3-stager unit "Grosse Bertha"

[before]

This project started few months ago. My aim was to turn my late 2-stager to a 3-stager.

Specifications:

1st stage
LG NP407P (25000Btu/h)
AC Condenser + Fan
Copper filter dryer
Captube .8mm (1.9m)
HX 14 plates
R404a

2nd stage
Matsuchita (sic) 2K32 (19,000Btu/h)
Manual Oil separator
REFCO manometer for HSP
Expansion tank
Desuperheater (huge Condenser split into 2 parts, one for 2nd stage and one for 3rd stage)
Johnson Controls Pressostat
2 Copper filter dryers (one before and another after 1st stage HX)
Captube .8mm (2.05m)
Coaxial HX
R1150 or R290/R50

3rd stage
Sanyo C-RV167H11AB (10,000Btu/h)
ShineYear Oil separator
REFCO manometer for HSP
Expansion tank
Desuperheater (huge Condenser split into 2 parts, one for 2nd stage and one for 3rd stage)
Johnson Controls Pressostat
Filter dryer
Captube .8mm (2.55m) passing into the succion line
Pgeo's "Marguerite" evaporator
R14/R290 or R14/R290/R50

I'd like to thx my friend Luigi for all he has done helping me for this project. Also, Xialis lent me a helpful hand...

Some results

1st Stage R404a @-46C
2nd Stage R1150 @-103C too cold IMO
3rd Stage R14/R290 @-130C unloaded

Cold-bugged X6800 at 5.2GHz @1.8v



1st Stage R404a @-41C
2nd Stage R290/R50 @-75C too "hot" IMO
3rd Stage R14/R290/R50 @-137C unloaded



Some pics: more to come





I'm currently waiting for the 23400Btu/h compressor in order to replace 1st stage compressor and finish the cascade.

[Gautam]

Bravo Xavier!

Even for an unloaded temp that is incredible...this unit might clock almost on an even plane with LN2 for some chips.

[gosmeyer]

Nice run on that chip. Any temps for third stage loaded?

[toaster]

nice unit !

[p8ntslinger676]

wow!

[Polizei]

Impressive temps.

[blind_ripper]

seen the build on OCM befor very nice cascade

[n00b 0f l337]

Bravo!

[before]

Thank you all

Bravo Xavier!

Even for an unloaded temp that is incredible...this unit might clock almost on an even plane with LN2 for some chips.

Thx bro My X6800 can't run under this cascade... As can be seen in the screenshot, it cold-bugs at -120C speedfan even with 5.2GHz and 1.8v...

Nice run on that chip. Any temps for third stage loaded?

Thx

Too early to make conlusions; about -125C to -130C with the X6800 at 5.2GHz @1.8v but because of cold-bug I can't try it really loaded during a Spi 32M for instance.

I need a good chip in order to perform loaded tests. Also, I'll replace 1st stage compressor by a huge one...

[ent]

Great work before! This unit will probally be as good as Ln2 for some CPUs and benching with a cascade is much easier than benching Ln2!

Put 2V in those coldbugged chips, maybe this will help them

[blind_ripper]

Great work before! This unit will probally be as good as Ln2 for some CPUs and benching with a cascade is much easier than benching Ln2!

Put 2V in those coldbugged chips, maybe this will help them

hehe cascade will never match LN²
the idle load temps one cascade is not the same like LN² , with LN² u can pump up the load temps

[Freddie123]

But having LN2 requires organization and an element of potential frost burn. A cascade you can hook up anytime without organizing LN2 and with less chance of frostburn :P Ricky managed -170 with his cascade remember, not far off LN2, if he'd have had pure methane it would have been better.

[LukeXE]

Very nice my friend.

I`m curious like Tom,so I`m waiting for answer for those questions.

Really R290/R50 blend on ST2 ?

[before]

Great work before! This unit will probally be as good as Ln2 for some CPUs and benching with a cascade is much easier than benching Ln2!

Put 2V in those coldbugged chips, maybe this will help them

Unfortunately, this chip tops at about 1.80v to 1.84v.

