Need Clarification on Fans

[Strayone]

I have been trying to put together a new computer with but have ran into a few problems: My desired setup ran into a problem when I discovered a H170i Elite LCD XT cannot be used on my Gigabyte RTX 3070 Gaming OC 8G V2. After I found Bykski Watercooler For GIGABYTE Geforce RTX 3070 GAMING OC 8G ,With Back Plate ,Full Cover Water Block so I decided to look into making a custom loop using an XD5.

Again I ran into problems, there was no Soft tubing so I would have to go with Hard Tubing. This would mean alot of setup costs for bending, parts etc. Price wise it just doesn't seem practical, ordering off your site, given the low value of the $AUD, it would be well over $1000 just for 2 x 420mm rads and the XD5. I also noted tubing max temp to be 60°Celsius which I have no doubt I will exceed. Given this is my first time using water cooling, I think I will just stick with the H170i for my CPU and create a decent airflow style to try and keep the GPU cool.

Given where I live in Australia, heat is a problem, and at times so is airflow (still dead air), so I need some fans with some decent Airflow and Static Pressure behind them. Given I have a 7000d Airflow Case and the Commander Core with the H170i, I decided to buy a Commander Pro for heat sensor monitoring and control. Now since I have 6 x 140mm fans from the case and H170i, those fans will be connected to RGB but they don't provide enough flow and pressure for my needs. I love Corsair and their  AR140 Elite RGB and ML140 Elite RGB are a good step up but still not enough.

So here is my problem, I would like to try Noctua 140mm NF-A14 industrialPPC IP67 PWM Max 2000RPM Fan. I have searched high and low for some clarity but I am still unsure. I have read the 3000rpm wont work with Pro, probably due to the high wattage and amperage demands but I have found articles suggesting the 2000rpm is compatible with ICUE software just no RGB, eg;

 https://forum.corsair.com/forums/topic/153404-corsair-commander-pro-issue/

Given this was written in 2019, and since I have never even tried all of this RGB and fan control stuff before can you confirm that this is indeed true. Also given the date it was written, I am not sure the IP67 (Water and Dust Resistant version) was out then, if not, has the new model changed things, or has ICUE software changed or updated to include them? Given I need so many fans, buying them and finding out they don't work would be a massive headache and possible loss of money!

My motherboard is a Gigabyte Z790 Aorus Elite X WIFI 7 which has USB Headers with 2 Amp and 24 Watt each so I know I can fit and run them normally but can someone please clarify if the 2000 rpm ip67 can be still used with the commander core or Pro, and has anything changed with the 3000rpm or is it still unusable?

Just FYI

Case Setup:

Case:      Corsair 7000D Airflow

MB:         Gigiabyte Z790 Aorus Elite X WIFI7

PSU:       Corsair RMx Shift 1000w

Cpu:       Intel I9-13900K

Gpu:       Gigabyte RTX 3070 Gaming 8G OC V2

Ram:      Corsair Vengence 64G 6400Mhz

Cpu Cooler:   Corsair H170i Elite LCD XT

Fans:     As many as needed

HDD & SSD: 1 Samsung 990pro, 2 HDD (for now)

 

[c-attack]

11 hours ago, Strayone said:

Given this was written in 2019, and since I have never even tried all of this RGB and fan control stuff before can you confirm that this is indeed true. Also given the date it was written, I am not sure the IP67 (Water and Dust Resistant version) was out then, if not, has the new model changed things,

Nothing has changed as the problem is physical versus software.  The IP67 model was out then and what makes it different is the housing and sealing rather than the motor itself.  The 2000 PWM version should work regardless of IP67 rating.

 

3-4 fans at 2000 rpm is a torrent of air unless you have done something to block the flow.  Remember the case fans control the internal case ambient temperature.  They won't lower your GPU directly, except by decreasing the baseline temperature (internal case temp).  There should be no need for 3000 rpm fans and those will be a control problem on the Commander Pro.  Presumably it is the GPU temp you are concerned about rather than actual internal temperature?  That you need to control with its own fans.  

 

Most people who go this route wind up disappointed and there is a minimal difference between a bunch of fans at 1600 vs 2000 rpm.   Maybe we should get into specifics about what parts of the case are too hot, general layout (assuming H170i is up top as exhaust), etc.

