreezer?

[charlie]

i don't have a digicam and even if I did, the guts are enclosed in insulating foam...all you can see is the compressor, expansion tanks, condensor and fan!

[DaBit]

Originally posted by charlie
[B]Dabit,
Very nice work. I'd be interested in seeing more info on your airtight mainboard housing!

Visit my website: http://www.icecoldcomputing.com, and look under the 'old projects' menu item.

BTW: my hosting provider will soon move the site to another server, so you might experience hickups.

[st0nedpenguin]

Originally posted by charlie
i don't have a digicam and even if I did, the guts are enclosed in insulating foam...all you can see is the compressor, expansion tanks, condensor and fan!

That'll do nicely, I just wanted a basic look at one, I wouldn't be able to figure out the internals anyway. :D

[FUGGER]

Nice buy, I saw that thing as well but I went for the baxter cryo-fridge for $89

Get more prices for repair of your polycold.

2x 1HP cascade 500/503
http://fugger.netfirms.com/cryo.jpg

[xDUCK]

Originally posted by FUGGER
Nice buy, I saw that thing as well but I went for the baxter cryo-fridge for $89

Get more prices for repair of your polycold.

2x 1HP cascade 500/503
http://fugger.netfirms.com/cryo.jpg

You bought a -89-capable Cryo-fridge for $89?! :eek:

What's the catch?

[charlie]

The catch is usually shipping :D
These surplus machinery can very often be found for cheap, it's the shipping that'll bite you. My $76 cryo-chiller came from Florida to SoCal, ouch. I can kick myself for missing a PolyCold cryo-chiller P-75 (the little brother to mine) right here in SoCal that went for $450 and it WORKED...I wouldn't have this problem of finding/affording refrigerant!
C

[Popcicle]

...... you oughta be ashamed.

[charlie]

i know....

:(

C

[FUGGER]

Shipping was $75 I had it picked up by GI trucking and it was delivered the next day.

[captaincascade]

i doubt the software would work dabit,,, i dont know the size of the heat exchanger its all foamed in.... i also dont know why there are 2 expansion tanks with only 1 feed to the each of them.


yes 1 has a single captube running to it ..
the other has a single 1/4 line runnin to it?

i see no signs of selanoid valves or anything like that?
any ideas?


ULT-1786-5-AUA huh fugger?

[FUGGER]

Baxter 2186 Cryo-freezer (http://used-line.com/cgi-bin/used/~4237911_Revco_ULT2186_DBA_20_2_CF_Upri.htm)

I got mine from ebay, just a link to one like it

[charlie]

My fingers have been crossed for like 24 hours that there's a less expensive way to charge this beast...
When I get the manual I think it will help.
C

[DaBit]

Originally posted by captaincascade
[B]i doubt the software would work dabit,,, i dont know the size of the heat exchanger its all foamed in.... i also dont know why there are 2 expansion tanks with only 1 feed to the each of them.

For an expansion tank this sounds normal, right? All it has to do is to keep off-time pressure reasonable.

Maybe there are two for this reason: the more volatile components cannot start condensing until the intermediate heat exchangers are down to temperature. They might have added one on the high side, connected with 1/4" piping to keep pressure down during pulldown of the heat exchangers.

The captube-connected tank would then be the regular expansion tank, connected to the low side.

Am I right?

About the software: it is built using common sense only. All it does is to take into account the amount of low-volatile refrigerant needed to flash off itself and condense the high-volatile refrigerant. This results in a set of equations which the computer can solve. I once verified it to a known mixture for an 80K/2W application, and it was pretty close. Now, that doesn't prove that it always works, but it is worth the try. And it should give you a reasonable mixture to start optimization.

I would dump the program here on the forum if it wasn't such a loose piece of source code where you have to enter the parameters directly in the code.

If anyone has RefProp, I would be glad to find out if I could use the RefProp DLL to provide the necessary parameters...

[captaincascade]

1 expansion tank= 1 line conected
1 expansion tank= 1 line conected
2 expansion tanks= 2 lines

????????

[captaincascade]

fugger that means you have a 21 cu ft. you tell me the serial # and ill tell you month and year it was manufactured.

