The PhaseChiller heat exchanger project...

[J7SC_Orion]

 For a while now, I have been staring at boxes full of old water-cooling blocks, as well as 1Kw phase change cooler and other assorted sub-ambient and sub-zero bits from almost a decade ago. At the same time, I (finally) finished the (4x mobo) Ravens Nests builds so I figure a new project is in order, even if it will be in slow-motion due to other time commitments. 

 

I knew there must have been another reason I used all those QD4s in the Ravens beyond just convenience... 

 

 

 

Specifically, I have two 'dual-GPU' blocks for Radeon 8990s  laying around I haven' used in a long time. Per pic below, they are absolutely massive and first need a thorough cleaning inside and out. But given their cooling and humangous copper surface areas (combined, they helped to easily dispose of 1200W of heat energy before), I am thinking of utilizing the Ravens' QD4s to hook in a loop with both of these blocks  in a single circuit (for bench days only). These old blocks would be submerged and stand parallelly and vertically in a 72L plastic water basin, and in between the copper ends of those two blocks would be the submerged copper tip of the phase cooler that goes to around -50 C in 'normal operation'.

 

I have no intention  of actually going much below the dew point, but Raven A's two setups can each easily slurp 900 W (or more, depending on the vbios for the 6900XT and 3090 Strix, respectively). 72 L of water - or even more / less if need be - should be plenty enough to help 'dampen' transient temp swings to the point that it should be manageable, while the sheer copper surface area of the old 8990 blocks would be great for 'heat exchanging'. If need be, I can add a little 110V aquarium pump into the 72L basin to move the 'ice' water around / past whatever components I want.  Obviously, I will try this out on some older setups first (dual 980 Classifieds w/EVBot ~ 800W combined) to see what the base data and transient deltas look like before I hook this into either the 3950X/6900XT loop or the 5950X/3090 Strix loop.

 

What do you think about this contraption ? Anything obvious jumping out at you I should watch out for ?

 

 

 

 

Edited February 24, 2022 by J7SC_Orion

[Avacado]

9 hours ago, J7SC_Orion said:

 For a while now, I have been staring at boxes full of old water-cooling blocks, as well as 1Kw phase change cooler and other assorted sub-ambient and sub-zero bits from almost a decade ago. At the same time, I (finally) finished the (4x mobo) Ravens Nests builds so I figure a new project is in order, even if it will be in slow-motion due to other time commitments. 

 

I knew there must have been another reason I used all those QD4s in the Ravens beyond just convenience... 

 

 

 

Specifically, I have two 'dual-GPU' blocks for Radeon 8990s  laying around I haven' used in a long time. Per pic below, they are absolutely massive and first need a thorough cleaning inside and out. But given their cooling and humangous copper surface areas (combined, they helped to easily dispose of 1200W of heat energy before), I am thinking of utilizing the Ravens' QD4s to hook in a loop with both of these blocks  in a single circuit (for bench days only). These old blocks would be submerged and stand parallelly and vertically in a 72L plastic water basin, and in between the copper ends of those two blocks would be the submerged copper tip of the phase cooler that goes to around -50 C in 'normal operation'.

 

I have no intention  of actually going much below the dew point, but Raven A's two setups can each easily slurp 900 W (or more, depending on the vbios for the 6900XT and 3090 Strix, respectively). 72 L of water - or even more / less if need be - should be plenty enough to help 'dampen' transient temp swings to the point that it should be manageable, while the sheer copper surface area of the old 8990 blocks would be great for 'heat exchanging'. If need be, I can add a little 110V aquarium pump into the 72L basin to move the 'ice' water around / past whatever components I want.  Obviously, I will try this out on some older setups first (dual 980 Classifieds w/EVBot ~ 800W combined) to see what the base data and transient deltas look like before I hook this into either the 3950X/6900XT loop or the 5950X/3090 Strix loop.

 

What do you think about this contraption ? Anything obvious jumping out at you I should watch out for ?

 

 

 

 

Watch out for being anointed into the awesomsauce HOF. Sounds really cool. This is the kind of * I want to do when I retire. 

[J7SC_Orion]

3 hours ago, Avacado said:

Watch out for being anointed into the awesomsauce HOF. Sounds really cool. This is the kind of * I want to do when I retire. 

 

...yeah, I have no idea if this will actually work, but theoretically 'it should', and I have all the individual components anyway, so apart from my time, it is 'free'...calibrating the thing will probably be the biggest challenge.

[bonami2]

Send us some video! I would love playing with setup like this. Gonna try a ice bucket with a 360mm to play with some old am3 cpu poor man setup

 

Out of topic but did you try hdr on your lg oled? Just tried it on 60hz 4k and it run youtube in hdr but doest seem to work compared to my ips 4k i had before.