 

On a side note, the 60C temperature rating for soft or hard tubing is the liquid temp, not air temp.  It is impossible to heat up the liquid in a custom loop or AIO to 60C without stopping or severely restricting the liquid flow.  It's also unlikely you air temp in the case is 60C.  That would make the AIO sit at 60C coolant temp as well and it has the same limit.  At 60C liquid/air temp, that becomes the minimum temperature of the CPU and GPU with no voltage.  The moment either component tries to do anything you'll hit the thermal limit, so rather than a physical limitation and product failure, you bounce of the hardware limits first.  

[Strayone]

1 hour ago, c-attack said:

Nothing has changed as the problem is physical versus software.  The IP67 model was out then and what makes it different is the housing and sealing rather than the motor itself.  The 2000 PWM version should work regardless of IP67 rating.

 

3-4 fans at 2000 rpm is a torrent of air unless you have done something to block the flow.  Remember the case fans control the internal case ambient temperature.  They won't lower your GPU directly, except by decreasing the baseline temperature (internal case temp).  There should be no need for 3000 rpm fans and those will be a control problem on the Commander Pro.  Presumably it is the GPU temp you are concerned about rather than actual internal temperature?  That you need to control with its own fans.  

 

Most people who go this route wind up disappointed and there is a minimal difference between a bunch of fans at 1600 vs 2000 rpm.   Maybe we should get into specifics about what parts of the case are too hot, general layout (assuming H170i is up top as exhaust), etc.

 

On a side note, the 60C temperature rating for soft or hard tubing is the liquid temp, not air temp.  It is impossible to heat up the liquid in a custom loop or AIO to 60C without stopping or severely restricting the liquid flow.  It's also unlikely you air temp in the case is 60C.  That would make the AIO sit at 60C coolant temp as well and it has the same limit.  At 60C liquid/air temp, that becomes the minimum temperature of the CPU and GPU with no voltage.  The moment either component tries to do anything you'll hit the thermal limit, so rather than a physical limitation and product failure, you bounce of the hardware limits first.  

Thank you c-attack, that helped a lot. Yes, my problem is the GPU and the AIO was going to be my solution but unfortunately, no way to attach an AIO to my GPU so for now I will have to work with fans and build up a proper water loop system when I can. Thankfully I don't think I will be pushing the GPU too hard and I certainly wont be overclocking it so I should be good if I am careful. I was concerned that a lack of enough airflow would cause ambient heat off the GPU to heat up the inside of the case, plus from what I have read about the I9-13900K it gets pretty hot even with a really good AIO like the H170i I am getting.  With the combined heat build of both, plus the reports of increase temperatures from weather in the future, I was concerned it would always be red lining in heat. I was hoping a strong airflow would help, the more the better given possible rise in outside of case temps of air.

 

When I asked around a few people said "Why bother, a pedestal fan works better", so when I saw a standard pedestal fan does average 2000 CFM I was shocked. I then had to buy a fan for another computer, seeing the CFM and Static pressures of fans I was hesitant but bought a Cooler Master MasterFan 200mm Addressable RGB Fan which has 90CFM and 0.88 SP. I have to say I was not impressed, it offered very little help towards cooling. I realize its an airflow fan with very little SP but it was facing very little resistance in the case. So when it came time to start building the new computer, the talk of heat from 13th/14th Gen Intel chips, especially the 13900k and the fact GPU's now produce really high heats themselves was of no small concern.

 

The Noctua 140mm NF-A14 industrialPPC IP67 PWM Max 2000RPM Fan and 3000RPM offer impressive CFM and SP specs so I certainly wanted to give them a try but not its not going to work with your products. I was concerned the I was going to overdo it. I had ideas of 7x140mm 3000rpm and 3x120mm 3000rpm plus the 6x140mm AF140 Elite from 7000d case and AIO, it was meant to keep the case and 2 radiator coolers cool. Given your advice, maybe I was overdoing it, maybe........just a little...

 

Given what you said about it also not directly affecting the GPU, I was a little concerned about that and I will have to deal with it at some point but for now I just need to keep the Internal temps under control and with the control and sensors of the Commander Pro, that should help a lot. Thank you so much for letting me know the Noctua 140mm NF-A14 industrialPPC IP67 PWM Max 2000RPM Fan still works, it was a concern that I was wasting money.

 

As for Coolant Temperature of tubing etc, I was concerned that because of the weather here, the fact the i9-13900K is always trying to push towards 90-100 °C, the RTX 3070 around 70-80 °C and combine that with long term usage and I am not sure if that would make the coolant reach over 60 °C. As I said, this is all new to me when it comes to water cooling which I why I went with an AIO originally. 

