Question PC Airflow Management

[mrjyyyy]

So I have 6 ID Cooling PWM Fans Rated at 56.5 CFM max Airflow and 900 - 2000 rpm

Layout:
3 are on front of the case for intake
2 are on top for exhaust
1 is on back for exhaust also

The 3 intake fans are connected to one Sys Fan Header on the mother board
The 3 exhaust fans also connected to one Sys Fan header but different to the header of intake fans.

The question is, I want to achieve a positive pressure layout to minimize/trap the dust to a dust filter on intake fans, does manually setting the intake fans to higher rpm and setting the exhaust fans to smart mode will create a positive pressure inside the case?

Does setting the rpm low / high have direct impact to its airflow rating?

Thank you!

 

[tennis2]

The amount of air a fan moves is directly related to its rotational speed. Think of a ceiling fan.
If all the fans are the same, it's easy to do the math.

 

[Paperdoc]

There is another way that uses all of the automatic fan speed control capabilities of your mobo. Then I'll offer a tip for fine-tuning your settings.

To start, a little background. The air flow ratings of fans are given for when they are dealing with absolutely no restrictions of air flow. But of course in real use, you will have dust filters on the intake fans, but none on the exhaust units. So, simply having all your fans identical and operating at the same speeds, your exhaust capacity is greater than intake capacity, meaning you probably will have slightly negative internal pressure. The solution is to ensure the exhaust fans run more slowly to create the positive pressure you seek.

On almost all mobo fan headers one option for their setting is to create your own custom "fan curve" specifying fan speed for any given temperature at the sensor. You already have your fans grouped suitably so that you could create slightly different fan curves for the front fans on one header, and the other fans on other header(s). To start, you could leave the front fans as they are, which uses a pre-set fan curve. You MAY be able to see that in the display of fan configuration for that SYS_FAN header. Now, go to the header(s) used for the three exhaust fans. For each, choose to set your own custom curve, and set that to run the fans slightly SLOWER than what the intake fans are doing over the entire range of temperatures. SAVE and EXIT, then let your system run and check the airflow direction (below) under several workload conditions. If you still have negative internal pressure, make come changes to your custom curve(s) to slow those exhaust fans more. Trial and error to fine-tune your settings will get you there, and the fans then ALL can be automatically adjusted by your mobo control system and still maintain modest positive pressure under varying load conditions.

To test air flow, try this. You need a source of smoke as a tracer. I have used an incense stick, but maybe a cigarette will do. Get the system working. Slowly move the smoke source around the outside of the case away from the fans, but near small openings and cracks where air can leak. Watch which direction the air (smoke) flows. If it flows away from your case, there is positive pressure inside preventing dust intake at those leak points. If smoke flows into the case, you have negative pressure.

 

[mrjyyyy]

There is another way that uses all of the automatic fan speed control capabilities of your mobo. Then I'll offer a tip for fine-tuning your settings.

To start, a little background. The air flow ratings of fans are given for when they are dealing with absolutely no restrictions of air flow. But of course in real use, you will have dust filters on the intake fans, but none on the exhaust units. So, simply having all your fans identical and operating at the same speeds, your exhaust capacity is greater than intake capacity, meaning you probably will have slightly negative internal pressure. The solution is to ensure the exhaust fans run more slowly to create the positive pressure you seek.

On almost all mobo fan headers one option for their setting is to create your own custom "fan curve" specifying fan speed for any given temperature at the sensor. You already have your fans grouped suitably so that you could create slightly different fan curves for the front fans on one header, and the other fans on other header(s). To start, you could leave the front fans as they are, which uses a pre-set fan curve. You MAY be able to see that in the display of fan configuration for that SYS_FAN header. Now, go to the header(s) used for the three exhaust fans. For each, choose to set your own custom curve, and set that to run the fans slightly SLOWER than what the intake fans are doing over the entire range of temperatures. SAVE and EXIT, then let your system run and check the airflow direction (below) under several workload conditions. If you still have negative internal pressure, make come changes to your custom curve(s) to slow those exhaust fans more. Trial and error to fine-tune your settings will get you there, and the fans then ALL can be automatically adjusted by your mobo control system and still maintain modest positive pressure under varying load conditions.

