5 Watercooling Myths You Shouldn't Believe
ฝัง
- เผยแพร่เมื่อ 27 ก.ย. 2024
- These are just some of the 5 most common myths I hear on a regular basis regarding watercooling your PC. Think I left something out, then be sure to let me know in the comments section or twitter!
Visit www.coolermaste... to learn more about the MasterKeys Pro RGB Keyboards
○○○ All music provided with permission by audiomicro.com ○○○
SUBSCRIBE! bit.ly/sub2Jayz...
WATCH MY MOST POPULAR VIDEOS!: (link)
••• Follow me on your favorite Social Media! •••
Facebook: / jayztwocents
Twitter: / jayztwocents
Instagram: / jayztwocents
About JayzTwoCents
Your one stop place to shop for fun and educational videos on all things tech! Oh, and don't forget gaming! Gotta make time for gaming!
5 Watercooling Myths You Shouldn't Believe | JayzTwoCents
• 5 Watercooling Myths Y...
JAYZTWOCENTS
/ jayztwocents
Do mythbusting on overclocking
DO it
damn that hair on my screen LOL :D :D :D
i dare you
YESSSSS
+1 for this :-)
5 myths about myths plz
1 - Myths are usually not true.
2 - Myths are close relatives to a lamp worshipping species of insects.
3 - Myths tend to be busted by a pair of funnymen.
4 - Myths start out as simple misinformation and fundamentalize over time.
5 - Number two is the utter truth and definitely not a myth.
Alternative #2: a myth is a female moth
Myths is not facts
Jay could you cover myths about overclocking gpu and cpu next?
I third this!
Why? He has no idea what he's talking about.
@@spudhead169 are you sure
@@saab35draken39 Just have to watch his videos. Like the one on water cooling "myths" where literally everything he said was wrong.
@@spudhead169 *_ O H _*
1:44
I like how you're sitting on the stairs. It's like you're talking directly at me. So casual. So cool. See everybody? He is in his 30s, but he understands what the kids can relate to. I like this. I like you.
Lol?
k
Nani?
I don't give a shit about stairs.
Give me a Gaming Chairs. LOL
the fuck are you on about?
"...bigger Reservoirs don´t increase your cooling capacity..."
Me looking on my pool-sized radiatorless Reservoir: "Dont cry my little Water, Uncle Jay just make a bad Joke."
yep he really doesn't know what he's doing. Just watch his skunkworks build. Putting in dead motherboard, wrong pcie slots for x16/x16, using AF fans on radiator, blowing hot air back into his case etc...
@@donjones2500 That was obviously a joke for hyperbole, it doesn't increase your cooling capacity per se, it just makes it takes longer to equalize. Youre the idiot here taking a joke about a pool serious lol. When you have a million motherboards and one is dead, can you really be so nit picky as to blame someone for picking the wrong one? You really want to be angry huh? He's talked numerous times before on what kind of fans for radiators and that you want static pressure blah blah (in videos that are, ya know, actually supposed to be instructive), skunk works is old as hell, you should prolly just go back to your grumpy hole..
Meanwhile my stock cooler cries in the corner. =(
Don Jones I’m betting you have a huge reservoir not knowing how thermal dynamics with water works
It also isn't "radiatorless" since the surface area around the swimming pool (walls/floor/surface) is the radiator of that system and has a huuuge surface area and the thermal conduction to the ground through the walls of a pool is really high
I haven’t subscribed yet but with videos like this where you’re testing and showing REAL WORLD results. I can’t help but want to be a part of your space. Thank you, Jay.
Loved the change of scenery between myths.
The reason WC systems may heat a room up more is you can push more power through the components without hitting a thermal throttle. As far as heating your room, our PC's are essentially low/medium powered space heaters (depending on build). The wattage they draw from the wall is released into the room as heat. If an air and water cooled system draw the same power from the wall, they will heat the room the same.
or because the average water cooling radiator, lets say a 240, has more surface area to dissipate heat than an average air cooler.. the purpose of a radiator/heat sink is to move heat away from the cooled component. away means into the ambient air. more heat dissipation = lower cooled component temp and higher ambient temp. a fine example would be something like a noctua d15, which has pretty similar total cooler surface area to a 240mm AIO and can maintain about the same temps.
You're on the right track. The first law of thermodynamics tells us that energy has to be conserved. That means all the power your system draws from the wall is being put into the room, regardless of the cooler type. As long as the system is consuming the same amount of power, the room temp increase will be identical regardless if using an air or liquid cooler. For example, if a d15 and h100 are both able to keep the chip cool so it doesn't throttle, the increase to ambient will be the same.
The reason I said liquid cooling might heat a room more is due to the fact it is likely run higher clocks/voltage/power consumption since it won't thermal throttle vs a stock air cooler. Since the system would be drawing more power from the wall, running the higher voltages and clocks, the ambient temperature would increase.
This is not how thermodynamics works. Heat in, heat out. Doesn't matter how "hot" your cpu gets; if its putting 150w of heat into the room, its 150w regardless of what cooling you have.
a larger reservoir doesn't affect max temp, but it does slow changes in temp, reducing swings in temp that cause mechanical stress.