 

For the case setup, I am planning to use the 420mm radiator of the H170i on the front as an intake. I was told that would be best for the CPU but the temps in the case would build up quickly and be harmful to the GPU which would be sucking in the heated air off the radiator. To get around this I was told to increase airflow and static pressure of front fans of case dramatically so the case never has a chance to heat up and cause a problem to the GPU. I was also told to create a positive pressure style cooling system to keep out dust. The Noctua fans seem to solve both problems, and after seeing that previous fan I bought, the industrial fan offered peace of mind that it could and would do the job. Cost a little bit more but from what I have heard, you get what you pay for. I would have preffered Corsair RGB style fans that had these specs but the AR 140 Elite's where the best I have found sadly.

 

Again, thank you so much for your help c-attack. I couldn't find any info about the fan other than your original reply post. I still have to look into figuring out and building a proper water loop style cooling system but for now knowing that the 2000RPM still works with Commander Pro is a huge relief and now that I know that maybe they are a bit stronger than I thought, I might not buy so many.....maybe....☺

 

Your a lifesaver c-attack, your help has been very much appreciated, I cannot thank you enough! ☺☺☺

[c-attack]

7 hours ago, Strayone said:

When I asked around a few people said "Why bother, a pedestal fan works better", so when I saw a standard pedestal fan does average 2000 CFM

Funny, that is what I do in Summer.  It works on both me and the open case build.  Even in a closed box, using a room fan to disperse the "heat bubble" around the case can knock several degrees off all internal values.  This is extremly important if you keep the case in a corner or other low room airflow space.  

7 hours ago, Strayone said:

As for Coolant Temperature of tubing etc, I was concerned that because of the weather here, the fact the i9-13900K is always trying to push towards 90-100 °C, the RTX 3070 around 70-80 °C and combine that with long term usage and I am not sure if that would make the coolant reach over 60 °C. As I said, this is all new to me when it comes to water cooling which I why I went with an AIO originally. 

It's not as simple as 90C CPU + 90C GPU = 90C water.  The individual device temps are a physical material temperatures resulting from the voltage created heat, less the heat conducted away by the various cooling methods.  Water temp changes are slightly more simple as Watts In, less Watts dissipated through radiator = change in liquid.  Very simple version is takes X amount of watts to raise the temperature of a specific volume of water by 1C. A tiny CPU at 90C but only 45W won't add much heat to the liquid.  A liquid cooled 4090 with the 600W BIOS will add a lot of heat to water even though it may be sitting at 43C at the GPU temp sensor.  

 

Now where things get more complicated is you may often see coolant temp changes because of case air temp changes equal to the change in temp resulting from the CPU or GPU.  We see this most often with a AIO or custom cooled CPU and air cooled GPU.  A big watt model like a 4080/90 that can output 450W+ continuously will heat of the air in the case.  Getting that heat out of there is important and does add degrees to all components in the box.  Now will a 3000 rpm fan do more for this than a 2000 rpm fan?  Possibly, but if 2000 rpm can't do the job than something else must be in play --- super tight dust filters, stacks of HDDs, something else blocking airflow.

 

On that note, a standard 30mm radiator will reduce the free air volume specification of most fans by 40%.  Sometimes it feels like more.  If I put 3x140 fans at 2000 rpm on the case with no restriction, it will feel like a force of nature.  When you drop the radiator in front of it, the airflow speed coming out the other side will diminish significantly.  This has relevance to the AIO radiator location.  

• Top exhaust - The benefit is it dumps CPU heat directly out of the case.  It also leaves the front 3 fans free to blast air through the chamber and across the GPU backplate.  That can have real temp value for VRAM and GPU VRM temps that pass their heat into the backplate.  It also may have some effect on your system DRAM.  The negative is with the AIO rad up top, that part of the case is going to be 2-3C warmer than the front.  If you can't get the GPU waste heat out the back, it will heat of the coolant during gaming.  However, since most games don't actually run more than 60C, you have a lot of headroom to give here with zero loss in performance.  Most 3000/4000 model GPUs are more likely to throttle because of VRM or VRAM temps than GPU Diode temp, so cooling the backplate does have some value.  
• Front intake AIO radiator - Benefit is it moves the radiator out of that top heat zone.  Theoretically slightly better CPU coolant potential because of the cooler intake air on the front of the case versus central air.  This is only likely to matter in some sort of CPU only bench test.  Good move for people than do CPU renders or encoding.  Not necessarily the best move for everyone else.  Negative are the previously mentioned restriction to airflow blowing through the center of the case.  Additionally, all internal components will increase in temp by about the same as the coolant temp delta.  In effect, if you blow 26C air at your RAM and their temp is now 45C, then the RAM temp will be 50C when you blow 31C air in warmed by the CPU.  For most users this will be negligible, but not the best choice if you are a RAM super clocker looking to ring every ounce from them.  DDR5 is fairly temperature dependent.  There may also be a question of hose length reach in a large case like the 7000.  