Edited February 25, 2022 by bonami2

[J7SC_Orion]

1 hour ago, bonami2 said:

Send us some video! I would love playing with setup like this. Gonna try a ice bucket with a 360mm to play with some old am3 cpu poor man setup

 

Out of topic but did you try hdr on your lg oled? Just tried it on 60hz 4k and it run youtube in hdr but doest seem to work compared to my ips 4k i had before.

 

Once I get around to 'building' the contraption, I'll take pics and vids. BTW, I've done the rad in the ice bucket back in the day, then added a few pounds of DICE - worked for a short while, but lots of commotion and  'smoke', and the poor rad getting beat up by dancing frozen bits  Nothing a coat of flat black Rustoleum didn't fix, though.

 

HDR is working great on my C1 OLED. 

 

 

[Storm-Chaser]

On 25/02/2022 at 03:07, J7SC_Orion said:

and the poor rad getting beat up by dancing frozen bits  Nothing a coat of flat black Rustoleum didn't fix, though.

If you replace the rad in the bucket with a plate heat exchanger it will be able to handle plenty of abuse at the hands of DICE. I think my 60 plate unit is good to about 500 PSI. It's a pressure vessel after all.

 

So let me see here, Im trying to follow, are you saying you are going to immerse these blocks in the ice cold water? So does that mean you are effectively using them as a heat exchanger for something down the line? Just trying to get a visualization on what you are doing here. 

 

Good call on the 72L water tower. With that you should be able to bench indefinitely. 

[ArchStanton]

1 hour ago, Storm-Chaser said:

good to about 500 PSI.

435.113 

 

(I know, I know...forgive me, but I exact figures lol)

[Storm-Chaser]

33 minutes ago, ArchStanton said:

435.113 

 

(I know, I know...forgive me, but I exact figures lol)

 

33 minutes ago, ArchStanton said:

435.113 

 

(I know, I know...forgive me, but I exact figures lol)

Then where is your fourth digit?? Lol

 

But in all seriousness, you can see from the picture the design pressure is 3.5KPA which when converted to PSI comes to 507.

 

Edited April 7, 2022 by Storm-Chaser

[ArchStanton]

22 minutes ago, Storm-Chaser said:

you can see

Apparently I cannot .  I thought it said it was only 3Mpa .  Well .

[Storm-Chaser]

24 minutes ago, ArchStanton said:

Apparently I cannot .  I thought it said it was only 3Mpa .  Well .

It's okay. storm-chaser gets lost once in a while. 

Edited April 7, 2022 by Storm-Chaser

[Storm-Chaser]

23 minutes ago, ArchStanton said:

Apparently I cannot .  I thought it said it was only 3Mpa .  Well .

Apparently I'm the one who has lost his mind lol. swear I was seeing 3.5 for some reason... but no you were right the whole time.

[ArchStanton]

3 minutes ago, Storm-Chaser said:

Apparently I'm the one who has lost his mind lol. swear I was seeing 3.5 for some reason... but no you were right the whole time.

, Oh man,  I made a mountain out of a mole hill (was just being goofy), and now we have both fallen of the mountain

[Storm-Chaser]

6 hours ago, ArchStanton said:

, Oh man,  I made a mountain out of a mole hill (was just being goofy), and now we have both fallen of the mountain

[Avacado]

8 hours ago, ArchStanton said:

Apparently I cannot .  I thought it said it was only 3Mpa .  Well .

I love a person who is able to own it and poke fun at themselves! +10 points to you sir. 

[J7SC_Orion]

16 hours ago, Storm-Chaser said:

If you replace the rad in the bucket with a plate heat exchanger it will be able to handle plenty of abuse at the hands of DICE. I think my 60 plate unit is good to about 500 PSI. It's a pressure vessel after all.

 

So let me see here, Im trying to follow, are you saying you are going to immerse these blocks in the ice cold water? So does that mean you are effectively using them as a heat exchanger for something down the line? Just trying to get a visualization on what you are doing here. 

 

Good call on the 72L water tower. With that you should be able to bench indefinitely. 

 

...I have been checking Amazon for an additional heat exchanger...you can't just dump the copper head of the chiller into water (tried that years ago) because it will just form a clump of ice around the head which acts as insulation...

 

For now, in addition to the items depicted in the first post, I have been using 2x BlackIce 180x60+ high-fpi full-brass rads I have...the idea of this 'on the back burner project' (pardon the pun) is to use existing bits that are just laying around. Still, and inexpensive heat exchanger or two which I can flange to the chiller head is on the menu...   