What if the reservoir was made of copper. Like a 3" diameter copper pipe cut to 8'' length and reduced on both sides for fittings. You would get some heat transfer with air movement in the case. Correct?
@@deesmods6696 With a large enough reservoir, the reservoir itself could function as a radiator. So, while in most cases he's probably correct here, in an absolute case, he's wrong!
0:40
" *hey Charles, can you press this button randomly for me?* "
" *certainly, sir* "
Love these type of videos. Keep it up. Proud of you
h3h3?
dennis045 keep it up!
YO WADDUP HILA KLEINERS
i see non-conductive fluid being best for people who have never done watercooling before so if they messed up a fitting or didnt connect a tube good enough its not gonna ruin anything when it leaks while they are filling the loop for the first time or the first week of the loops use.
Looking forward to your loop order video. Love seeing your builds. They are just wonderful. Keep up the great work.
Having water cooling that keeps your rig at a stable 40C while gaming and around room temperature when not gaming is going to keep the room cooler than if you're running a cooler that keeps your system at 60C while gaming and 40 while idle.
I've actually experienced this, as my gaming room got a few degrees colder after longer gaming sessions with the water cooler.
Eventually both systems would heat the room up to the same temperature, but it would take much longer with the watercooling and people don't generally game for more than 8 hours at a time.
next, can you go over Overclocking your GPU. the other one is kinda outof date
it's not out of date at all still applies
For the water to heat up say your cpu after having gone through your gpu block the water would have to be at a higher temperature then the cpu. So it's technically possible but if thats a problem that you have then I think you have other things to worry about.
Not necessarily. The bigger the difference in temperature between two surfaces, the faster heat will transfer from one to the other. Meaning if your water temp is 50° below your CPU temp, it will conduct away a whole heap more heat energy than if it were say, 20° below your CPU temp.
Jayz2cents spittin that watercoolin game in back alleys and shit.
Unfortunately, when it comes to large overclocks, loop order does matter. I had a build where a 7700k would go unstable if the GPU was in the loop before it as that added temperature did take the CPU just a few degrees higher and lost stability, Switched them around and it was rock solid.
Your cpu was not stable. End of story.
I'd like to see a video where we can learn more about custom loops and tube bending. It's really intimidating as a beginner for water cooling but looks amazing.
Jay please do a tutorial on how to change out the fluid in your loop
Floating Paradise I think it depends on how the loop was set up. If you set up your own loop you shouldn't need someone to tell you how to empty it. If you had someone do it for you, that should have been the first question asked. Just my 2 cents.
Grey Roots Not necessarily. I always have trouble draining my loops. I can never get all of the fluid out before having to disassemble. These videos make it look so easy..
Scott G. Helps to have a drain valve at the lowest point.
SkinUpMonkey I know. I always have one, but water is still always in the blocks, or doesn't make it around the bends in the tubing.
Scott G. For blocks make it so the block is flat then tip it around about a bit. Watercooling is all about the tipping lol.
I have been using H2o and Propolene Glycol for years with zero growth (1 year between flush) goop or corrosion, I'm now trying H2o, Glycerin and silver dilution in my hard line loop. So far so good, keeps the pump quiet and seems to cool fine, Glycerin has a higher heat transfer rate/capacity than Glycol and is still antimicrobial.
Please this "myths" a series. :)
Not for nothing but these videos really do help me. I've always wanted to dive into custom and I'll be honest I've been afraid and have always settled for aio coolers but for my next build I'm thinking of buying the aluminum ek that you reviewed and giving it a shot. Thanks Jay.
A large reservoir also looks better
That's an opinion and not even always true. A large reservoir in a tiny case looks cramped and not necessarily better.
Wade Bingham some of us have big cases so need to fill the space with something
exp10der *A large reservoir looks better except when it doesn't.
I prefer one radiator per CPU or per GPU so that I can swap cards. I like the reservoir integrated into the radiator solution to save space. However, no manufacturer makes such a system for custom make.
Just a bit of complementing with some other details on the topic:
- There seems to always have been some discussion regarding tubing diameter and how that affects temps.
And the thing is as long as you're using good pumps (laing or go home) the difference in temps is pretty much negligible so this choice should be dictated by aesthetics and ease of access instead (also thinner tubing is easier to route).
- With regards on VGAs in series vs parallel.
If you have your VGAs on their own loop, pick what you think looks best. If you have them on the same loop as your CPU, you can use that to "shift" cooling capacity between your CPU and and your VGAs in a way.
Tl;DR: If you're more worried about your CPU, run your VGAs in parallel, if your GPUs are the ones you want to focus on, run them in series.
This is a rule of thumb and it has to do with the fact that water flow through the different components does affect a bit your temps. If you run your GPUs in series, your whole loop has a bit more resistance, but if you run them in parallel, the GPUs divide the flow flow between them. So flow for the rest of the system is a bit better but each GPU gets a less flow. So their blocks are performing a bit worse, but your CPU gets better flow instead.