[DaBit]

Originally posted by captaincascade
1 expansion tank= 1 line conected
1 expansion tank= 1 line conected
2 expansion tanks= 2 lines

????????

I don't know what you mean. But am I right that the tank connected with 1/4" line is connected to the high side, and the tank connected with the captube is connected to the low side?

I would even say that the low side tank (captube) is equal or larger in size than the high side tank.

[charlie]

Dabit,
Indeed you are correct. The expansion tank attached to the captube is the larger tank. I'm very interested in your program, as the proprietary blend is simply cost prohibitive...
C

[DaBit]

Hmm, I will see if I can make it work for 'the general public'.
Currently you must hack parameters here and there in the few lines of C code, without logic at all. It was never meant for general public use.

[charlie]

Originally posted by DaBit
I have written some software which can solve the set of equations needed to give you a basic mixture with just about the right molar fractions of the various components.

I could run this software to give you a basic hydrocarbon mixture consisting of methane, ethane, propane and butane, which should work fine. From there you can experiment. I once ran this software against a mixture with various known molar fractions for a 80K cooler, and results were off by about 15%.

It is the result of my efforts to build an autocascade, but I decided to try regular cascade first, and then autocascade the low stage.

As initial input I need the design low side pressure, design high side pressure and target evaporation temperature.

I don't have the manual yet, but on the spec plate it says:

MAX OPERATING PRESSURE 325
HIGH SIDE TEST PRESSURE 450
LOW SIDE TEST PRESSURE 150

Don't know if that's the spec's we're looking for, but if it is...and we're shooting for a evap temp of -120C to -140C...what's the proggie say?
Thanks!
C

[DaBit]

No, they are not. These are the pressures you don't want to exceed.

Just another question: does the unit use captubes for expansion or thermostatic valves? In the latter case we might have a problem since vapour pressures must match, but not sure yet.

Can you give me every single detail you can find?

[captaincascade]

ok, steppin back......... good luck. most expansion tanks have an inlet and a outlet...... im sure theres a reason theres only one but all i can do is speculate

[charlie]

CC,
Lol...that avatar is too much...
I found out a bit more from an IGC-Polycold tech guy today...
And it's good news. You see these units are used for a ton of different things...even aircooling in -150C freezers! The evaporator (or heat exchanger or whatever) is completely "changeable". Knowing the specific load, any evap may be put on. Even a couple CPU/GPU evaps!! :D
He can give the tuning info depending on the total load, So I'm guessing 160ish for the CPU and 90ish for the GPU....about 250W total load. Unfortunately he didn't give an option on the refrigerant...it was quite funny actually, I got the impression they'd like to have it in their shop to do the mod :)
C

[DaBit]

Originally posted by captaincascade
ok, steppin back......... good luck.

Why? We probably need your experience hard enough.

most expansion tanks have an inlet and a outlet...... im sure theres a reason theres only one but all i can do is speculate

I always thought that an expansion tank was just meant to keep off-time pressures limited. In that case, a single connection would do.

To what lines is a regular expansion tank connected?

[charlie]

even without a digi cam I can still share with ya' guys what a -140C autocascade looks like :D
Yes it's crude but should get the point across.
remember the top of the unit is a compartment sealed in foam, this is where all the separators, etc reside. But I will show what is where and how the tubes run

[charlie]

and like this:
for you experts, that's a Tecumseh AJ4512A 230V 1PH

[captaincascade]

dabit, you are correct about the expansion tank be ing used for the off cycle. i just spoke with my father, who has worked with polycold in the past. Sometimes you dont thinkl about the simplist things:rolleyes: :rolleyes: :)

[charlie]

Yeah,
I just got off the line with PC, and indeed the larger tank is an expansion tank and the smaller tank is a buffer tank, serving the purpose of adding additional refrigerant volume and allowing the compressor to "start up". I'm waiting for an Apps engineer to get back with me about calculating the proper load for this model and how that applies to a couple CPU/GPU evaporators...
C

[captaincascade]

im just used to them bein piped from the high side with a selanoid and a high pressure switch. the outlet has a cap tube rolled and going to suction.

hey ask em if they remember jack lester from l.a. cascade.

