We Hooked This up to a PC

[LinusTech]

1 hour ago, Luscious said:

So repeat the test using ambient/room temperature water - the pump speed will have even less effect. A D5 running at 80% (#4 on the pot) will do 317 gal/hr or a little over 5 gallons per minute. That will circulate the entire 1 gallon coolant volume in my rig every 12 seconds, and typically well under that on the average (smaller) build. Your ghetto setup assuming that is a 40 gallon bucket would take an entire minute to circulate through. Not only does the water temperature take longer to equalize but it becomes impossible to measure a water temperature difference based on the heat load. Again you're using chilled water to remove the heat rather than allowing the components to push the water temperature above ambient - pump speed becomes irrelevant. And assuming that really was chilled water you then introduce condensation issues... but I don't care at this point. You do whatever you need to for those $$$ YouTube views.

You seem a little confused about how we tested. We tested the low flow setup BEFORE we tested the power fist. We didn't change the water in between. If anything thr power fist should be at a disadvantage.

 

Also it doesn't matter. Even if the power fist was 10 degrees cooler no one should use it to cool their PC, so who cares? This was always just meant to be a silly, entertaining video. 

[Lurick]

2 hours ago, Luscious said:

So repeat the test using ambient/room temperature water - the pump speed will have even less effect. A D5 running at 80% (#4 on the pot) will do 317 gal/hr or a little over 5 gallons per minute. That will circulate the entire 1 gallon coolant volume in my rig every 12 seconds, and typically well under that on the average (smaller) build. Your ghetto setup assuming that is a 40 gallon bucket would take an entire minute to circulate through. Not only does the water temperature take longer to equalize but it becomes impossible to measure a water temperature difference based on the heat load. Again you're using chilled water to remove the heat rather than allowing the components to push the water temperature above ambient - pump speed becomes irrelevant. And assuming that really was chilled water you then introduce condensation issues... but I don't care at this point. You do whatever you need to for those $$$ YouTube views.

Man, how dare you! Because of this video I already ordered a PowerFist and now you're telling me it might be a stupid idea that I shouldn't bother with that won't lower my temps like magic?!?!

[Nystemy]

1 hour ago, LinusTech said:

You seem a little confused about how we tested. We tested the low flow setup BEFORE we tested the power fist. We didn't change the water in between. If anything thr power fist should be at a disadvantage.

 

Also it doesn't matter. Even if the power fist was 10 degrees cooler no one should use it to cool their PC, so who cares? This was always just meant to be a silly, entertaining video. 

I can understand their argument around the thermal mass of all that water, and how fast one recirculates the given volume. But personally I don't think that is important for what were showcased in this video.

 

The thing I found interesting with this video were how drastic the difference in the performance of the water block were by just increasing flow, even if the incoming water more or less had a similar temperature.

 

From the point of "does flow speed affect the performance of a water block" we wouldn't be interested in a system reaching steady state for the test, since the steady state temperature would largely be affected by how well we can get rid of the heat from the water. Ie steady state wouldn't just look at the performance of the water block but also the radiator. _{(or in this case barrel of water.)}

 

However, it would have been nice to know the temperature of the water entering the block. And one could go the extra mile of measuring the flow rate as well, and then compare different water blocks at the same heat load and flow rate. _{(though here one would preferably want a thermal test platform like the ones Steve has over at GN.)}

 

Another topic all together would be how adding volume to a loop affects CPU/GPU boosting over the longer term, especially relevant for more varied workloads like an FPS capped game. _{(capping the FPS to reduce power output when it technically isn't needed, reducing loop temperature and providing a thermal buffer for when the workload gets harder at another point in time. Likely making for an overall more stutter free experience.)}

[darknessblade]

On 4/7/2022 at 7:07 PM, LinusTech said:

It's definitely the worst.

Lets add a couple 100000RPM fans to it to make it a even bigger nightmare.

 

Dub it the Silent sleeper.

[Sir Chadwick]

Hey Folks! Been a lurker on the Youtube channel since NCIX, always love the mod vids, whether legit or crazy, it's all good.

 

I'm also an HVAC tech and a Tinkerer... so here's an idea for your Version 2.0:

 

Think back to the vids with the hacked window A/C unit in a cooler of water with the coil of copper tubing...  HVAC parts will be your friend here  

Keep your low-pressure coolant loop separate from your hi-pressure water from the pump. To do this, you can use a "coaxial heat exchanger" from a water-source heat pump. Lots of options on Aliexpress or Ebay, or source one at your local HVAC Wholesaler. I'm on the east coast (New Brunswick) otherwise I'd track one down for you.

 

Run the loop coolant through the exchanger's inner copper line. This would normally be for the refrigerant in a geothermal system, so it will be medical-grade pristine on the inside, and as long as your CPU block is also copper, you won't have issues with dissimilar metals. Since this loop is low pressure, a couple hose clamps should be more than enough to keep your tubing in place.

Next, plumb your PowerFist pump to the water inlet on the exchanger, which should have standard NPT threads and will be designed for a decent pressure. Just make sure the flow directions are in opposite directions; a.k.a. counterflow (See attached picture). As long as the water coming out of the pump is a couple degrees cooler than the loop coolant, it will suck the heat out easily.   
 

Optional add-on.. Find a new (never used) automotive radiator or a water-to-air heater coil and plumb it in-line with the pump before the heat exchanger and use large fans (furnace blower size) to blow air through it... think of it as an upgrade from that cute little 560MM radiator and 200mm fans . Heat added to the water from the pump will be removed with the radiator, meaning it can absorb more heat from the coolant loop.

Optional add-on #2. Dust off the old hacked A/C water cooler and use it to chill the water in the garbage can!

 

Thanks for the chuckles

Sir Chadwick

[Sir Chadwick]

2 minutes ago, Sir Chadwick said:

Next, plumb your PowerFist pump to the water inlet on the exchanger, which should have standard NPT threads and will be designed for a decent pressure. Just make sure the flow directions are in opposite directions; a.k.a. counterflow (See attached picture). As long as the water coming out of the pump is a couple degrees cooler than the loop coolant, it will suck the heat out easily.   

Oh! If you were to plumb the pump so it sucks the water through the heat exchanger instead of pushing it through, the water entering the heat exchanger wouldnt have the heat from the pump in it yet, so more capacity to absorb heat from the coolant loop!

[Luscious]

On 4/8/2022 at 8:18 AM, LinusTech said:

This was always just meant to be a silly, entertaining video. 

Precisely the answer I was looking for - now I can sleep better at night