Ultimately this is very much into nitpicky territory so to speak and the differences shouldn't really be too pronounced, unless you're relying on a weak pump.
- Bonus: pumps in parallel vs series
Pumps in parallel offer no benefit. Don' do it. Pumps in series are useful. You get better flow and redundancy. A stalled pump (true for DDCs and D5s) won't really hinder you much either
You should be able to go lower than room temerature if you spray water on the heatsink.
In theory the air passing by should evaporate the water quickly, sucking energy out of the system, that might be a fun one to test.
Man I can't believe how much better jays videos look compared to the other content creators I watch. Thank you jay !
@JayzTwoCents can you test a water-cooled parallel against a series loop in regards to cpu & Gpu's temperatures, water flow rate and If it has any impact onto noise generation?
Well, thermodynamics state that heat in vs heat out is a net function; the laws of physics dont care about parallel or series loops. The ONLY consideration is head loss, but it winds up being a very negligible.
+JgHaverty I think you misunterstood that law. Heat is not temperature. Heat is energy. 1kg of 20°C water has way more heat energy than 1kg of 20°C air. You can only equalize temperatures and not the heat energy.
If you have a higher temperature difference, you can equalize temperatures faster. In a serial loop, you pre heat the water for the next block and in the radiator you pre cool it down for the next radiator. So the difference become lower and you get higher temperatures.
So the energy that goes in and the energy that stays in the water and leaves it, are equal. But with a serial loop more energy stays in the water and it becomes more warm.
I hope i explained it well, english is not my main language!
This is all above my pay grade...
I like that Switching of spots. Especially that spot on the stairs goes very well with that kind of Vidoe.
how old is jay his hair makes him looks like hes 50 but his face makes him look like hes 25 and he acts like a 10 year old
42
yoda won about 40ish but he could also be dyeing his hair, I've seen tons of people using grey/silver hair dye recently
you won this battle Jay but you really don't seem to be 42 if you're 42 i thought you was like early 30's to mid 30's. and i always see you with grey hair if that's your real hair color. and you still like to act like a little kid and be serious at times to so i was always thinking you was around that age group. that's why i like you jay cause of your kiddie side and your love of computers and cars
jay did a draw my life a few years ago and gave us his birthday in that video. look it up. interesring story too.
He's 36, if I remember right. He just had a bday not too long ago. I thought he was older me. Nope. He's younger.
Funny video. Now we need a myth busting video to clarify the new myths introduced in this video.
My favorite was: water cooling increases your room temperature more because it is more efficient at getting the best of of your pc.
Ymmd!
A video about flowrate would be interesting.
I agree. You could probably find the exact amount of radiators that would guarantee the most heat removal given a specified pump head.
We get it, you vape.
Flow rate won't really affect your cooling much at all, unless you literally turn the pump off. One could argue that a lower flow rate would cause the system to take a bit longer to ramp up to a stable temperature (like if you have a larger reservoir), but I think the difference would be negligible. In terms of max temps for your components, you wouldn't see a difference, because even with a low flow rate, all of the water in your loop will be roughly the same temperature everywhere (within a few degrees at least). That temperature that the cooling liquid stabilizes at will depend on the heat output of your components, the ambient temperature, and your radiator.
+Sebastian George, The argument about flow rate isn't about how long a given bit of water is over the heating/cooling components but rather how that water moves over that component. It's a fluid dynamics thing.
If the water is smoothly moving over a flat plate and most of the water molecules are moving in a straight line, the effective surface area is much lower than if the water is turbulent over that plate. The more effective surface area you have the more the gpu/cpu is cooled.
Much of this is handled with well-designed water blocks, but each water block will have some specific flow rate for optimal heat transfer. Now how important this is in a modern loop, I don't know, but in theory it could be very significant.
It's kind of like if you used a thicker block (more metal between the cpu and the water). It would take longer to remove the heat and, despite the loop being likely the same temp, the cpu would stabilize at a higher temp. Same with poor thermal paste.
+slicedtoad I agree that it COULD be a significant effect in theory. However, I don't think it is in reality. If you open up a modern water block, you'll see that directly about the CPU/GPU, the water is forced through a bunch of very narrow channels cut into the metal. Given the viscosity of water and the sorts of flow rates that you'll get from any PC watercooling pump, the flow through those channels will be pretty much laminar (i.e., not very turbulent at all).
Not only that, but the water flows past the CPU/GPU fast enough that its temperature is only raised very slightly, regardless of whether there's turbulence or not. You constantly have new water molecules coming into contact with the heat sink, but it's because you have a net flow, not because of turbulence. Like I said above, it might make SOME difference if you have a super low flow rate, but I don't think modern watercooling pumps can give low enough flow rates to cause a significant effect without being turned off.
I have a 1080ti and a 5960x both cooled on water with 2 Bitspower 480 x 30mm radiators and my temperature difference on AIDA64 under load is 3 degrees whether my fans are at 800rpm or 1500rpm. So the extra radiators help a lot with allowing the fans to run quietly and efficiently. My single 240mm AIO on the same setup would run about 15 degrees warmer at idle and would have about a 10 degree delta between those same fan speeds (same exact fans as well Corsair ML120 Pro.)