So which is best?  I don't know.  I can lay out the theoreticals, but ultimately each case is different and I have not done this in a 7000D.  

 

 

7 hours ago, Strayone said:

I was also told to create a positive pressure style cooling system to keep out dust.

This one is never going away and has been an internet staple for years.  If you have a case with substantially more exhaust than intake, it will pull air in through whatever cracks and spaces offer the least resistance.  So in a non-sealed case, you can pull in dust through the gaps.  However, even if you have every fan in your case set to intake and thus the 'ultimate expression of positive pressure', when a particle of dust gets near any of those intake fans, it will enter the case and stick to the first charged surface it comes in contact with.  "Positive pressure" only prevents dust from coming in mesh gaps and other places, not the fans themselves.  

 

The only true way to dust proof a case is put some very tight filters on all intakes and then use a sealed case with no mesh or gaps.  That will prevent dust getting inside.  The bad thing about dust is can increase components temps.  The bad thing about sealed cases and dense dust filters are they increase your case temp from the moment you power on.  No free lunch on this.  Some people are going to be in the dust phobia category and be perfectly happy with a tight sealed case.  Rinse dust filters, don't hear much noise, but that also keeps heat inside.  I am on the other side.  I use open cases and I can blow the weekly dust off in 10 seconds on Sunday morning and be done, while losing all the 'contained heat' from not being trapped in the box.  Most people are going to be in the middle and regardless of what you decide to do with the radiator, you always have control over the intake/exhaust balance by adjusting the speed of individual fan banks.  If you use the radiator on the front, you won't want to blast the top 3x140 + rear in order to maintain "positive pressure".  In fact, you would need to keep them at half the speed of the front radiator fans to do so.  That may be a strike against that placement for you.  In the top position, you deliberately limit the radiator fan speed to a moderate level while blasting the rear exhaust and front intake.  This is an attempt to blow the heat through the case and out the back mesh/exhaust vs sucking too much through the top radiator.  

Edited November 5, 2023 by c-attack

[Strayone]

I have to say, your a treasure trove of information c-attack 

 

9 hours ago, c-attack said:

Funny, that is what I do in Summer.  It works on both me and the open case build.  Even in a closed box, using a room fan to disperse the "heat bubble" around the case can knock several degrees off all internal values.  This is extremly important if you keep the case in a corner or other low room airflow space. 

That "heat bubble" is what concerns me, I thought maybe a powerful intake/exhaust would help move air around better, I still intend to use a pedestal fan & overhead fan until temps reach around 35°C then I will turn on my A/C. If heatwaves become more frequent and longer, A/C may have to be used less and less though to keep power cost down, so that also played into my thinking of a lot of airflow. The problem I found in the past though is if the temperature in the case rises quickly and constantly enough, even with the fans, it becomes a large heat sink, making the entire room heat up quickly, then you end up caught in a constant catch 22 situation. 

 

9 hours ago, c-attack said:

It's not as simple as 90C CPU + 90C GPU = 90C water.  The individual device temps are a physical material temperatures resulting from the voltage created heat, less the heat conducted away by the various cooling methods.  Water temp changes are slightly more simple as Watts In, less Watts dissipated through radiator = change in liquid.  Very simple version is takes X amount of watts to raise the temperature of a specific volume of water by 1C. A tiny CPU at 90C but only 45W won't add much heat to the liquid.  A liquid cooled 4090 with the 600W BIOS will add a lot of heat to water even though it may be sitting at 43C at the GPU temp sensor.  

This has helped alot, I kept getting conflicting opinions on how it all works with no maths or facts behind any of them. Thank you for this, now I have something to work with in calculating my needs for water cooling. After reading this I did a little research into the wattage of both the I9-13900k and RTX 3070. The Intel I9-13900k at defaulf settings is 125 watts base and 253 watts on Turbo. Since the only way to 5.8Ghz is to run on Turbo mode, I am guessing it will be at 253 watts when system is under heavy usage. As for the RTX 3070, it consumes around 220 watts base and overclocking takes it up to around 240-250 watts. I have no intention of OC the RTX 3070, so I guess its 220 watts for that. Considering I am both a gamer and I creates mods and trainers, I am sort of between both worlds on which will be used but since Turbo is used when I play both games and decode/encode, it seems my CPU will be the biggest heat producer unless I am idling. Is there a formula or way to tell how much a AIO or Water cooler can disperse, or a way to find out how much a heat radiator can dissipate, just some formula or guideline to calculate heat produced and heat exchanged I can work with would help me figure this all out would be very helpful.