[Storm-Chaser]

On 07/04/2022 at 13:33, J7SC_Orion said:

 

...I have been checking Amazon for an additional heat exchanger...you can't just dump the copper head of the chiller into water (tried that years ago) because it will just form a clump of ice around the head which acts as insulation...

 

For now, in addition to the items depicted in the first post, I have been using 2x BlackIce 180x60+ high-fpi full-brass rads I have...the idea of this 'on the back burner project' (pardon the pun) is to use existing bits that are just laying around. Still, and inexpensive heat exchanger or two which I can flange to the chiller head is on the menu...   

If you run methanol in the bucket, you can get around the icing up issue. Unless you are going lower than -25*F, then you would have to increase the methanol content. 

 

How is the project going, any progress?

[J7SC_Orion]

3 hours ago, Storm-Chaser said:

If you run methanol in the bucket, you can get around the icing up issue. Unless you are going lower than -25*F, then you would have to increase the methanol content. 

 

How is the project going, any progress?

 

...I have briefly tried w/ 99.9% isopropanol in the config w/o home-made heat exchanger for some GPU fun, and it seems to work...just trying to figure out how o-rings etc react before I change the config...still wanting to get a real heat exchanger from Amazon but it is a low-priority project for now. I also plan to use the isopropanol as a medium in the LN2 pot and the phase as the source of subzero in the pot, and compare it to a DICE run...

[Avacado]

1 hour ago, J7SC_Orion said:

 

...I have briefly tried w/ 99.9% isopropanol in the config w/o home-made heat exchanger for some GPU fun, and it seems to work...just trying to figure out how o-rings etc react before I change the config...still wanting to get a real heat exchanger from Amazon but it is a low-priority project for now. I also plan to use the isopropanol as a medium in the LN2 pot and the phase as the source of subzero in the pot, and compare it to a DICE run...

Video or it didn't happen. 

[J7SC_Orion]

3 hours ago, Avacado said:

Video or it didn't happen. 

 

...oddly enough, it happened w/o video as I had my hands full 

 

...but when I get a real heat exchanger added in (to avoid running isopropanol or acetone directly through the GPU blocks, O-rings etc), I'll take some pics and may be do a vid. 

[Bastiaan_NL]

On 28/04/2022 at 00:05, J7SC_Orion said:

 

...oddly enough, it happened w/o video as I had my hands full 

 

...but when I get a real heat exchanger added in (to avoid running isopropanol or acetone directly through the GPU blocks, O-rings etc), I'll take some pics and may be do a vid. 

+1 for a video!

[Storm-Chaser]

@J7SC_Orion

 

You could always try the vortex cooler if you don't mind sounding on par with a wide open throttle chainsaw. Maybe wireless mouse and keyboard and put the machine in another room? 

 

 

 

 

This one pictured is good for about 800 watts cooling performance. Yes you heard me correctly...

Edited May 4, 2022 by Storm-Chaser

[ArchStanton]

@Storm-Chaserany connection between the device above and the GM "Vortec" cylinder heads?

[Storm-Chaser]

They both contribute to turbulent air, but I doubt there is much connection beyond that.

 

Here is a crash course.

 

Part 1: Gary Talks about the Vortec combustion chamber design.

"Here is the short version of the importance of combustion chamber design. First some assumptions, air fuel ratio is right, meaning enough present oxegen to support total combustion of the fuel. Second, forget about the "crevice", the area above the top ring between piston and cylinder wall where the fuel will not burn efficiently. Third, assume flat top piston, best design, promotes most rapid flame travel. Now compare a theoretical combustion chamber with piston at tdc and no air movement. This of coarse would never happen, but bare with me. To burn the fuel would require maybe 60 degrees of timing lead, because you have only the close proximity of the fuel atoms to spread the burn from the plug accross the area of the chamber. This would mean your engine would "fight" against building combustion pressure as piston approached tdc. Very inefficient. "old fashioned" combustion chambers are more desireable than our theoretical case because they induce turbulence. The turbulence speeds the spread of the burn, but in a random and uncontrollable fashion. Requireing maybe 34-38 degrees ignition timing lead. Better, but still not great. This is all heads built prior to the early 90's. And, most aftermarket heads, which are "tweaked" copies of "old fashioned" production heads. Now consider the Vortec head with its predictable and repeatable swirll patterns. Faster burn rate than our previous examples, requireing even less timing lead, about 30-32 degrees, therefore engine has to fight the mounting pressure less than the previous examples. This means more power available to put to the ground, or up in smoke I guess. The theoretical ideal engine would have such a fast burn rate that total combustion would happen at the moment of ignition. Then you could run 0 or 1 degree timing lead. Don't hold your breath waiting for this to be reality, the laws of physics aren't cooperating with us on that one. Make sense to ya?
Regarding port design, "total area under the curve" is most important. Most buy heads based on peak claimed flow at some valve lift (ie. 250 cfm at .600" lift.) But more important than that is flow at all levels and particularly low lift flow. If you chart your flow at all valve lifts from open to close it looks like a bell graph (the line will be shaped like a bell). The fatter the bell the better. A narrow pointy bell is no good. Also, you need efficient flow management, that is laminar flow without turbulence in the ports. Turbulence creates "voids" or low pressure pockets that mean your port will not pass as much air as the same port witout turbulence would. Laminar flow takes the air through the port intact, essentially fills the port with quality air without "voids". Another thing to consider is velocity and momentum. Air and fuel have mass, so they can be accelerated and will have momentum. The Vortec head uses a venturi built into the port shape to accelerate the air/fuel after it makes the turn toward the valve. This creates low pressure behind it, drawing more air/fuel. And it gives the air fuel high velocity and momentum so that it has a lot of energy when it hits the back side of the open valve, because at this point it has to be redirected and loses energy as it's redirected.
So, bottom line is best high velocity laminar flow at all valve lifts with a controlled high swirll chamber, combined with a flat top piston. Currently, no aftermarket head can begin to compare with the Vortec, even the Fast Burn lags a little behind in the quality of these characteristics."

