Fan set up in modded C70.

[wgusler]

New guy to the Forums here, did allot of modding and PC builds 15+ years ago and then moved on to other things. I am doing some small mods to a Corsair C70 and a build and have a question on using a H60 water cooler.

The C70 has two 120mm intake fans at the front and one exhaust at the rear in stock configuration. I have moved the intake fans to the front of the case (they were mounted behind the HD cages) and plan on mounting the H60 where the rear fan is now. I have replaced the window and will have airflow restrictions on the top fan mounts.

So, should I just mount the cooler fan and blow out the back or follow Corsairs recommendation and mount fan first and rad second to push fresh outside air in? Don't thing the hoses on the H60 will reach for a bottom mount. Don't like the idea of intake air from the rear. With the H60 rad pushing in from the rear, would a 140mm fan in the rear mount on top of the case blowing out work?


Computer specs:
I7 4790K / H60 cooler
Gigabyte Z97 UD5h
8MB DDR3 1866 Gskill Sniper
Corsair CS650M power supply
MX100 128 & 512 SSD'S
Seagate 1TB 7200
Asus GTX980

 

[NatePo717]

So I did water cooling in my c70 back when I was using it for my main case. I did a custom loop. Had the following setup.

Top: Pull exhaust on a 240mm radiator fans on case side
Bottom: Push/Pull 240mm radiator intake
Back: 120mm radiator exhaust fan on case side
Front: 2 fans intake between face and chassis.

I got pretty good temps that way. Though I didn't really try other configurations.

 

[wgusler]

Thanks Nate. My side panel and the front fan on the top of the case will not be avalible, so I have run thru the combinations for this.



This is my idea I want to do :

or this:


here is why I have no side fans

 

[BonzaiDuck]

Thanks Nate. My side panel and the front fan on the top of the case will not be avalible, so I have run thru the combinations for this.
This is the configuration that Corsair recomends:


This is my idea #2 using the above setup;

what I want to do #3:

or this:#4


here is why I have no side fans

"Gun! Everybody hit the deck!" D: What's that on the wall: an AR-15?

Seriously, though. I think you have the right idea, but it boils down to this: What goes in -- must come out.

I'm skeptical about an H60 being sufficient for a Devil's Canyon. Maybe you don't plan on overclocking it though . . . Just my opinion, but I'd pick a double-sized AiO like an H110 for this. And the problem is fitting it in that case . . . ah . . ammo box. . .

I notice this case is vented on the bottom -- something I didn't see at the Newegg showcase for the C70. You'd fit a PSU in the bottom rear, but there's no drive-cage obstruction in the front. wondering if you could fit a 240mm AiO forward from the PSU.

But I agree about blocking off unused vents, as you plainly seem intent doing to the top vents. [or am I wrong?]

It might not much matter if you're pushing warmer air into the case with a cooler at intake, if there is sufficient airflow in and out of the case.

Have you yet installed this H60 on the i7-4790K? What do the temperatures look like under stock loading?

 

[thescreensavers]

Front/Side in

Top/Rear Out

Is How I do it.

 

[wgusler]

Thanks for the reply. I haven't built the board or ran this yet, and I don't plan on overclocking it. The top of the case will be covered with a bunch of linked .50 cal ammo, cus what good is an ammo box without ammo?

I don't know if the hoses on the 60 will reach the front, but I can check that out. They might reach if I move the drive cage to the bottom and mount the rad above.

This will blow the hot air on the GPU, but the Asus 980 with the ref cooler that blows out the rear and runs pretty cool. ( not overclocking it either)

I can make a shroud to direct the air from the rad up away from the GPU but I don't know..............

 

[wgusler]

I checked, the H60 hoses will not reach the lower front, only way is moving the drive cage down. Or mount the rad in the top rear blowing out?

 

[wgusler]

this?


Add a lower 140 fan to overpressure the case ?

 

[BonzaiDuck]

this?


Add a lower 140 fan to overpressure the case ?

If you want to pressurize the case, then block off the intakes that are not being used for anything. Also, drawing air from the floor would be better if you could lift the case up an inch or two. Me -- I just install double-wheeled braked casters on the case bottom. You can also buy "tall" hard-rubber "feet" at an electronics warehouse.