 

9 hours ago, c-attack said:

Now where things get more complicated is you may often see coolant temp changes because of case air temp changes equal to the change in temp resulting from the CPU or GPU.  We see this most often with a AIO or custom cooled CPU and air cooled GPU.  A big watt model like a 4080/90 that can output 450W+ continuously will heat of the air in the case.  Getting that heat out of there is important and does add degrees to all components in the box.  Now will a 3000 rpm fan do more for this than a 2000 rpm fan?  Possibly, but if 2000 rpm can't do the job than something else must be in play --- super tight dust filters, stacks of HDDs, something else blocking airflow.

 

On that note, a standard 30mm radiator will reduce the free air volume specification of most fans by 40%.  Sometimes it feels like more.  If I put 3x140 fans at 2000 rpm on the case with no restriction, it will feel like a force of nature.  When you drop the radiator in front of it, the airflow speed coming out the other side will diminish significantly.

This was my original (now defunct) plan, based on using 2 AIO with radiators top and front and using only stats to create a theoretical idea:

Noctua NF-F12 & NF-A14 industrialPPC IP67 PWM Fan 2000RPM

Intake
3 x 140mm 2000 RPM     = 322.23   CFM                         3 x 140mm 2000 RPM    = 12.54 SP      
3 x 120mm 2000 RPM     = 215.07   CFM                         3 x 120mm 2000 RPM    = 11.82 SP       
3 x 140mm AR140 Elite  = 58.5-267 CFM                       3 x 140mm AR140 Elite  = 0.48 - 6 SP
Total Intake CFM               = 581.8 - 790.3 CFM
Total Intake SP                   = 24.84 - 30.36 SP

Exhaust
4 x 140mm 2000 RPM     = 429.64 CFM                           4 x 140mm 2000 RPM    = 16.72 SP
3 x 140mm AR140 Elite  = 58.5-267 CFM                       3 x 140mm AR140 Elite  = 0.48 - 6 SP
Total Exhaust CFM           = 488.14 - 696.64 CFM
Total Exhaust SP               = 17.2 - 22.72 SP

=====================================================================================================
Noctua NF-F12 & NF-A14 industrialPPC IP67 PWM Fan 3000 RPM

Intake
3 x 140mm 3000 RPM      = 475.5 CFM                              3 x 140mm 3000 RPM       = 31.56 SP
3 x 120mm 3000 RPM      = 329.7 CFM                              3 x 120mm 3000 RPM       = 7.63 SP
3 x 140mm AR140 Elite   = 58.5-267 CFM                       3 x 140mm AR140 Elite     = 0.48 - 6 SP
Total Intake CFM                = 863.7 - 1072.2 CFM
Total Intake SP                    = 39.67 - 45.19 SP

Exhaust 
4 x 140mm 3000 RPM      = 634 CFM                                  4 x 140mm 3000 RPM       = 42.08 SP
3 x 140mm AR140 Elite   = 58.5-267 CFM                        3 x 140mm AR140 Elite    = 0.48 - 6 SP
Total Exhaust CFM            = 692.5 - 901 CFM
Total Exhaust SP                = 42.56 - 48.08 SP

==================================================================================================

As you can see the theoretical potential of 1000 CFM, half of a pedestal fan was what I was hoping for.  Given your advise, maybe I was being a little over ambitious ☺. As a note though, my plan was never to run them at full speed constantly. I was hoping to use the Commander Pro with its 4 heat sensors, the 2 AIO and CPU & MB sensors, to create a fan speed curve to match the PC needs. For example, if the AIO/CPU heat was getting high, the Noctua fans would be activated to give extra strength airflow when needed and return to a quieter speed when the AIO/CPU where cooler again. Rear Exhaust was to be tied into a internal case temp or maybe if possible, use sensors on Intake & Exhaust, if the difference was getting too high, the rear Exhaust fan would speed up to suck more heat out of the case. Basically a system to activate different fans when needed given the situation, then they return to a quieter slow speed, a backup and support to the AR140 Elites.

