ECI 33+ (3 Stage Cascade Worklog)
 

Here is the project log for my current refrigeration project. A 3 stage cascade for benching CPU's to their very max at the flick of a switch, no liquid nitrogen around here. Better. Adjustable STABLE temperatures without pouring or filling.

Sponsors:
ECions, Extreme CreatIons LLC
Maybe more to come? Who knows who might want to see their product get frozen over.



I joined (followed Buckeye actually) to share with you guys some current phase projects.
Hope you don't mind! :wave:

So here's some autocad!

Estimated Temperature: ;) <-- Ha like I'd tell you that ;)

http://img8.imageshack.us/img8/5355/northeast.jpg

http://img8.imageshack.us/img8/5359/northwest.jpg

http://img8.imageshack.us/img8/8343/southeast.jpg

http://img8.imageshack.us/img8/2424/southwest.jpg

UPDATE JUNE12th.

Little update.
Got home today, gotta start unpacking a huge amount of stuff from school. But, ran down stairs, and threw some crap on the table.
Hope you enjoy.

http://img35.imageshack.us/img35/120/eci33parts2.jpg

Some fairly sloppy grinding and heavy red krylon paint later,

http://img40.imageshack.us/img40/765/0622092009.jpg

A milling machine! Richard got it into the workshop on this dolley of his. I was scared watching this thing jump up and down with each step. (Had to underinflate the tires on the dolley to take 300lbs safely)

Welding up a new work table, mainly for the project, but also for the milling machine.
Then we can start plotting the cage :)

http://img36.imageshack.us/img36/3205/0623091838.jpg

http://img235.imageshack.us/img235/8387/0624092241.jpg

Table 1 done. Milling machine going on up tomorrow.

http://img265.imageshack.us/img265/3509/0626090014.jpg

Winch + ceiling beam = Mill on Table!
I'm happy it made it, table seems more then okay, but conducts vibration a bit much.
Going to get some more MDF under it, get it all bolted down. And then figure out why the spindle ain't spinnin!

http://img198.imageshack.us/img198/597/0630092229.jpg

Everything moved over, so time to start laying out parts.
Might have a lead on the condenser I'd like for this, so the front is a little empty.
Compressors that seem "used" are working pulls from my own personal cascade that well, I built and never got around to finishing. So they're brand new still.

Layout time.

Brazed the first suction stem on compressor numero uno. It's nice to do them like this, since they'll shoot straight up to the heat exchanger coils. Nice and easy to insulate.

http://img148.imageshack.us/img148/6873/0701091913.jpg

http://img9.imageshack.us/img9/7853/0701092106.jpg

Heat exchanger time.
You can never have enough copper.

Making some fitting attachments.
Cleaned and shined;
http://img254.imageshack.us/img254/2241/0701092213.jpg

Survey says...
OIL SEPARATORS!
http://img29.imageshack.us/img29/2681/0701092217.jpg

Very nice man!!! Keep the pics comin.

It's Nice Pics NoL
Good Work Men Let's Some pic More

Let me explain this picture a bit.
It's fairly hard to shove one tube inside another tube. It's especially hard as the lengths required get long and longer. When making a heat exchanger using 3/16 and 3/8" tube, it becomes an exercise in insanity.

Both tubes must be nearly perfectly straight.
That means, even the tinyiest imperfections become a problem. As you begin to push one through, friction becomes a huge problem. Even a degree or two at a tiny spot pushes parts of the tube against the inside of the other tube.
This isn't a problem with performance, simply that it becomes impossible to push one tube in any farther. You must then withdraw it, figure out what's stopping you, touch it a tiny little bit, and try again.

My method involves 5 ft long steel beams and clamps.

Clamp the tubes straight as possible.

And try.

When your looking down a 20ft long barrel though, things get interesting.
This took 3 days.

http://img13.imageshack.us/img13/9/0705092354.jpg

And bent on up.
20" by 12". 4 loops or so tall.
Should be adequate heat exchange. And nice and easy to insulate, I hope.

http://img39.imageshack.us/img39/9324/0706090050.jpg

http://img110.imageshack.us/img110/8...0706090113.jpg

Wow NoL awesome project you got going here :thumbup: Personal build or for someone else?

Just mine at the moment. Doubt anyone can afford this one :P

I can't afford any of this shat but somehow I keep right on buying..... pullhair

Ain't that the truth.

Just a guess but $3,000 about right?

Around, yeah.
Bit more with the full load of features I'd like to build into it.
Something about making the best god damn 3 stager in the USA comes back to my mind. :)

I thought u already did that with sdumper's :D

That's why this has to be better ;)
I tend to hit a similar temperature though, but with better adjustment.

Don't tell me your making another Cascade on a Red Harbor Freight Cart :rofl

Or if this is the Mystery Man Cascade :rofl

Oh come on Buckeye. We all know I'm more talented then that :P
I'm MIG welding up a frame actually. Starting on it tonight,

Second heat exchanger rolled, CPEV's tested played. Nice and up front. They will be "separately" insulated in their own little sealed box.

Now thats sexy. And under 3 inches tall.
http://img269.imageshack.us/img269/1151/0706092329.jpg

Nice work NoL, so your shooting for temps around -130 but with better load handling?

I would suspect after reading about your other 3 stage and the amonut of load that one could hold it would be some where in the range of -130c unloaded and below -100c fully loaded.