I am thinking of buying a proper pc case from caselabs. A Magnum TH10A or SMA8 would suffice, but i am thinking #yolo because TX10 is back for a limited time. Might aswell pay the extra few bucks. I had a watercooled PC 6-7years ago and then i used a old Thermaltake Mozart TX case which i did some light modding on. I bet any caselab case would be great for waterbuild. But i am a bit facinated by TX10. Have 1.5weeks to decide whether to buy it or not.
Whoa @ the TX-10. If I had that money right at the moment, or was likely to have it before they took it away again, I'd totally pick one up. A -D version. Possibility to fit up to three systems in one? Connect all of them to a single epic oversized radiator loop? Hells yeah.
My only gripe about it is that they don't seem to have as an option a desktop style case that can fit HPTX boards. In fact, the TX-10 is the only case they have that can fit HPTX boards at all, and they don't even sell it consistently. What's up with that anyway? Why would you do limited runs only of a mega case as your only case that can handle larger systems at all? A somewhat more reasonable case could easily be designed for HPTX, it just takes a few extra cm of depth. Why even limit availability of the mega case? I understand only doing limited runs of it, sure, but keep it permanently listed, and when you run out of stock, do another limited run. It's not rocket science to keep availability of cases like this. They aren't going to sell by the hundreds or thousands, of course not, but they'll sell, just more slowly.
What are you putting in it? I'm a big fan of finding the smallest possible space to accommodate the components.
rosewell1983
Why? Their cases are no nicer than a Fractal and are 5x the cost...there is ZERO reason for anyone to buy one of those.....ive built in 6 of them and all but 1 of them the customer switched out to other brands..... The other one didn't because he got the case for free Pre- modded from a giveaway
Really? I've never heard of, nor, when I just tried browsing Fractal's catalog just now could I find, a single case from Fractal that could support anything like the modularity of a CaseLabs case. None of them can handle a repurposed automotive radiator without extensive re-machining of the case itself. None of them can house more than one system in a single case. None of them allow swap ins and swap outs to the extent that CaseLabs demonstrates. And they're a helluva lot more expensive than other cases equally as dismal at modularity.
Propylene glycol I feel might be the magic cooling fluid that everyone is searching for, it provides corrosive protection, its molecular which means it should be non-conductive and should have good thermal conductivity. Would love to see a test bench please. Please try a leak simulation by spraying some on to a old system just to see what happens, try it straight then 60/40 with distilled water. Can use food grade PG which is used in E-cig fluids or Amsoil makes a nontoxic antifreeze which is industrial grade PG. Thank you.
Surely if I stick the radiator into my freezer it would cool below ambient room temp??? 😂
Jas George for a short while, probably. If you live somewhere cold the trick is to stick the rad outside of your building 👍
Jas George bitwit did it with an ice water bath for the radiator and it did work. Linus already showed that putting the heat into a refrigerator or freezer won't work as it can't handle the heat being put out by the components.
It would destroy your fridge at the same time. Linus tried that one for science and gave a mini fridge a major hernia.
Depends on the output of the compressor, the size of the condenser and radiator in your fridge. Realistically a consumer fridge or freezer will not have the capacity to do so as most of the time they have less heat transfer than a small A/c, proof of which is seen when you hook up your new freezer and it takes forever to freeze stuff.
just use a peltier cooler on it. sub ambient for cheap
About the water flow rate: the german magazine PC Games Hardware did a test that ended with 4 CPU and 5 GPU blocks (for increased flow resistance) and temperatures due to resistance increased by just 2.5°C
"Boils Down" you really don't want your loop to boil.
ever since i built my new system using the corsair h150i aio my room sits at a very toasty 85 f at all times. I use a 9900k and the cpu temps never reach over 40c even on a high load for multiple hours on end so i can definitely vouch for cpu heat dispersal heating your exterior environment up even more while keeping your system much cooler.
I am pretty sure watercooling isn't going to make your room temperature higher than air cooling. Or at least not for the reasons you mention.
Yes, watercoolers remove heat faster and more efficiently, but they still remove the same amount of heat energy. If your PC pulls the same amount of power from the wall, the same amount is going to end up as heat in the room, regardless of cooling solution. It is just going to do this by removing the same amount of heat, but at a lower equilibrium temperature (or Delta T).
That being said, better cooling usually means more overclocking, either by TurboBoost/GPU-Boost technology, or manually by the user, which increases power draw, and of course the heat being produced in the room.
Dude when you remove heat faster and more efficiently that means you're removing more heat. I think you've confused the heat produced with the amount of heat transfer and the temperature, and while heat produced will be the same for both systems in the same conditions, not all of that heat will translate into higher room temperatures since it's still somewhat isolated within the case. The more efficiently your remove the heat and dispel it outside the case the easier it's able to increase the ambient temps.