Also, if you're going to add bottom intake, why not use two 140's? If all the air is going through the cooler which you deploy as an exhaust, that would be better. If you keep the rear exhaust along with mounting the radiator/fan on top, you CAN build a duct that would pull air through a narrow aperture across the motherboard -- see another recent thread in this forum where I explained that in detail. It would also assure more pressure outside the duct and in the case.

There have been a few examples you'll find in this forum of small cases with great custom water-cooling, but it's a lot of work, a tradeoff between storage and radiators in the case.

I'm building a spare computer with air-cooling that should be better than an H110i AiO using the EVGA ACX heatpipe tower. For your case, I'm guessing that would be easier than this. You could make everything but the exhaust an intake fan, seal the case, duct the cooler to the rear exhaust, and it would still be quiet -- especially with motherboard thermal fan control. The ACX is going to run you about $55+ with shipping and maybe tax.

You're on the right track, though, even for using the H60.

 

[wgusler]

Ok, change of plans. Removed both drive cages and ordered 2 HD mounts.

I am going to mount the SSD's in one and the back up 3.5 in the other and mount them on the floor in front of the PS. Ordered 2 AF140's HP and will alter the case to make them both fit in the front. Going with the stock H60 fan, sealed and with a 25mm shroud pulling out the rear of the case. Top and bottom vents will be sealed, forcing the 120+ cfm from the 140's out the back. We will see how this works.

With the drive cages out of the way, there should be good airflow across the board and GPU. I will update with the results. Thanks for you help guys!

 

[wgusler]

Here is what I came up with.

I have modded the front to fit two 140MM Corsair QE fans and opened the back of the case at the exhaust to eliminate any restrictions. 2 Gentle Typhoon AP-15's 120mm in a push / pull on the H60 with a homemade shroud.







Found that the Corsair stock H60 fan dosen't fit flush on the radiator, it has a small gap on all 4 sides that will hurt the airflow in a pull config. Ordered the GT as they are completely square and should have as better seal. Made a shroud and installed with a rubber gasket, still has a small gap. Sealed the shroud to the rad with silicone.


As you can see, no window = I don't give a s*** what it looks like on the inside.

 

[BonzaiDuck]

This will blow the hot air on the GPU, but the Asus 980 with the ref cooler that blows out the rear and runs pretty cool. ( not overclocking it either)
. . .
I can make a shroud to direct the air from the rad up away from the GPU but I don't know..............

It depends on how fast in CFM the air is exhausting from your case. The real issue is one of how many degrees C increase over external and internal air ambient does it take to reduce cooling effectiveness of a radiator of a certain size and capacity by 1C?

Just off the top of my head -- which is a little "funny" this time of the morning -- I think you can compensate by upping CFM and probably control the fan dynamic thermally -- maybe just linking it to CPU temperature.

It's a shame those GT AP-30's aren't available anymore. I know one member who stockpiled the suckers, and now? -- Boy, is he havin' fun, I bet.

I only bought two, and after that, I'll either have to use those fans on the best project and replace with something less, or try and compensate by buying still-available AP-15's. And by the way . . .

Did you get those in the 3-pin or PWM flavor? On all the later motherboards I know of, you can use both with the onboard fan-controller. But the PWM option doesn't require anything more than available CPU or CPU_OPT fan ports and a $10 splitter. Were we talking about the Swiftech 8W-SPL . . . whatchamacallit? so no fan voltage draw from the motherboard. That would free the controller to work better with every 3-pin CHA_FAN port you need.

Here's a question for all the pizza-cists.

"What is the effect on CFM of stacking two identical fans, push-pull and front-to-back?" Is there mostly just an increase in final exhaust pressure or maybe static pressure? Or could it possibly enhance CFM, if you could somehow increase overall intake surface? You could build a narrow box between the fans, and jerry-rig a means of porting another fan to it.

But the main question deals only with a stacked-fan configuration.

 

[BonzaiDuck]

I just looked again at your last post, wgusler.

This is getting interesting. I see the method in your madness. I KNOW where you're goin' with this!!