 

I was hoping to use the 3000 RPM because I was also concerned  about what you said about the radiators impeding airflow, although I didn't know it was as much as 40%. Given the 10.52 Static Pressure on the 3000 RPM, I was hoping they would easily suck air through the radiator vents without loosing too much CFM airflow but since I cant tie them into the Commander Pro, I will have to fall back to the 2000 RPM. Also, given reading the Pedestal Fan 2000 CFM average, the 158.5 CFM / 10.52 SP on the 3000 RPM was certainly more attractive than the 2000 RPM 107.41 CFM / 4.18 SP or other brand fans.

10 hours ago, c-attack said:

On that note, a standard 30mm radiator will reduce the free air volume specification of most fans by 40%.  Sometimes it feels like more.  If I put 3x140 fans at 2000 rpm on the case with no restriction, it will feel like a force of nature.  When you drop the radiator in front of it, the airflow speed coming out the other side will diminish significantly.  This has relevance to the AIO radiator location.  

• Top exhaust - The benefit is it dumps CPU heat directly out of the case.  It also leaves the front 3 fans free to blast air through the chamber and across the GPU backplate.  That can have real temp value for VRAM and GPU VRM temps that pass their heat into the backplate.  It also may have some effect on your system DRAM.  The negative is with the AIO rad up top, that part of the case is going to be 2-3C warmer than the front.  If you can't get the GPU waste heat out the back, it will heat of the coolant during gaming.  However, since most games don't actually run more than 60C, you have a lot of headroom to give here with zero loss in performance.  Most 3000/4000 model GPUs are more likely to throttle because of VRM or VRAM temps than GPU Diode temp, so cooling the backplate does have some value.  
• Front intake AIO radiator - Benefit is it moves the radiator out of that top heat zone.  Theoretically slightly better CPU coolant potential because of the cooler intake air on the front of the case versus central air.  This is only likely to matter in some sort of CPU only bench test.  Good move for people than do CPU renders or encoding.  Not necessarily the best move for everyone else.  Negative are the previously mentioned restriction to airflow blowing through the center of the case.  Additionally, all internal components will increase in temp by about the same as the coolant temp delta.  In effect, if you blow 26C air at your RAM and their temp is now 45C, then the RAM temp will be 50C when you blow 31C air in warmed by the CPU.  For most users this will be negligible, but not the best choice if you are a RAM super clocker looking to ring every ounce from them.  DDR5 is fairly temperature dependent.  There may also be a question of hose length reach in a large case like the 7000.  

So which is best?  I don't know.  I can lay out the theoreticals, but ultimately each case is different and I have not done this in a 7000D.  

 

 

This one is never going away and has been an internet staple for years.  If you have a case with substantially more exhaust than intake, it will pull air in through whatever cracks and spaces offer the least resistance.  So in a non-sealed case, you can pull in dust through the gaps.  However, even if you have every fan in your case set to intake and thus the 'ultimate expression of positive pressure', when a particle of dust gets near any of those intake fans, it will enter the case and stick to the first charged surface it comes in contact with.  "Positive pressure" only prevents dust from coming in mesh gaps and other places, not the fans themselves.  

 

The only true way to dust proof a case is put some very tight filters on all intakes and then use a sealed case with no mesh or gaps.  That will prevent dust getting inside.  The bad thing about dust is can increase components temps.  The bad thing about sealed cases and dense dust filters are they increase your case temp from the moment you power on.  No free lunch on this.  Some people are going to be in the dust phobia category and be perfectly happy with a tight sealed case.  Rinse dust filters, don't hear much noise, but that also keeps heat inside.  I am on the other side.  I use open cases and I can blow the weekly dust off in 10 seconds on Sunday morning and be done, while losing all the 'contained heat' from not being trapped in the box.  Most people are going to be in the middle and regardless of what you decide to do with the radiator, you always have control over the intake/exhaust balance by adjusting the speed of individual fan banks.  If you use the radiator on the front, you won't want to blast the top 3x140 + rear in order to maintain "positive pressure".  In fact, you would need to keep them at half the speed of the front radiator fans to do so.  That may be a strike against that placement for you.  In the top position, you deliberately limit the radiator fan speed to a moderate level while blasting the rear exhaust and front intake.  This is an attempt to blow the heat through the case and out the back mesh/exhaust vs sucking too much through the top radiator.  

