2-Stager unit

[jinu117]

but it might be interesting to add some 0.9mm captude.

Uhmmm BEFORE or after? (I couldn't help it... :P)

[before]

hehe

between actual 2mm captube and evap

[before]

It runs quite good

I got this E6850 up to 540x9 @1.72v (max FSB is lower than 550) and the temp didn't really move... in fact, it was a bit colder at the end thx to a lower ambient.













I'm gonna move for a better chip in order to achieve higher loads.

[4Qman]

great pics before. Look forward to a betetr chip under there man.

[before]

Load is well held; evap pressure stays under 1bar.

[Marvin]

Congrats Xavier.
what are the dimension of the suction line ID and OD ?
and the SL insulation tube wall and ID ?
sorry for the questions ,but seem you dont have condensation.

[before]

I've switched to R404a and temps are now quite stables. BTW they are very very cold; I mean HX inlet temp is about -70C loaded!!! Less than 0.3bar...

I don't have exact HX outlet temp in mind, but delta between both is a bit high in my opinion.

Anyway, amount of refrigerant is good... I mean, adding more doesn't increase HSP but only static's.

IMHO, that's not good.

Additionnaly, I've added 3m of 0.9mm captube to the 1.5m of 2mm captube already installed between CPEV and evap in 2nd stage. Pressure has significantly decreased... from 0.6bar (-112C) with CPEV completely shut to 0.25bar (-124C) with CPEV completely open.

Taking speedfan as a temperature measurement, CPU is now clearly colder:
E6850 at 5100MHz @1.73v
Before: -92C idle / -75C SuperPI 32M
After: -106C idle / -90C SuperPI 32M !! (similar to my 3-stager temps with R1150 into 3rd stage)

BTW, I'll remove a bit of captube to properly use the CPEV...

Congrats Xavier.
what are the dimension of the suction line ID and OD ?
and the SL insulation tube wall and ID ?
sorry for the questions ,but seem you dont have condensation.

Thanks Marvin

Succion lines are made of 14/16mm copper pipes to perfectly fit compressors succion connection.

What do you mean by "tube wall"? Insulation thickness? I think it's 13mm.

[gosmeyer]

Nice, Way to tweek it.

[before]

Thank you

I've removed 1.9m of 2nd stage 0.9mm captube.

Temp is now -109C with CPEV fully opened... Obviously, closing it decreases the temp.

[before]

It has been a long time since I've last updated this thread...

So here are some results from what I did recently.

-118C unloaded

-112C Super PI 32M with QX9650 5.49GHz @1.90v



-107.5C 3DMark 2005 CPU tests with QX9650 5.46GHz @1.944v (OMG, these tests are strong! )



So; as you can see, it allows some good results.

BTW I still have a problem with flaring the oil return captube. It looks to leak a bit everytimes I unmount and then remount it. Will probably braze a simple captube here; but it pisses me off to have bought a special captube with flaring nuts and not using it.

Anyway; what you guys honestly think about these temperatures? Those are pretty cold, that's true, but can I expect less?

I'm not sure about the good functioning of the second stage CPEV... Maybe I'll remove it and braze a captube there.

[before]

I've removed the CPEV from 2nd stage, and try to get a good amount of R1150 in order to achieve some decent temperatures. (CPEV is now used as a 1st stage expansion metric for my 3-stager)

First tries weren't very good cause without CPEV I've lost about 10C+, pressure for pressure, load for load.

So, I've though that it was time to remove the remaining captube which was used along with the CPEV (~2m of 2mm + ~2.9m of 0.9mm). Was thinking about 3.5 to 3.6m of 0.9mm as a replacement.

But today morning, before changing the captube, I've decided to go for a bit less of R1150... and I finally got pretty nice loaded temperatures... at about 5.6GHz using 1.97vCORE for Spi 32M, top/'hotter' part of the evaporator shown -103.5C with an evap pressure of 0.6bar which means -112C.