I want to see how this turns out. Did you -- perchance -- take any baseline data with all your hardware and the H60 in that case? Got notes?

But anybody could simply match a processor or even mobo to your configuration and report some thermal data.

The 140's in the top as intake? Was the case too small for even 180mm fans? There are some good 140's that seem to live up to a CFM rating of >= 100 CFM. Quiet, and PWM-controllable.

This isn't just some guy who turned his case into an ammo-box, folks!! This could turn out pretty good -- very well -- capital! I want to see the thermal statistics on this sucker when you're finished with it and "up and running."

 

[BonzaiDuck]

So . . . . what's with the cheesecake?

Moderator?

That's fine, but if you want to share exciting pictures, run along to the glossy mag site.

We're discussing hardware here. Softwear -- that's somethin' else.

 

[wgusler]

No baseline as I haven't even bought the MB and CPU. I will run the H60 with shroud sealed in the pull only and then with the push fan added. I plan on using a Gigabyte x97 UD5H with fan control in the software. Might set up the H60 fans to thermal control and the case fan intakes to a temp sensor inside the case.

Would be nice to have a single software system to monitor all the temps at once( CPU, GPU, MB chips, drives and inside case temps) to create a termp/ fan speed config base on any max temps or a ave temp across all sensors. I need to sniff around the nets and get up to speed on fan conrols/ software. Like I said, been out of building PC for a loooooong time.

 

[BonzaiDuck]

No baseline as I haven't even bought the MB and CPU. I will run the H60 with shroud sealed in the pull only and then with the push fan added. I plan on using a Gigabyte x97 UD5H with fan control in the software. Might set up the H60 fans to thermal control and the case fan intakes to a temp sensor inside the case.

Would be nice to have a single software system to monitor all the temps at once( CPU, GPU, MB chips, drives and inside case temps) to create a termp/ fan speed config base on any max temps or a ave temp across all sensors. I need to sniff around the nets and get up to speed on fan conrols/ software. Like I said, been out of building PC for a loooooong time.

That's why we share experiences here at the forums.

I was on a "vision quest" for maybe six years for thermal fan control. First time I "did it right," I had a 120x38mm Delta Tri-Blade on a heatpipe cooler and a Springdale ASUS board -- p4p800-SE I think it was. I could only control the CPU fan. I spent too much change on controllers. I started backing away from SpeedFan: the project has attempted almost to do too much -- keep track of all the sensors on all the boards ever made.

Gigabyte has a proprietary software called "EZ-Tune," and it has a module or feature that works similar to ASUS AI Suite and either Fan Xpert or Thermal Radar -- to set fan curves for duty-cycle plotted against temperature C. On the ASUS board, one has both CPU and Chassis fan control -- for two fan curves. Both controlled according to CPU temperature. There's a chance that the EZ-Tune software also allows for multiple fan curves. The only version I'd worked with belonged to an LGA-775 Gigabyte board.

I'd say you should plot a course on this issue with a goal of avoiding extra software, extra sensors, etc. An ingenious solution will be the most simple.

You can GET monitoring software that serves up everything, and first I can think of are CPUID HWMonitor and Aida-64. There should be a limited trial version of the latter, but you have to buy the full enchilada.

Point of it: Unless you're setting up a peculiar airflow situation as pertains to GPU, VRM or chipset motherboard components, you shouldn't need to do more than just "monitor" those. More recent GPU cards come with their own monitoring software and features for defining fan curves of the GPU.

I find it very fine -- even elegant -- to define a fan curve for all the chassis fans and a fan curve for all the CPU fans. But both groups of fans ramp up according to CPU temperature. They just follow different fan curve schedules or sets of RPM/temperature points.

However, if you want to at least LOOK at add-on, "smart," thermally controlled fans by a front panel controller or even a simple little circuit card, look at the Aquaero units:

http://www.frozencpu.com/search.htm..._id=oFIcSjJh&searchspec=Aquaero&go.x=0&go.y=0

Again -- adds a level of complexity. But it (the Aquaero 5 units) communicate with the motherboard and CPU via internal USB connection (as I recall). Otherwise, with some controllers you'd be "on your own" with separate analog and digital sensors, harder to place and measure core temperatures. And even with the Aquaero, it's another device, which connects to additional "devices" (fans and pumps). Better if you can organize the fans, distribute them so as not to overload the mobo fan ports, or -- with PWM fans, you can control them from a splitter connected to the CPU_FAN PWM port and powered directly from the PSU 12V Molex or SATA. the Swiftech 8W-SPL-PWM-SAT model has ports for 8 PWM fans.