Sir Gildo So basically the equilibrium temperature of the room will be a bit higher with watercooling since heat enters the room from the case faster, but passively leaves the room just as slowly
", not all of that heat will translate into higher room temperatures since it's still somewhat isolated within the case."
You're an idiot and you're the one who's confused. Shut up.
You have two possible situations: Either the pc is heating the room, or the pc is heating itself. If the pc is at a stable temperature, no matter what it is, and it's pulling 500 W from the wall, those exact same 500 W will end up as heat in your room. Period. You can't avoid it. It's called thermodynamics.
The only way a water cooling solution can heat a room more than air is if it allows the cpu to generate more heat via overclocking and/or reduced throttling. Though it's possible in a room with poor insulation for there to be a steeper radiant as you move closer to the cpu, the actual total heat energy put into the room from the system remains constant to a specific power draw. I suppose you could argue the tiny amount of heat generated by losses in your cooling itself may contribute more than a lower power using system of cooling would through it's losses.
Indeed, christopher90, if a PC takes x watts to run, it will put almost all of that power into heat regardless of the method that heat is removed from the PC. I'm frankly a little surprised that Jay said that watercooling might heat the room faster, since he has had all other cooling-related arguments correct. We could nitpick and say that the pump in watercooling adds an extra tiny amount of heat (in the order of a couple of watts), but that is about a few percent of the total amount in idle and much less in load conditions.
Can't wait for the loop order video. Have some friends who still don't believe it doesn't matter. With that video I can finally shut them up... Great work Jay.
OMG jay, can you do a loop with a refrigerated rad????
like with one of these
thereefshop.com.au/index.php?main_page=product_info&cPath=43&products_id=756&zenid=5abf43e21a7a321b82649dbc75546fee
Robert Goodfellow you can but you would need a wider pipe or a converter for the connections to a fridges rad you would also need to cool it
adamania its not a fridges rad its a cooler for an aquarium and there are ones specifically for computers there are already a few youtube videos on them (go search it lol) but the videos are poor quality, no where near jays level, and they also dont try overclock or show any benchmark results.
I am a Mechanic, and I can tell you that the size of the reservoir is ultimately crucial to the cooling capacity to any liquid cooling system. Whether you are dealing with a CPU/ GPU , or a nuclear reactor, the amount of coolant is the key. Water tends to hold heat very well (Even better than Iron). Also, speeding up the pump might not be the best idea, as the liquid in the radiator needs time to cool down. Air is a very poor conductor of heat.
That being said, listen to Jay, (and those like him that are trustworthy), cause they have already done the research, and made most of the mistakes that you should avoid. (Be it for cost, pride, or both) I'm no different.
Just remember that "Everyday's A School Day !"
Thanks Jay!
Matt
mostly correct, a few things are wrong tough.
the order used to matter alot more in the past, especialy if you are cooling 2gfxcards before the cpu. what you didnt touch on was flowrate. more radiators will increase resistance within the loop and slowdown your flowrate. if the flowrate gets realy low, the order of your components starts to matter.
For some reason I do not understand there are people who still believe that Mac Pro is better than a regular computer in the same price range in video editing. I would love to see you make a comparison between the two
Because they are stupid. I saw a $1100 mac a month ago with a 1.9ghz duel core CPU, integrated graphics. 4gb of ram, a 1tb HDD, and a 1080p screen. Just because it's apple doesn't make the shitty specs better than a pc with a 7700l/8700/ryzen 7 and a 1070. Editing computers still use the GPU and CPU so having good ones is not exclusive to gaming. At best a mac has a 4gb AMD card at the price range where it should have a 1080 TI or 2 of them. Apple fanboys seem to think the apple logo magically transforms their low spec hardware into some ultimate editing computer.
forza1sra they do have an amazingly efficient OS i still heavily prefer Windows but just saying
They used to have hardware in the cpu that helped with this sort of thing. however with the switch to x66/amd64 intell chips they lost that advantage. They can get some gains with a tight optimization of software to known hardware configurations that is harder to do with a more varied ecosystem like you see in the rest of the pc space. However there is only so much you can get that way. So for an identical spec system they might be a few percentage points faster, but if the same money can buy you a significantly faster hardware spec, then the value quickly shifts away from them.
@@drieshuybens4594 so u would pay 1100 for just an OS. Thats like buying a ferrari for the looks but the engine they gave u is for a camry.
7:00 Loop order does affect temperatures. By a couple degrees C at most, typically, so one could say it doesn't matter. But it does make a difference. Any block coming right after a radiator is going to have cooler water flowing through it, which means heat transfer into the water will be faster, due to the increased delta T.
The one component that most stands to benefit from that is the CPU, due to the high thermal density of the interface. So putting the CPU block after all radiators might net you a couple degrees off top temperatures. GPU's are much easier to cool due to the much larger die area and much lower clock speed. And with all radiators together, the water entering them will be the hottest it is in the loop, making heat transfer out of the radiators faster.
Water temperatures never reach equilibrium across the loop. There are always gradients, where cooler water flows out of radiators, and warmer water flows out of blocks.