[J7SC_Orion]

52 minutes ago, Storm-Chaser said:

They both contribute to turbulent air, but I doubt there is much connection beyond that.

 

Here is a crash course.

 

Part 1: Gary Talks about the Vortec combustion chamber design.

"Here is the short version of the importance of combustion chamber design. First some assumptions, air fuel ratio is right, meaning enough present oxegen to support total combustion of the fuel. Second, forget about the "crevice", the area above the top ring between piston and cylinder wall where the fuel will not burn efficiently. Third, assume flat top piston, best design, promotes most rapid flame travel. Now compare a theoretical combustion chamber with piston at tdc and no air movement. This of coarse would never happen, but bare with me. To burn the fuel would require maybe 60 degrees of timing lead, because you have only the close proximity of the fuel atoms to spread the burn from the plug accross the area of the chamber. This would mean your engine would "fight" against building combustion pressure as piston approached tdc. Very inefficient. "old fashioned" combustion chambers are more desireable than our theoretical case because they induce turbulence. The turbulence speeds the spread of the burn, but in a random and uncontrollable fashion. Requireing maybe 34-38 degrees ignition timing lead. Better, but still not great. This is all heads built prior to the early 90's. And, most aftermarket heads, which are "tweaked" copies of "old fashioned" production heads. Now consider the Vortec head with its predictable and repeatable swirll patterns. Faster burn rate than our previous examples, requireing even less timing lead, about 30-32 degrees, therefore engine has to fight the mounting pressure less than the previous examples. This means more power available to put to the ground, or up in smoke I guess. The theoretical ideal engine would have such a fast burn rate that total combustion would happen at the moment of ignition. Then you could run 0 or 1 degree timing lead. Don't hold your breath waiting for this to be reality, the laws of physics aren't cooperating with us on that one. Make sense to ya?
Regarding port design, "total area under the curve" is most important. Most buy heads based on peak claimed flow at some valve lift (ie. 250 cfm at .600" lift.) But more important than that is flow at all levels and particularly low lift flow. If you chart your flow at all valve lifts from open to close it looks like a bell graph (the line will be shaped like a bell). The fatter the bell the better. A narrow pointy bell is no good. Also, you need efficient flow management, that is laminar flow without turbulence in the ports. Turbulence creates "voids" or low pressure pockets that mean your port will not pass as much air as the same port witout turbulence would. Laminar flow takes the air through the port intact, essentially fills the port with quality air without "voids". Another thing to consider is velocity and momentum. Air and fuel have mass, so they can be accelerated and will have momentum. The Vortec head uses a venturi built into the port shape to accelerate the air/fuel after it makes the turn toward the valve. This creates low pressure behind it, drawing more air/fuel. And it gives the air fuel high velocity and momentum so that it has a lot of energy when it hits the back side of the open valve, because at this point it has to be redirected and loses energy as it's redirected.
So, bottom line is best high velocity laminar flow at all valve lifts with a controlled high swirll chamber, combined with a flat top piston. Currently, no aftermarket head can begin to compare with the Vortec, even the Fast Burn lags a little behind in the quality of these characteristics."

 

...Linus has a related vid:

 

 

[Storm-Chaser]

5 minutes ago, J7SC_Orion said:

 

...Linus has a related vid:

 

 

Yup that's where I discovered it.