Based on your advice, I might move the Hi70i radiator to the roof, given I am getting a XT version, I am hoping the longer tubes will help. I do get fairly involved with games but I am also into creating mods and trainers. So I do use Hex Editors, and other decoding/encoding programs so CPU can be put under a bit of stress with them but I try to conserve resources by running as little programs as possible etc. But that said, as you have pointed out, that heat generation will be returned back to the system if I run it at the front. As you pointed out, as long as I can quickly exhaust out the back heat from the GPU, the top radiator shouldn't be too badly affected, and with a decent burst of cool air coming from the front Intake, a top mounted radiator should work much better than a front mounted radiator in keeping my system cool. As for Positive Pressure system, I will abandon that idea, since the 7000d Airflow is made for airflow, with top, side and front dust filters. I usually maintain my computer to get rid of dust on regular but I was concerned with an increase airflow brings more dust. So I guess I will just stick to a more stricter maintenance schedule.

 

Once again, thanks for you help c-attack, you have help immeasurably. You have saved me and my computer quite a few headaches and solved quite a few problems with your advice. My hat's off to you my friend, I cannot thank you enough for all this. 

[c-attack]

12 hours ago, Strayone said:

I am guessing it will be at 253 watts when system is under heavy usage. As for the RTX 3070, it consumes around 220 watts base and overclocking takes it up to around 240-250 watts.

You will only hit 253W if you are running some type of synthetic benchmark or CPU render (like Cinebench).  The most I ever see in normal usage is around 160-170 with all cores at 5.7 GHz.  CPU wattage tends to fluctuate quite a lot and the CPU power algorithms are very sophisticated.  Continuous sustained CPU power is not typical outside of specialized tasks.  Even when you are under a Intel limited max 253W, it will take some time to raise the coolant temp to the point where it has adverse effects.  +1C to coolant temp = +1C to CPU temp.  No need to worry over +2-4C if the CPU temps are below your personal limit.  Knocking 4C off the liquid temp is no easy feat.  

 

Anything water cooled works by physically conducting the heat away from processor, then dumping it in the water stream.  That makes the radiator fans something more like taking out the trash to curb.  You don't need to do it every time you drop in a piece of garbage (heat), but if you don't do it it will build up and cause problems.  Because load cycles vary, the radiator fans simply need to keep a reasonable pace to keep taking out the trash.  It's totally fine to use the Commander and it's temp sensors for fan control, but you won't need to make it super reactive to changes.  

[Strayone]

3 hours ago, c-attack said:

You will only hit 253W if you are running some type of synthetic benchmark or CPU render (like Cinebench).  The most I ever see in normal usage is around 160-170 with all cores at 5.7 GHz.  CPU wattage tends to fluctuate quite a lot and the CPU power algorithms are very sophisticated.  Continuous sustained CPU power is not typical outside of specialized tasks.  Even when you are under a Intel limited max 253W, it will take some time to raise the coolant temp to the point where it has adverse effects.  +1C to coolant temp = +1C to CPU temp.  No need to worry over +2-4C if the CPU temps are below your personal limit.  Knocking 4C off the liquid temp is no easy feat.  

 

Anything water cooled works by physically conducting the heat away from processor, then dumping it in the water stream.  That makes the radiator fans something more like taking out the trash to curb.  You don't need to do it every time you drop in a piece of garbage (heat), but if you don't do it it will build up and cause problems.  Because load cycles vary, the radiator fans simply need to keep a reasonable pace to keep taking out the trash.  It's totally fine to use the Commander and it's temp sensors for fan control, but you won't need to make it super reactive to changes.  

I was under the impression the wattage was set, once turbo mode engaged (which is when any heavy load is placed upon it) it immediately switched to 253 watts. Nice to know its a bit more flexible, is the GPU the same way? Nice to know also that I wont need a reactive response style system. Having only dealt with air CPU coolers before its hard to get this mentality of expecting heat spikes out of my head. The AR140 Elites where to be the "trash removers" but I wasn't sure it would be enough given environment and these new 13/14th gen Intel chips I have heard and read produce enormous heat even with a top quality AIO. The 4 thermal Sensors included and the control of fan speed with the Commander PRO appealed to me for safety of the system and ability to lower speeds when not needed to keep down noise, with the fans running from 38-52 dB on full speed.

 

Thanks again c-attack, your knowledge and advice is always very much appreciated!

[Strayone]

9 hours ago, c-attack said:

You will only hit 253W if you are running some type of synthetic benchmark or CPU render (like Cinebench).  The most I ever see in normal usage is around 160-170 with all cores at 5.7 GHz.  CPU wattage tends to fluctuate quite a lot and the CPU power algorithms are very sophisticated.  Continuous sustained CPU power is not typical outside of specialized tasks. 