Are two 240mm rads nore efficient than one 480?
*more
Very interesting question though.
I would say yes.
2 240mm rads should have more copper than 1 480mm rad, think about the ends of a rad where the fittings are. So you should have more surface area to transfer heat (Very small difference though).
But there is more things to consider, the placement of the rads will make a difference. So if you can mount 2 240mm so that they only get ambient air (Probably easier than mounting a 480mm as an intake, never seen a case that supports a 480mm intake rad) then they will out perform the 480mm.
Would be an interesting to see a video on it though.
Depends of fins per inch and surface area
I wonder what the lens he's got on that rig on the left at the end of the vid. Honestly a filming kit video would be super cool
Hey, You should do a build "Extreme watercooled" where like you don't watch price or difficulty to install/wash, just the pure look and the big creasyness, just like How clean and how mush sick bend you can put in it like I think 2 lines, two colors, most place they get like over and under and beside and re-over, etc... like the highest difficulty level that you can do.
Also say in the video to not reproduce the build because it might cost kind of a lot of money to do those bend and every thing
But most importantly:
MILK IS NOT A GOOD COOLING LIQUID
doesn't loop order only matter for the fact that the water will be slightly heated if it goes trough the second component? but having a proper balance in heat dissipation and mass/flow of the water will probably neglect this.
for the follow up video it would be nice to have a temperature sensor (calibrated ) in between each component of the loop to show the temperature change of the water in the loop.
"Having water cooled systems is probably going to heat your room up even more"... WRONG
Jay, Jay, Jay. Man, so, so, so close.
What you're referring to (loosely) is the law of physics regarding the conservation of energy. That people still don't do their own research is why you're addressing this issue but basically, there is no additional energy being pushed into the rest of the room for it to be 'warmer'. It may very well exchange heat to the rest of the room faster, but the room will still retain/dissipate heat at the same overall rate. If your room evens out at 30C for an air cooled PC system in it, it'll do the same with a water cooled PC system in it, it'll just get to that temperature faster; much like, in fact, one of your other myths regarding loop temperature equilibrium.
So good work Jay, but your specific use of words here may very well have implied something other than what you intended. Please keep up the excellent content, and please consider the idea of designing/modding your own case for something truly unique.
Water cooling keeps the cpu cooler overall, so that extra heat goes to the room instead of the cpu, thus making the room warmer
@@tandlose That comment's over 3 years old, wtf are you even doing replying to it? Anyway, unless the room you're referring to is some sort of hermetically sealed box, any energy flowing into the room will escape out somewhere, so no, the room won't get 'warmer' to any greater degree between air and water cooling. It'll heat up *faster*, sure, but it'll level off at the same peak temperature when equilibrium is reached between and .
@@daiphelion7002 lmao 3 years ago
Well, one thing a bigger reservoir will do is to help against short bursts of high load. This might not be a common thing while gaming but during normal usage of a PC, it's quite common to have full load for a few tens of seconds. My CPU air cooler for example immediately goes full throttle the moment the CPU starts running at 4.5 GHz. The larger the reservoir, the stronger the potential dampening effect on load vs. fan speed and therefore the quieter the machine with such workloads, up to the point where even extended compiling or exporting sessions will not lead to thermal equilibrium and therefore not max fan speed.
I think its time you used a thermal camera and benchmarked different loop orders
What about just looking at the system temps lol
@@jeremymcadam7400 I like the thermal camera idea better. I think it would be really neat to actually see the heat travel through the loop and get dissipated until equalization
@@SirChaosS yeah fair enough, that would either prove or disprove the equalization theory
Thank you for the humor. Please bring back "Transitions", they were a fun touch. - Thank you for explaining water cooling to your fans and viewers. - Keep up the good work. - BTW, this is a water cooling video, however, my Noctua NH-U14S is a great CPU air cooler.
Parallel vs serious loop? 6:56 :)
Peter Kostovčík serial
Cereal
Love the scene change for every myth. Youre produktion quality has skyrocketed in the new studio.
I would be very interested in seeing numbers of how many radiators are needed for different setups if you want sub 1000rpm fans and working towards silent systems.
Im typing this from the cooler master masterkeys pro L rgb with mx blues and it is by far the best keyboard ive ever used, the rgb looks really bright and the mx blues feel really good.
Public service announcement please stop saying first
Bad no
Streets Speed 484 first
Tucker Bergantino it's not hard, try starting a discussion instead about the video or just something else
PENIS
Streets Speed 484 FIRST
Hey Jay. I love your new studio and all. But you should do videos at the house more. I miss seeing skunkworks in the backround and the more down to earth vibe your older videos had. Don't get me wrong I love the step up you've taken with your channel and the content has been getting better and better but every once in a while it would be really cool to see a video at home for old times sake😉
Jay cmon dude.... :-/
1) Wrong, a larger reservoir gives more thermal mass in the system. If the radiator can support the cooling function, it will give you "lower temps" because you aren't heating the total mass the same amount. Q=MC(T1-T2); MASS is a function of thermal power. Your transient response time will also be slower proportionately to the amount of mass gained.