[charlie]

Wusup Captain?
They seemed nervous about the whole remove the probe/replace with evap's thing....he transferred me to the Apps engineer to calculate if the CPU/GPU evaps would work. The tech guy said they do NOT under any circumstances change the length of the cap tube...so If I were to add evaps, I'd probably have to remove the whole flex line/cap tube (the part coming OUT of foam) it's probably like 30"....and use like 12" of flex and 2 x 12" flex for the evaps off of a manifold?? Just have to keep the overall design/enginering consistent with it's original purpose/application.
The tech guy also said they'd walk another tech through charging.
If I lived in SF, he says they'd do it for free, a leaktest and charge...
C

[DaBit]

Originally posted by charlie

for you experts, that's a Tecumseh AJ4512A 230V 1PH

If model numbering adheres to what Tecumseh says here (http://www.tpc-nacg.com/mdlnmbr1.htm), it is a 12k BTU @ 45F evap R12 compressor with high starting torque motor. At least this says something about the low side and high side pressures which can be expected.

But to be honest I am not sure. This compressor seems a bit light to do -140C at a reasonable Qmax.

Originally posted by captaincascade
[B]dabit, you are correct about the expansion tank be ing used for the off cycle. i just spoke with my father, who has worked with polycold in the past. Sometimes you dont thinkl about the simplist things:rolleyes: :rolleyes: :)

Nevermind. I am sorry if I sound a bit arrogant sometimes; it is not meant that way. I am a bit stubborn though, and I won't easily take things for granted without good reasoning behind it. I just heard a bit too much bull§§§§ from local ACR engineers for that (the kind of guys who wanted me to use an R404a valve for R134a application since 'R404a is newer and thus better').

But I'd like to cherish guys like you, Gary, Herefishy, Prof Sporlan, Bowman, and many others. You guys are the experts.

The 2-connections expansion tank with the solenoid+HP switch sounds like a good way to do it. This saves you the second expansion tank.

Originally posted by charlie
[B]Wusup Captain?
They seemed nervous about the whole remove the probe/replace with evap's thing.... The tech guy said they do NOT under any circumstances change the length of the cap tube...

Which is not strange. Autocascade balance is a very, very, and then I mean very delicate thing. I discovered that when trying to design my own (a Prom-like autocascade is much smaller than my 'classic' cascade I have built now, and therefore preferred). Modify one variable without adjusting the others, and you are left with a non-working autocascade.

so If I were to add evaps, I'd probably have to remove the whole flex line/cap tube (the part coming OUT of foam) it's probably like 30"....and use like 12" of flex and 2 x 12" flex for the evaps off of a manifold?? Just have to keep the overall design/enginering consistent with it's original purpose/application.

The inner volume of the flex line is important, but not nearly as important as captube length and restriction.

I would remove the current captube completely, and replace it with two captubes having the same resistance. You will have to measure the current captube, and match the two captubes against this tube. This way you won't disturb operation.

Here (http://www.cubigel.com/java/X?cgi=cubigel.infotecnica.InformacionTecnicaArticu lo2.pattern&seccion=informaciontecnica) is a description on how to do this.

The tech guy also said they'd walk another tech through charging.
If I lived in SF, he says they'd do it for free, a leaktest and charge...
C

Ship it to someone who lives there is not an options?

[Master Chief]

Any news? Running yet, lol.
Pictures would be nice, too.
MasterChief

[charlie]

OK
Well maybe it's a $76 paper weight :(
Today I spoke with the IGC-PolyCold engineer at great length about this project. This unit is designed (quite precisely) to run at a load of 100W +/- 10% at -130C, go out of the range and it will simply shut down. So it looks like this unit won't work for my intended purpose...unless I have it filled and put a Baker18 R9800 block on it (thats around 80 or 90w) and use it as a GPU chiller. Unfortunately the compressor can't be changed out for a larger one as the autocascade system is optimized for this EXACT charge/load. So that's it...end of story.
Now I'm considering tearing it all apart (to see all the fun stuff hidden in the foam block) and removing the expansion tanks and compressor and replacing them with a suitable 115V compressor, using the PolyCold condensor/fan and adding a dual evap setup with 2 Baker18 blocks...so then I'd have a portable rolling cooling unit working off 115V... maybe only -50C temps :( but that's OK too. BTW, if anyone ever sees the P-200 model rated at 200W, jump on it, it'd be JUST right!
C

[charlie]

Hmmmm.....what to do with all these autocascade parts, phase separators, stuff I have no freekin idea what it is, lol.........
C

[DaBit]

Noooooooo!