Just a side check on that, I got some conflicting info from Gamers Nexus review of the chip, apparently unlike the 12th Gen Chips, the 13900K is pulling full power all the time whenever under load:

https://www.youtube.com/watch?v=yWw6q6fRnnI

 

Relative information is found at 3:06 into the clip. Apparently 12th gen and previous gen chips had a 50-60 sec window and turbo was reduced to conserve power but not so in the new 13/14th gen Raptor Lakes Chips. I don't think it will affect me too much, but if this is true according to Gamers Nexus, its 250 to 290 watts constantly as soon as Turbo mode is engaged, quote "full power all the time".  I am concerned that when i am decompiling, using 32 threads might push her to heat up but like I said, I need to get this heat spike mentality out of my damn head. 

[LeDoyen]

12 and 13th gen boost the same way. both have the 56 seconds boost before falling back from 253 to 125W.

The issue may be that there is a difference between testing "stock", and testing "out of the box".

"Stock" would be with the intel limits. In that regard all gens boost the same way.

"Out of the box" is a free for all. That's where GN testing is basically. Some motherboard bioses remove the intel limits by default and let the CPU boost as much as the cooling allows, which is where you see figures above 253W. Their test was on an Asus board, and Asus is notorious for disabling limits and enabling auto overclock (Multicore enhancement enabled, PL1 and PL2 power limits set to 4096W, basically disabled..), so their test is absolutely not Intel stock, it's heavily overclocked. It's not a bad test as they show you what the average consumer would face without going to fiddle in the bios.

So no matter what motherboard you get, you have to go in the bios to at least check if the intel limits are in place (PL1 = 125W, PL2 = 253W / 56 seconds). Otherwise you may be in the GN case where the thing goes too high and is impossible to cool.

If you have good cooling, you can then manually change PL1 to a higher sustained load value and gain some time on your long workloads.

[Strayone]

3 hours ago, LeDoyen said:

12 and 13th gen boost the same way. both have the 56 seconds boost before falling back from 253 to 125W.

The issue may be that there is a difference between testing "stock", and testing "out of the box".

"Stock" would be with the intel limits. In that regard all gens boost the same way.

"Out of the box" is a free for all. That's where GN testing is basically. Some motherboard bioses remove the intel limits by default and let the CPU boost as much as the cooling allows, which is where you see figures above 253W. Their test was on an Asus board, and Asus is notorious for disabling limits and enabling auto overclock (Multicore enhancement enabled, PL1 and PL2 power limits set to 4096W, basically disabled..), so their test is absolutely not Intel stock, it's heavily overclocked. It's not a bad test as they show you what the average consumer would face without going to fiddle in the bios.

Thank you for clarifying that and for the information on setting the BIOS. I had in intended to get an ASUS Motherboard after getting told they are more durable and tuned towards performance with more frequent and stable BIOS updates etc. I have always used ASUS in the past and have never had a problem before so I stuck with them. I was not aware they can have the intel power limits removed by default, in the past I have left it set to factory default, not one to venture into overclocking etc unless it was necessary so that was some very useful information, thank you. That does clear up the issue of heat control etc if it was going full speed/heat all the time while its under load because then I would have to beef up the cooling. 

 

3 hours ago, LeDoyen said:

So no matter what motherboard you get, you have to go in the bios to at least check if the intel limits are in place (PL1 = 125W, PL2 = 253W / 56 seconds). Otherwise you may be in the GN case where the thing goes too high and is impossible to cool.

If you have good cooling, you can then manually change PL1 to a higher sustained load value and gain some time on your long workloads.

Thank you for this, since I am not one to fiddle with CPU or RAM voltage and settings, I have never been quite confident in my skill and knowledge in this area. With your previous comment came the immediate thought to reinstate limits and straight away you gave me the information I needed, much appreciated! I bought a h170i Elite LCD XT as I was told it was the best and has the longer tubes I will need. So hopefully when all the parts for the computer arrive and I get it up and running, thanks to c-attack and your advise, she will run like a dream.

 

Thank you LeDoyen, your help also is very much appreciated, the peace of mind you and c-attack bring certainly helps save me from a world of stress, worry and a possible visit to a psychiatric hospital. Nice to have somewhere to turn to when I get out of my depths, thanks again! 

Edited November 7, 2023 by Strayone