But having LN2 requires organization and an element of potential frost burn. A cascade you can hook up anytime without organizing LN2 and with less chance of frostburn :P Ricky managed -170 with his cascade remember, not far off LN2, if he'd have had pure methane it would have been better.

Yep It'll not be an LN² replacement, but for some chips LN² will not produce significant improvement IMO.

Very nice before

Can you confirm that in the second case you're using only propane and methane on your second stage and that it isn't a typo?

What ratio are you using propane and methane in? Something like "i charged to 50PSI propane then increased to 200PSI with methane" would be great.

What discharge pressure do you have on your second stage and what is the second stage liquid temperature at this pressure?

I thought that perhaps propane and methane could possibly form a useable refrigerant mixture but didn't know anywhere near enough about chemistry to proove it, to be honest i didn't think they'd work well. But you might have just prooved that they do - the best way

If you have access to ethane, i think that ethane with methane could possibly work better than propane with methane at the temperatures you seek...

Tom

Very nice my friend.

I`m curious like Tom,so I`m waiting for answer for those questions.

Really R290/R50 blend on ST2 ?

Thx to both of you

Yea, currently, 2nd stage refrigerant is a mixture of R290 and R50: about 11bars of R290 and 4bars of R50. I'm not sure about exact ratio, but that's close to actual blend.

TBH, I've removed R1150 from this stage because of a small leak suspicion. Then, I've filled in with R290 and R50 to make a try.

I think R1150 is too cold for this stage; with the HX @-103C, R14 of 3rd stage is too easy to condensate. That's why I've tried to get something into the -80C range to properly use R14 at the lowest stage. Actually, R290/R50 mixture of 2nd stage isn't cold enough to hold load (at least, that's what I imagine); but that's a matter of 5C, not more.

Unloaded HX temp of 2nd stage is about -80C with R290/R50 and it goes up to -75C with 3rd stage turned on (LSP is about 0.2Bar; need to confirm this). That's quite impressive IMO. Now, I'm impatiently waiting for the huge compressor to use into 1st stage; it might be the key for my particular need... BTW to use this exotic mixture, 1st and 2nd stages will run respectivily a 23,400Btu/h and a 17,000Btu/h rotary compressor. Also, I'm thinking about making something like an SLHX for R404a of 1st stage, it may helps a bit (AFAIK, heat exchange into SL gives good results with R404a http://www.irc.wisc.edu/file.php?id=49)

I may get some R170 if needed

[n00b 0f l337]

Beauty of hydrocarbons, they mix so well
Nice job Before, if you really have a cold bug trouble, time to make some shims
Pickup a 1.5" hole saw and some 1/8" thick copper, cut a few discs out, lap them to be perfectly flat, add thermal paste and try them between your evap and your cpu.

[Pete]

Beauty of hydrocarbons, they mix so well
Nice job Before, if you really have a cold bug trouble, time to make some shims
Pickup a 1.5" hole saw and some 1/8" thick copper, cut a few discs out, lap them to be perfectly flat, add thermal paste and try them between your evap and your cpu.

Odd but intresting idea

[bazx]

great work before

and i know you will put it to good work

baz

[one_servant]

Great job there! Keep us posted

[kayl]

very nice work before

[n00b 0f l337]

Pete its what a bunch of people did for the amd cold bugs

[gosmeyer]

Thicker base, Warmer chip. Go figure.

[n00b 0f l337]

Yep yep!

[[486]]

you would need to clamp it down and use a drill press, otherwise you would need to use a holesaw with a drill bit in the middle, because of degraded stability, but you could use a short piece of 1/4 inch rod as a arbor though, some hole saws require a drill bit to hold the assembly in the chuck

edited for clarity

[n00b 0f l337]

Yep that is true, or just remove the drill bit and try and not make it slip.

[Blaster]

Thicker base, Warmer chip. Go figure.

Thicker base, Warmer chip, Evap colder. Go figure.









(you forgot this)