Find us a Gentle Typhoon AP-30 -- preferably with PWM all wired and ready -- I can show you what you can do with your duct and that lil' ol' H60!! But you're going to want to wrap the assembly in acoustic-deadening foam rubber . . adhesive padding.. . . . The closest I can come to an AP-30 as I scan the online stores is a 120mm Noctua "Industrial" IPPC 3,000 RPM fan . .. which . . . looks promising. PWM, all the features and reputation you'd need to build with confidence. And ONLY 41 dBA noise rating!!

Peace!! Through superior . . . fire . . . power!!

Incidentally, now that I scroll back through your photos. Get some rubber fan mounts for those top fans. If you have some strategy about a custom panel to cover them with a grille, fine -- but the little screws and nuts will propagate fan vibration and motor noise throughout the case. You'll be glad you spent the chump-change on those little rubber suckers. Sidewinder Computers has them; so does FrozenCPU, I think.

 

[wgusler]

Sorry so late with a reply, was gone for a few on a thanksgiving trip to Iowa. Thanks for the info. I looked at the Aquaero 5 LT, without the display and it looks like that may be a good solution without having to buy 4 new PWM fans.
I like the fact that I can spool up or down 4 separate fans on different curves with their software. I can set the push pull fans on the H60 with the CPU temp only, and then the front fans on two different sensor curves: #1 on the MB/case temps and #2 on the GPU temps. Or run both intake front fans as one and use two curves on the push and pulls on the H60. I am curious if there is a "sweet spot" on the push pull by changing the fan speeds independently of each other.
With one fan control option left on the Aquaero...............the GPU fan? I can't tell from their documentation if the GPU fan can be controlled if not directly plugged into the Aquaero. I am not comfortable tearing into my new 780ti ref cooler.
The H60 rad, shroud and fans are all set up and in the case, and I feel like the sealing around shroud and fans should help quite a bit on the efficiency of the system. On automotive cooling systems, the shroud ( in a pull set up) is very sensitive to air leaks between the core and the fan, makes sense that it would be the same here.
Not having a baseline to compare to sucks, but I am not tearing it apart to put it in a stock configuration just to test. I imagine there is enough data on water cooled 4790K's to get a good idea on how well my system is cooling.
I will have the computer built by next weekend and will have some results after.

 

[XavierMace]

I know I'm late to this party, but the only thing I'd mention about the setup you are doing is that you're degrading the H60's performance by pulling hot air through it. That's why Corsair gave the recommendation they did. I'm sure it's still going to be up to the task, but it's far preferable to pull cool air over it rather than hot air.

 

[wgusler]

I know that it is not ideal, but that is why I am bringing in more intake and using the ref cooler 780ti. With no side or top fans avalible, it needs to flow front to back.

 

[BonzaiDuck]

I know that it is not ideal, but that is why I am bringing in more intake and using the ref cooler 780ti. With no side or top fans avalible, it needs to flow front to back.

I said it before. I could be right, I could be wrong. But it all depends on overall CFM going through the case, pressurization as well. With enough CFM, case interior ambient might not be much more than external ambient to any measureable degree. If there were actually a difference, it would be insignificant to any improvements in airflow that would further enhance the actual efficiency of the cooler.

Now if you're going to have three graphics cards, that would add something no matter how you installed the radiator, but then as I said -- enough CFMs with cards exhausting through the PCI plate more than they dissipate in the case -- it might not matter there so much either.

 

[wgusler]

Bonzi,



Just ordered the Aquaero 5 LT controller, heat sink and mounting bracket. I like the idea of a "virtual" sensor by combining multiple sensors and setting a curve off of it. I will also be able to tie individual fans to sensors: lower intake fan to the board temp and the upper intake to the GPU sensor and set curves for each.