2) At the same fan speed, larger radiators will have a faster thermal transient response; albeit not below ambient, and point of diminishing returns applies
2b) wrong, heat in, heat out. If youre cpu is putting out 150w, its 150w. doesn't matter if youre cooling it with air or water; youre still putting 150w of heat in your room.
3) Certain chemicals don't strip ions; some of the nano based fluids don't strip ions because they dont have water in it. Think of mineral oil cooled pc's. The only ionization that will occur will be from friction stripping.
4) Dye's are so all over the place; youre right about this one
5) Right on.
Yea, I can put a cup of water on the stove, or a gallon of water on the stove. Sure, eventually they'll both reach 100C but it'll take so much longer for the gallon to reach 100C that in practical use like gaming it's literally going to keep your temps lower.
he's actually correct, regardless of the amount of fluid being used, eventually, they would both equalize at the same themp, sure the larger amount would take longer, but it would still reach the same temp. Ran this test in my thermodynamics class years ago.
They are arguing that if they have a larger res it takes longer for the system to reach equilibrium. Which, they're right, it will take maybe a second longer.
Thermal transfer is faster across larger gradients. A larger reservoir at most creates a small delay in reaching equilibrium.
@Jg. You're wrong. He's right on #1. #2 he's right again. Water cool is going to be better at moving heat from the component to the ambient air. Why would people water cool if it didn't lower the temp of their components. If you air cool to 90C but water cool to 85C you have 5 degrees of temperature moved from your cousin to your room... Hence warming the room more that air cooling
Digging the filming method with this one
@JayzTwoCents This comment is for JayzTwoCents and anyone else that could weigh in on it. First of, I have very little knowledge of water cooling. I do NOT have any experience with water cooling myself, moving on. When running water cooling, is there a filter you have to change? and if so, what would the interval be? Next question, Say you are running a loop that includes your GPU and CPU and it's temperature is at 105 degress. Is it possible to drop the temperature of the water to effectively drop the temperature of the components? And lastly, if you have a system on one side of the room and your plumbing and radiator 8 feet away. Would it run colder than if you had your water cooling system inside your box?
actually what jay says at 2:54 is a myth: The amount your room heats up depends only on the total power consumption of your system (law of conservation of energy). The only thing that makes your room hotter when watercooling is the additional power your pump (and maybe additional fans) need, which most likely is negligible. In other words: a 1000 watt drawing pc will heat your room exactly as much as a 1000 watt heater, no matter how you cool them.
Speaking of cooling fluids... pH neutral water is very good but what is pH neutral. There are a lot of pH test sets for sale on pet stores or aquarium stores. You will always be able to test your loop.
Myth one: watercooling has reasonable prices.
I think the thermal capacity of water is so big that it makes no big diffrence in wich order you mount the blocks or radiators.
But i would say in a really small loop with only a small radiator, a reservoir could make a small diffrence. Because it increases the total volume of water, thus allowing a little more heat to be absorbed into the water, wich could help in smoothing out thermal spikes.
But unless you come up with some ridiculous borderline situation, the diffrence would be barely measurable. And even then, it would be more effective to add another radiator instead of a bigger reservoir.
Thermaltake Opaque dyes will destroy your clear hard tubing eventually. Destroy is a strong word, but you'll be stuck using large pipe cleaners to clean the residue out of the loop. the powder used to make the fluid opaque will eventually stick to the inside of the tubes near the bends, leaving a sticky white powder that makes your tubing look dirty, cloudy or straight white.
Please test static pressure fans on case, are they better or worse than airflow fans? thanks in advance. I love your unique and useful contents! :)
Jay, were you into computer around the early 90s when we used dorm freezers as computer cases with absorber pads to keep moisture away from the motherboards
Really useful info. I've never built a liquid-cooled system, but I want to do this for my next build!
Thanks Jay! Going to do custom water loops for the first time on my next build and this info is great, looking forward to the follow up vids as well!
This was very helpful. But I am wondering how do I know when it is time to change out my fluid? Are you supposed to change it yearly? If you don’t change it out will that cause the system to turn off? Thanks so much I have learned a lot from your videos!
what would be really good for debunking the loop order myth would be to have some thermocouples installed at various points around the loop. If what Jay is saying is true then each thermocouple should read nearly the same temperature on either side of each waterblock.
You could also do some quick math using flow rate(assuming you could determine that), specific heat and the wattage of each component under a waterblock to find the theoretical increase in temperature over a waterblock.
I just calculated the temperature change over a 500 watt component would be about 3 degrees celsius.
That was assuming a flow rate of 0.05 L/s and a specific heat of 4180 J/kg*K.
So yea, Jay is right. Not a big change in temperature.
Woah that bonus really threw me, I assumed that was a major benefit of water cooling having the exhaust air be cooler and not having the room be super heated.
Loop order doesn't matter in a standard sequential loop. But if you have your GPU running in parallel with your CPU and the CPU in sequence with a radiator then your CPU is going to have less flow because there's more resistance in the route. When planning a complex loop it's handy to think of it like an electric circuit... Or just keep it simple.