I don't remember what loads were used in that past 3 stager tho.

I would like to see a load handling ability of 300watts or more to be able to handle the high heat loads of Core I7 chips.

From my tests with my 920 D0 and LN2 I was able to take my chip to -130c and was still able to boot at that with a 5.3ghz over clock and vcore of 1.575.

How ever the sweet spot for this chip was in the range of -90c to -110c and did not find a real use of going lower for high OC's.

How ever with other chips this might be different.

Also AMD chips would love the extra cold and lower temps are better with them.

My thoughts on single or multi stage units is to go for max load temps and generally try for higher load handling ability as you never really know what chips you will put this on later.

From what I understand about these units is a low unloaded temp is not allways the best for handling loads. Better tuning of units will result in better load handling vs lower unloaded temps.


That stack is looking very nice tho NOL, good work !

Thanks for the clarification buckeye :thumbsup: Im no phase expert (I prefer LN2) I just remembered seeing sdumpers cascade at -130 coldbugging everything it touched lol.

I would eventually like to get a nice 2 stager but costs are very prohibitive for me. I got a wife n kids (hungry kids!) to feed. :pile: (thats a pile of kids to feed)

Yeah I read that also about cold bugs. Different chips do different things, my 920 can boot at -130c under LN2 but some others do not have that luck.

I don't remember if that first 3 stager had a hot gas bypass on it. I think those are really needed for these things.

On mine if I start getting CB's with other chips I have to shut down the 2nd stage and wait until pressure has equalized on the 2nd stage before firing it back up. Usually what will happen after I shut it down is temps will slowly start getting warmer and about the time pressure has equalized I am good to go to fire it back up again.

Before I do fire it back up I start the compuer so I am in windows before I reach max temps. It's a pain to do tho. but workable.

Oh yes. Sdumpers unloaded towards -150C. Coldbugged motherboards and RAM too ;)
The CPEV on all stages will allow a certian amount of dialing in of the temperature. Unloaded might be below -150C with the valve closed, but real useful tuning around -125C at i7 loads is a good goal. It will have a hot gas bypass as well for expressedly warm situations, and a controller on that to actually have it control a temperature automatically.

Now we're talking, nice :good: Git er dun!

-150c unloaded :eek:

I really like that -125c loaded Core I7 temp tho ;)
With a setup like that you really would not need LN2 for 90% of benching, except for newer AMD chips that it. But very useful for getting everything setup and tuned before going to LN2 :)

And for that you could still dial her down colder for lighter loads. Not much use past quick Pi and such, but still not bad ;) The mass I'm delivering in the evaporator should be interesting.
Let's just say it competes with many top end ln2 pots.

Though finding the condenser I want is starting to become harder and harder.
Might have to shell out some serious cash to someone in Euroland.

I dream of the day I have a cascade setup. :this:

They're pricey, but if you bench Ln2 on CPU only 20 times, you've paid more.

Not with the rates I pay for LN2 :D ($0.50/Liter)

So $25 for 50L.
$300 for a CPU pot.
$400 for a dewar.
How many runs to hit $1600? ;)
28?

I prolly have just about paid for my Cascade in teh amount of use i get out of it, almost everyday its running which I could not do with LN2. I save that for chips that can really use it and for GPU's when I run them.

To say nothing about ease of use. You can walk away and do others things when the MonstaCade is running and not have to worry about the pot running dry. Being able to take your time in tweaking is also a big plus.

It has not taken away my need for LN2 to tho, just made things much easyer.

How ever, colder is better :)

Exactly. I hate rushed tweaking.
Here it costs you no real cash but electricity to play around.

Update for you later boys. Going to start cutting and measuring steel, and pressure testing the heat exchangers, AND creating the big massive pink box of doom!

Pink time!
Those CPEV's are looking beastley aren't they?


http://img33.imageshack.us/img33/5800/0708092205.jpg

Kool, what do they do?

CPEV's allow adjustment of the suction pressure via the screw on the front.
In short, as suction pressure changes, so does evaporation temperature. Thus I can control the evaporation temperature of EACH stage!

Left loosy, righty tighty? Which way is colder/hotter?

Frick if I know. Probably righty tighty brings the screw into the body, more pressure on the screw, and pressure goes up? Pretty easy to test when its running. Normally they're set at 5-10psi to start. Even if its set at max of 55psi low side, it won't make a difference.

I think I went a bit overboard height wise. Don't need that much! (Yay the cascade just got shorter!) I think I'll shave 2 inches off of her. Having the "low" HX in the center helps with the insulation of it. Normally -80 to -90C would require 3" of foam.

http://img23.imageshack.us/img23/874/0708092235.jpg

ROFL :rofl

I'll stop buggin u now and leave u to your work, lol. :morpheus:

Hahaha bug away, I think I'm almost done for the night. Cutting down that foam. Add some tube insulation inside the box, some tapering nails to hold it together, then break out some duct tape for an air tight seal. Maybe autocad the cage up. Keep your eyes peeled!

5 of 6 gauge kits.
2 ft of 1/8" tube on each, after brazing in an inch and a half of 1/4" tube to a tube adapator.
Will find that last bit or just pickup another while I'm at homedepot getting JB Weld for these joints. I don't like teflon tape, it's just not robust enough for me.

http://img189.imageshack.us/img189/3709/0708092343.jpg