Don't tear it apart. Get Baker's blocks installed, and then we will see what we can squeeze out of this unit. I bet it will be more than -50C, but we might not reach the -120C.

edit:

And *if* you decide to tear it apart and ditch the pase separator etc, I want to be the first on your list to receive them :D

[]JR[]

Originally posted by DaBit
And *if* you decide to tear it apart and ditch the pase separator etc, I want to be the first on your list to receive them :D

Me 2nd...

]JR[

[charlie]

Originally posted by DaBit
Noooooooo!

Don't tear it apart. Get Baker's blocks installed, and then we will see what we can squeeze out of this unit. I bet it will be more than -50C, but we might not reach the -120C.

edit:

And *if* you decide to tear it apart and ditch the pase separator etc, I want to be the first on your list to receive them :D

No, unfortunately it just won't work. The engineer described to me how the compressor will simply STOP if the temp goes too high. The whole unit will shut down if the unit temp were to rise too far. It's a very strange unit, designed to work in a very narrow set of parameters, outside of this it just doesn't work, at all. As I disassemble this, I'll try to post pic's of what I pull out for everyone to see.
The only suggestion she gave me was to use the unit as a first phase for a second unit.......sounds a bit Large for a CPU rig.
C

[DaBit]

so, the unit has a protection which stops the system if the temp rises too much.

Now, that's great stuff to try your new shiny wire cutter on.

I simply won't believe that it is not possible to make an existing autocascade system with everything in place work. Maybe we won't reach the -120C, but only -70C or so. Maybe we have to modify the protection circuitry. Maybe we even have to modify the cooling system itself.

[charlie]

I'd certainly hope so....but the cost of experimentation is quite HIGH....$1075 for 1 (one) CHARGE!!! Last night I removed the factory "probe" (evap) and cut off about 1M of refrig line and cap tube. This way the unit moves around a bit easier. I will hold off on disecting the unit for now, the engineer "did" tell me that I could keep the system balanced by replicating the refrig line/cap tube length of the original....but the internal volume of the first evap (probe) must have been greater than that of an evap??
C

[FUGGER]

Charlie if you want to get me the unit I can get my Engineer to look at it. I already picked up a few pounds of second stage gas for my projects.

Want to take a Vegas vacation? I belive he can get your unit operational again.

[Hobocrow]

Charlie, wow!....Let's see...S. Cals' temps are in triple digits and it's on fire....Sounds like the solution to a $76.00 paper weight. I don't know about you, but....Seems like a great offer to me! :D

Edit: I forgot the show girls!!:banana:

[charlie]

Girls?
Ahhh...well, but I told my WIFE that Vegas doesn't do THAT anymore, you know it's like DisneyLand now! ;) Go to Vegas, though...she'd drag my a** to see Celine Dion I'm sure ;)
Don't know what's worse, air cooling the rest of my life or sitting through CD show??

[Hobocrow]

You're right...I wasn't thinking! :ROTF: ;)

[Hobocrow]

Originally posted by charlie
Girls?
Ahhh...well, but I told my WIFE that Vegas doesn't do THAT anymore, you know it's like DisneyLand now! ;) Go to Vegas, though...she'd drag my a** to see Celine Dion I'm sure ;)
Don't know what's worse, air cooling the rest of my life or sitting through CD show??

I know we need to get back to the serious topic and it is something to seriouly consider, I 'm sure!
Ok, ok, I promise not to go OT anymore....But, just this last one...time.
My wife has 2 sisters and they exchange Celine Dion cd's all the time...any time my wife wants some yard work done ...she plays the CD...makes the outside yardwork seem a pleasure! :D

Good luck with the "auto" :)

[charlie]

lol!

We gotta' do what we've gotta' do....

C

[Hobocrow]

Originally posted by charlie
lol!

We gotta' do what we've gotta' do....

C

:toast: True! :)