I can confirm that loop order dossent really matter.
I use a big external 140x9 phobya rad, I installed 2
temperature transmitters, one in, and one out of the rad,
the temperature difference is only 2 degrees celsius.
That means if I change the loop order, I can max gain 2 degrees on any of the waterblocks
Loving this content, one I've heard is that having your pump run faster will give you lower tempretures
I am loving the videos to remind us of the basics and now even more so since I will be doing a new build and want to watercolling because the system will have top of the line components and I want the most i can get out of them.
Corsair H60 on a XEON E5-2690. Average temp=100 degrees, 900 RPM fan. Max all 8 cores 16 threads for 10-12 minutes compiling a Linux Kernel ... Temp 120 degrees, fan 1800-2000 RPM. Stock CPU Heatsink/Fan would run 3000 RPM and Temps to 190 degrees. The H60 is my first liquid cooler. And it's awesome.
i feel like u should do a HUGE reservoir like a bucket and put it in a loop.. i feel like that would cool it down. surface area is the biggest factor in cooling so i would think something that water can expand in would cool a system.
As far as loop order goes, I like to have my GPU first in the loop and pretty much straight out of the radiator as that is the coolest the liquid will ever be in the system. No idea if the temperature of the fluid out of radiator is significantly lower than overall loop temp, but it makes me feel good so I do it. :)
The only link I can think of for Large Reservoir being better for cooling is if you have fans blowing across it, making it a really bad Radiator as well :)
well a larger resoirvoir can give you lower temperatures if you are not at 100% load all the time because it takes longer to heat up the water it "smoothes out" the temperature spikes, but as you say under 100% load it won't make a difference
A few weeks ago I was browsing through Facebook ads, and a tower with a custom water loop popped up for a reasonable-ish price, so I looked at the pictures to get a better look at what condition it would be in. I tried to follow the loop through by zooming in and I figured out that it was in the opposite way. More specifically, the pump was pushing water to the radiator, which then sent it to the components, which ended up in the water tank. Basically the water was hitting the radiator BEFORE the components it had to cool down. I really wonder if the guy even managed to sell this thing.
1:40 No you just said it DOES do something for temperature, it causes it to take longer to reach higher temps, should have made that distinction a tiny bit more clear.
Can't wait for the myth buster episode!
Thanks Jay great video, i was interested in the bit about the length of time it takes for non conductive liquids to become conductive, it would be cool to see a follow up about that as well. Thanks again keep it up.
Have a massive reservoir where the water is used and not returned to it, just pumped down the drain. Keep the massive reservoir super cool, or as cool as practical without starting the condensation issue. By the way, can't you control the humidity level in the room to zero? If so, condensation will never be an issue and you can go way sub zero.
Changing the type of cooler being liquid or air won't change the amount of heat being produced. So unless your fans/radiator exhaust to a different room or outside, the room will heat up at the same rate.
If the dye was deigned for water-cooling, it's fine. I believe ek includes a dye with their kit and you mix it with distilled water and it won't clog it. If you look at food coloring, then yes it will.
You should have the radiator before the pump and reservoir because the pump does not like hot water. A D5 maximum operating temperature is 50°c / 120°f, so too get the longest life line of the pump give it cool water. And yes I know the water is almost the same temperature every were in the loop but even a 5°c difference can mean a lot. Say that ambiant temperature is 25 degrees and the water before the hot components is 10 degrees over ambient that makes the water before the pump 40 degrees c
Could you please do a top tips for buying used parts. Great channel. Thanks Derek.
Jay please make a video on cooling effectiveness of parallel vs series water cooling loops.
Another good video. Your quality is constantly going up and inspires me for my own channel.
Hey Jay,
I would like to see more reviews on monitors that have G-sync, freesync, freesync 2, and some monitors without any sync but would be good budget options for 1080, 1440, and 4k gaming. Keep up the great videos. Here are a few monitors I had in mind. Samsung Chg70 27"(freesync 2 monitor), Samsung UH75 28" 4k monitor, Samsung CFG70 Series 27-Inch, Samsung Chg70 31.5-Inch, Acer G257HU smidpx 25-Inch WQHD, ASUS MG279Q WQHD, 178° 27-Inch FreeSync Gaming Monitor,
Acer XF270HU 27-inch WQHD
From what i understand, if the aio is not correct then air could get caught in the pump and shorten the lifespan. I was gonna run 2 more front fans but ppl are saying thats where the radiator needs to be.
There is one way to get a lower than room temperature for a loop, that does not heat the room, and that is if the loop is fed through an external chiller. This isn't a new idea, IBM did this with the 308X series in the early 1980's. The Cray-2 used something similar also in the early 1980's.
If you don't plan on switchign colors every couple of months, I recommend food die. I use it myself, no impact on temps, vibrant colors, very cheap. Your tubing does get stained though, and it fades faster than other alternatives, but considering how easy it is to recolor again, it shouldn't be a problem.
Ironically, food die is the only die Jay recommends against :V