Exploring Below Ambient Water Cooling

[Motopsychojdn]

I'll still dump the copper in here to tweak the watermass so that my working Pc is at least at ambient,
but I am going to seriously investigate and start work on the other plan as well, that one is more desk friendly hehe
Moto

 

[4Ryan6]

I'll still dump the copper in here to tweak the watermass so that my working Pc is at least at ambient,
but I am going to seriously investigate and start work on the other plan as well, that one is more desk friendly hehe
Moto

Don't wait too long or I'll do it myself!

 

[Motopsychojdn]

Replied to Pm
Off for sleepzies now, work tonight, have fun mate
Moto

 

[4Ryan6]

Just letting you know I ordered what I need to conduct the experiment Moto!

I'll post whatever happens right here one way or the other, good results or bad. Ry

 

[dylanc0100]

wow this is hectic but would love to do it

 

[Motopsychojdn]

@Ryan, nice one! Can't wait to see your results, I'm still hung waiting to see if I have spare money this month yet
@Dylan, don't wish, do it
Moto

 

[4Ryan6]

I'm taking detailed photos but the only way you'll get to see them is if it is a successful experiment, if it is not successful I'll show one picture of the finished cooler so I can claim my idiot award!

 

[Lutfij]

:lol: we're all for the good of the people. Galileo was branded a lunatic for thinking the world was round so if you'd call yourself a big idiot(via that award), which isle would that put us in ?

 

[Motopsychojdn]

Hehe, this is a hybrid adaptation of both mine and Ryans methods,
I won't say anymore until something happens,
but it promises to be good
Moto

 

[4Ryan6]

First leak test after the modifications passed, thankfully!

So far it's all coming together I've run into a couple of minor roadblocks, still to overcome, but the finished look is turning out very nice.

Once it's all together I'll start testing it, one of my main concerns, is the heat sink that will be cooling the hot side of the peltier sufficient??????

Hope So!

That's the first test and if that test fails, back to the drawing board!

That will only be a speed bump though, because I have other heat sinks

 

[Motopsychojdn]

Feel free to Pm if you'd rather not discuss it here mate, but I'm sure it will be fine, look at your current results, the true does what it should yes?
I'm with you in spirit and holding my breath
Moto

 

[4Ryan6]

I used Marine Grade Clear silicone to bond the acrylic adapter plate to the acrylic reservoir and it takes a full undisturbed 48hrs to completely cure before water contact so this will not be continuing at all, until after 4:00 AM Monday Morning my time.

 

[Motopsychojdn]

No rush man, I'm eager to see results, but not at the cost of success
Moto

 

[4Ryan6]

Easy for you to say but I'm chomping at the bits to see this thing in operation!

 

[Dark Lord of Tech]

Can't wait to see your results,Ry.

 

[4Ryan6]

I'm taking detailed photos but the only way you'll get to see them is if it is a successful experiment, if it is not successful I'll show one picture of the finished cooler so I can claim my idiot award!

Well this post is a disappointment to have to make, I promised a picture but to fully understand what I created and experimented with, here's three views.

Any of you that have followed this thread the best way to sum up why this was not successful was it produced the same results as the previous massive copper water block did.

Remember this massive monster!

Using the modified dual bay reservoir with the copper fins submerged in the water could not maintain a thermal cold load under testing load and once the water heated in the reservoir, it took too long to recover.

End results of radiators cooling the GPUs, Idle = 32c, Load = 40c.

End results of Copper Heatsink over Copper Heatsink in acrylic reservoir, Idle = 32c, Load = 52c,

12c performance cooling loss, so this particular project is, "Not Da Mama!"

This confirms, as far as I am concerned, that using a traditional CPU water block to pick up the cold from the peltier and an insulated reservoir to hold and store the cold is the way to go.

This method of transferring the cold to the water is unsuccessful!

Ryan

 

[Motopsychojdn]

Initial testing not what it could have been, whats the ports on the side for?
As Mr Burns would say,
'continue the research...'

Moto

 

[4Ryan6]

Initial testing not what it could have been, whats the ports on the side for?
As Mr Burns would say,
'continue the research...'

Moto

The ports on the side was a Serenity flashback, "What the hell is this ?", "It's the new engine coupler capn!", Shiny!

Actually my CPU cooling is working better than I could have ever hoped with my TECBox, so I'll just stick with what I know works!

 

[Motopsychojdn]

Maybe the catalysers are shot and thats why results weren't as good as we hoped
Moto

 

[Lutfij]

@ Ryan - All this hi-tech talk for a guy who watches alot of Dinosaurs! :lol:

"NOD-da-mama, hehehehehe"

 

[Motopsychojdn]

Dinosaurs featured in Serenity as well Lutfij,
http://www.youtube.com/watch?v=znxFrgql5dc
Mine is an evil laugh hahaaa
Moto

 

[Lutfij]

:lol:

 

[4745454b]

Perhaps, but not the momma came from one show.

 

[4Ryan6]

A CPU water block transfers the cold fast!

Because the base contact copper plate is only 3mm thick on most all water blocks.

The large copper finned heat sink has a massive amount of copper base approximately 8mm thick.

The more copper, the more retention of cold and hot!

Once the cold side begins to loose it's cold because of the heat from either a GPU or CPU, the cold conduction from the peltier transferring to the water, has to be faster than the cold loss, or else the waters thermal mass begins to slowly warm and cannot recover the cold it is loosing.

Because when the cold side has been steadily running a balance with the hot side and then the cold side becomes too warm, it causes the peltier to literally stall and throws off the balance of hot and cold.

The same loss of balance occurs if you allow the hot side to get too hot, because if it gets too hot it warms the cold side and keeps it from reaching it's coldest point, throwing off the balance.

I know this sounds totally ridiculous but it is a well known side effect of peltier use, if you do not learn to take advantage of the balance you'll never get what you want from peltier cooling.

The solid copper heat sink I used in the experiment allowed the peltier hot side to get too hot, from my experimentation a hot side temperature of 120f ~ 130f allows a balance chill factor down to 50f or 10c.

The heat pipe cooler is definitely the way to go cooling the hot side of the peltier, to achieve the balance needed between hot and cold for maximum cold conduction to the water block, to store in the insulated reservoir.

This solid finned copper heat sink experiment even though not successful has confirmed the method I am already running is definitely the way to go regarding peltier/TEC cooling.

It has also confirmed that using a CPU water block is the most efficient way to transfer the cold generated by the peltier to the water, and using an insulated reservoir to store the cold water, which is the storage buffer needed to get the maximum benefits from the peltier cooling.

 

[4Ryan6]

Continuing sharing my discoveries.

If you energize a peltier keeping the hot side cool of course, (because if not you'll burn it out!), frost begins to form on the cold side of the peltier almost instantly, the thicker the copper base the longer it takes to form frost on the cold side.

With the solid base copper finned heat sink approximately 8mm thick, I saw no frost forming even after 15 minutes of operation, and I already knew from that, it probably was not going to end up successful.

I really did not want to have spent that money for nothing so I continued on with the testing to be 100% sure.

So it's another lesson learned the hard way!

Just so there is no confusion, the test above was not a change of my configuration to something new, I want to make that perfectly clear, below is my setup in operation.

View of main computer on the left with the GPU Mini RadBox to the right.

Full view of GPU Mini RadBox with the CPU TecBox to the right.

View of CPU TecBox to right.

Full view of CPU TecBox.

Ryan