May install 4 sensors on the H60 intake, exhaust, inside the case and outside the case for ambient. I foresee many hours playing with fan speeds and curves



Thanks for the tip.

 

[BonzaiDuck]

Bonzi,



Just ordered the Aquaero 5 LT controller, heat sink and mounting bracket. I like the idea of a "virtual" sensor by combining multiple sensors and setting a curve off of it. I will also be able to tie individual fans to sensors: lower intake fan to the board temp and the upper intake to the GPU sensor and set curves for each.



May install 4 sensors on the H60 intake, exhaust, inside the case and outside the case for ambient. I foresee many hours playing with fan speeds and curves



Thanks for the tip.

Just my opinion, but with the smaller case, simple with elegance may work well. Even so, I'm testing a late-entry, end-of-lifecycle Z77-A board right now. It only has CPU_FAN, CPU_FAN_OPT only maybe, and CHA1 and CHA2 fans with an uncontrollable PWR FAN. I see the ASUS software is still divided between CPU and CHASSIS, but you can control the two CHA headers separately just from within the BIOS.

If I thought I needed a separate controller, though, I'd pick the Aquaero 5 if just the circuit-board.

But it should work for you, no less for the chosen case.

You know? I opened up a Dell system that belongs to my dentist (and we're negotiating on these fillings, so . . I do what I can.) This is a Dell PRecision T3400. They had done with their proprietary heatsink what you are doing with the H60 and what I do with my heatpipe coolers.

So I'm interested very much in how this works out for you. I'm looking at a custom water rig, and I have my own ideas how to hybridize the best of air and water. I want to see what load temperatures you're going to get with more CFM throughput at a given thermal wattage. Or just a known wattage.

 

[BonzaiDuck]

Just another thought, but I'd like to see this thread continue with ideas of OP and others.

This is an opinion, and I have a bias for using a HAF midtower. There WILL LIKELY be a noticeable difference between room-ambient and internal case ambient at IDLE. And like someone said, maybe XavierMace, it could even be a delta of 20F -- translating into something like 10C.

BUT!! BUT!! If the fans are thermally controlled, there is sufficient or even overpowering case intake CFMs, and ALL the case exhaust and intake fans follow CPU temperature, the mid-range or LOAD delta between ambient and case interior will dynamically attenuate, and the cooler's potential efficiency will not be reduced at temperatures where that efficiency matters!! I'm interested in seeing the results of this H60 configuration!!

 

[wgusler]

Yes, that is why I am thinking of using the temp sensors included with the board to do an inside/outside temp reading and before and after on the H60.

I am curious about the thermal extraction of the H60, what is the increase in air temp from the intake to exhaust and can this be increased or decreased linearly by fan speed. Or what diff the push pull vs the pull only in heat extraction.

Want to test the same thing on the GTX 780 Ti blower cooler, does fan speed/airflow have a linear curve on heat extraction? I think there may be a drop in the amount of extraction at higher fan speed.


For example: GPU at 100% runs at 80C with the blower fan at 25%. The GPU runs at 70C at 50% fan speed. So from a purely logical standpoint this would continue at 60C at 75% fan speed and 50C at 100%. That isn't going to happen in the real world.


I do however believe that somewhere in the upper fan speeds there is a diminishing return on cooling vs fan noise and power consumption. Also can the air flow can be too fast? I do know that the thermal exchange in an air conditioner varies with air flow. If you test the temp on your car AC at high and then medium with no other changes, the temp will be lower at the medium blower speed. Does this work in reverse? It should, so can a fan at 80% speed remove more heat than the same fan at 100%?


Without PWM fans, testing this may be difficult. Hopefully the Aquaero will allow fan speed changes with some degree of accuracy.

 

[wgusler]

Installed the Aquaero 5 LT and had a thought: To ensure overpressure on the intake side, what if I mounted an additional 140mm PWM intake fan on the bottom like below.



I can control the H60 fans individually and the two intake fans as one, leaving the 4th fan header on the LT (PWM) for the added bottom intake.

I can build a curve in the software based on internal case temps or a virtual sensor with the CPU/GPU and run it from that. Quite at 800RPM at idle and then cranked up under load.

Now I need to find a monster airflow 140MM PWM fan!