This Cooler Draws 2400W!??

[AlexTheGreatish]

Ryzen 7000 just launched which means one thing - we need to sub-zero it, and this time we're doing it on the cheap.

 

Buy an AMD 7950X: https://geni.us/k61Br

Check out the AsRock X670E Taichi: https://geni.us/NsOJJ

Buy an MSI 3090 Suprim: https://geni.us/r4Gu9K

Purchases made through some store links may provide some compensation to Linus Media Group.

 

 

 

[mariushm]

I was surprised you didn't attach the hot sides of the peltier elements to a big aluminum plate or some big heatsink ... don't you have to cool down the hot side to keep them efficient?

 

[Retro_GamerC64]

13 minutes ago, mariushm said:

I was surprised you didn't attach the hot sides of the peltier elements to a big aluminum plate or some big heatsink ... don't you have to cool down the hot side to keep them efficient?

 

They did, it was the side with the purple dye.

[Ubersonic]

This brings back so many great memories, back in the day before companies brought out "specialist" PC cooling fluids to cash in it was common to use engine coolant in water cooled PCs (well, apart from the guys who used straight tap water then blamed aluminium when their loops turned to sludge xD).  Prestone was always my go too fluid in the 00s for that cool greeny yellow colour, ah great times.

[Bitter]

1 hour ago, Ubersonic said:

This brings back so many great memories, back in the day before companies brought out "specialist" PC cooling fluids to cash in it was common to use engine coolant in water cooled PCs (well, apart from the guys who used straight tap water then blamed aluminium when their loops turned to sludge xD).  Prestone was always my go too fluid in the 00s for that cool greeny yellow colour, ah great times.

Hot tip, that Prestone also glowed under black light.

[darknessblade]

For the loop why did Alex, not use those valve connectors, that allow you to close certain parts of the loop?

 

That would make removing the air a lot easier.

 

Or have a free tube, that is just for air-drain

[Bitter]

I feel like having a pool of cold water to draw off of might have been beneficial, seemed like either you didn't have enough heat capacity in the cold side, the cold plate wasn't optimized, the mounting sucked, or the flow rate was too slow...or all the above.

[SolarNova]

A fun video to watch for entertainment ..but .. Roman did this better tbh, felt more 'professional.

TECs can work to get subzero, but u need to run them as efficiently as possible (as they are inherently very inefficient), with plenty of insulation, correct voltage tuning of the TECs, and proper flow rate of coolants

 

You can even stack TECs to get colder at a reduced cooling capacity, which would have been great for your setup as you had multiple blocks.

You take a smaller TEC and cool the hot side with a larger one, you can go further than 2 ofc by then get an even larger one to cool that TEC ..and so on.

You tune the TECs so that you achieve a lower total cooling capacity but high Delta allowing for temps well into subzero , like -50c.

You then make lots of these stacked TEC units to achieve the cooling capacity you need.

[Dutch_Master]

As a refrigerant engineer, I'd suggest the following:

 

1. mount a decently sized reservoir on top of the Peltier-frame
2. mount a normal pump at the height of the bottom of said reservoir, allow space for a 2nd pump

This allows for easy filling (gravity is free!) and the "law of communicating vats" dictates there's an even level of fluids on the other side of the loop. This means the pump only has to cover the distance from the bottom to the top of the reservoir. Using valves, as suggested earlier, is an excellent idea as it allows flow control to the various heat-exchange elements. The pump is not on the top of the loop, that's the fluid level in the reservoir, so no problems there. For air-bleeding, add a T-junction just below the pump with another valve.

 

Enjoy!

[YellowJersey]

One question I've always had about sub-zero cooling and the condensation problem: Would it make any sense to mount the motherboard upside down? That way condensation would fall off rather than start building up? Or is there something I'm missing that means that wouldn't work?

[Needfuldoer]

On 10/9/2022 at 9:43 PM, YellowJersey said:

One question I've always had about sub-zero cooling and the condensation problem: Would it make any sense to mount the motherboard upside down? That way condensation would fall off rather than start building up? Or is there something I'm missing that means that wouldn't work?

Wouldn't that just move the puddling to the back of the board? It's still going to accumulate condensation once it gets below the ambient dew point.

 

How about spraying as much of the board as possible with conformal coating, or running the board inside a chamber filled with pure nitrogen instead of air? (I'm sure the extreme overclockers already went through plenty of ideas back in the NetBurst days...)