Well, that's nearly the same than it was with the CPEV installed, that's damn good considering the load.

Click the thumbnail to view full-size image

[gosmeyer]

I finally got pretty nice loaded temperatures... at about 5.6GHz using 1.97vCORE for Spi 32M, top/'hotter' part of the evaporator shown -103.5C with an evap pressure of 0.6bar which means -112C.

I think "pretty nice" is an understatement

[n00b 0f l337]

0.6 bar? As in above atmospheric? Shouldnt that be warmer then the BP at 1 atm no?

[boshuter]

Awesome job before

Nice to finally see a cascade tested with a real load

[before]

Thanks guys

0.6 bar? As in above atmospheric? Shouldnt that be warmer then the BP at 1 atm no?

Not sure if I've clrearly understood what you said.

Here's what I'm talking about.



As for the observations, -112C evaporating temperature (0.6bar) to -103.5C evaporator top (K-type probe) is quite consistent. I mean, exit of the evaporator has to be hotter than evaporating temperature... obviously. About 10C more is not that much for actual load IMHO, isn't it?

[sacha35]

Nice results there before, now all you need is that golden 6Gig chip to put under this cooling, but even so the 5.6Gig you are getting is outstanding and it shows this unit holds good temps with 1.97 on the core when running SP, can you run a Wrime 1024 and show us some loaded temps please.

Well done mate outstanding.

[Clemmaster]

About 10C more is not that much for actual load IMHO, isn't it?

No you don't have to worry about it, unless it has effect on the temperature of the base of the evap

[Gautam]

5600 on a 2 stage for 32M!

[before]

Thanks guys

I've decided to change the 2nd stage captube...yeah I know, I've to stop myself making mods everytime

In fact, taking into account the lenght of 0.9mm captube used by two XS builders for their cascades, I've performed a quick extrapolation to get a correspondance table for captube adjustment relativily to compressor power (rotary R407C compressor). Dunno if it makes sense or not...

BTW, I've brazed a piece of 3.8m. Here are the first results.



Windows idle: -114.1C evap ; -62.4C HX in



Loading Spi 32M at 5535MHz (582.6FSBx9.5) @1.82v: -108.2C evap ;-56.1C HX out; HP = 7bars

[LittleDevil]

Great Can i ask you what is your room temp?

Regards

[n00b 0f l337]

That really can't be good for the first stage compressor

[before]

Great Can i ask you what is your room temp?

Regards

Thanks my friend

Temp was about 18C to 20C I guess. (Outside temp was 24C and the cascade is next to the garage main door which was wide opened)

That really can't be good for the first stage compressor

Right. I've to adjust the amount of R1150 and then tune the first stage TXV to match the need.

[before]

Adjusting R1150 amount, and retuning TXV.

Unloaded: -118.0C evap (-57.1C HX out)



-118C ins't the coldest; but the pic I've taken. Guess -120C or so is highly probable with a bit more time. Ambient temp is 20C.

Also; I'm not looking for the absolute unloaded/unuseable coldest temp to take a pic, but for the right balance between 1st and 2nd stages.

[sacha35]

Looking good before, is see you have posted up a chart for cap lengths, would this not change though on the amount of load required to sustain the given temperatures or was these calculations given on unloaded temperatures.

[before]

Thanks

Of course the lenght has to be adjusted upon the need. BTW I've checked the properties of 2 cascades that shown very good held of load (IIRC 250W+
@ about -107C). Obviously, unloaded temp and loaded one are highly correlated.

Also; this chart doesn't stand to be an 'end of all charts' you know (1st stage is so important as well even if the 1st compressor is also highly predictible knowing the 2nd one), but an approximate table gathering values one might expect to use with R1150 relativily to the 2nd stage compressor power. Assuming a linear relationship (which is pure guess), I would say the lenght is to be taken + or - 0.10m or something.