To test air flow, try this. You need a source of smoke as a tracer. I have used an incense stick, but maybe a cigarette will do. Get the system working. Slowly move the smoke source around the outside of the case away from the fans, but near small openings and cracks where air can leak. Watch which direction the air (smoke) flows. If it flows away from your case, there is positive pressure inside preventing dust intake at those leak points. If smoke flows into the case, you have negative pressure.

Thanks for this, I was also thinking about this but my problem is how should I test if it's working, Great suggestion, will definitely try those steps, I was also thinking if I will leave the intake fans at 90% to max speed to ensure the positive pressure, and leave the exhaust fans on smart fan mode, Is it advisable?

Also I am thinking if it's balance since my bottom intake fan is directly blowing to my psu shroud, so i dont think it contributes to the airflow.

 

[Karadjgne]

Some software operates their curves by temp. Now those headers may or may not be linked to different sensors, being in different areas of the mobo, or maybe tied to cpu temps. So you'll have to test your airflow at idle and loads before changing anything. Given cpu temps as source, simply changing the curve on exhausts to respond to a higher temp (default for max speed is 70°, so setting exhaust to 80° will lower rpm at cpu temps).

If source is mobo sensors, you'll have to see what the temps really are, gaming gpus will drastically change some sensors in that area, but don't do much for temps near the intakes, so this can require serious adjustment of curves to keep intakes proportionally slightly faster than exhausts.

 

[Paperdoc]

All automatic fan speed control systems built into a mobo are based on measured temperatures. In fact, although we call them "fan speed controls", they really are TEMPERATURE control systems. Each focuses on the actual Temperature measured at a sensor, and manipulates the fan speed to whatever it takes to keep that temperature at the target. A key item here is the temperature SENSOR used for each control system. The CPU_FAN header always uses ONLY the temperature measure by a sensor built into the CPU chip by its maker, and sent out on one chip pin to the mobo. Normally, SYS_FAN or CHA_FAN headers use instead a sensor built into the mobo by its maker and positioned to emphasize a particular component area selected by its designers. SOME mobos offer to let you use the CPU's internal sensor instead, in case you want to re-configure a header to cool something related to the CPU directly. And then SOME mobos also include extra sensors associated with other key components (say, the power regulation section or the North Bridge chip) and allow you to use those sensors for some fan headers in case you wish to arrange a fan directed at those components for focused cooling.

VERY few systems allow you access to temperature sensors on a graphics card because there are no "standard" ways to do that. There are a few mobo makers that allow this IF you use one of their graphics cards with non-standard features. None of them have cooling of the graphics card to be done by the mobo - that is done by the graphics card itself. For this purpose some graphics card drivers provide tools to monitor and adjust that internal cooling control system and "see" the temperatures on that card, but the mobo does not know about that data. However, as Karadjgne says, the flow of heat from of the graphics card MAY affect adjacent areas of the mobo and case.

OP, if your bottom intake fan feeds into a chamber for the PSU that really does not blow further into the case, then its is not part of the airflow you are working on. However, if the air though the PSU then does proceed into the case, it MAY not contribute much to cooling the case, but it DOES contribute to total air flow and pressure balance inside.

Your original plan - to keep all intakes at full speed or nearly and let the mobo automatic system control only the exhaust fans - WILL give you positive internal pressure. I only suggested the more complicated approach of developing custom fan curves as a means of reducing the intake fans' speeds and noise generation when your workload is low. If that is not important to you, your simpler plan will work well.

 

[Karadjgne]

Heh yeah, my pc used to be confusing. It uses mobo sensors in areas, sys_fan1 is somewhere near the socket. Up until recently I had a 280mm aio as intake. Now in cpu intensive stuff, cpu_fan would increase the intakes on the rad but that isn't enough heat to really change sys_fan1, so exhausts stayed low, positive pressure for sure. But in simpler games with high graphics, the heat from my seriously OC'd gtx970 would drive sys_fan1 crazy and while intakes spun relatively slow (600rpm) the exhausts (2x 140mm) would be upwards of 900rpm. Nice negative system.

Sometimes there's just no easy answer or way around things.

 

[alceryes]

Some great info in this thread!
I just wanted to add that you will also want to take into account air vortices and air pressure at different locations inside your case. You never want to have fans fighting against each other for the same air. At that point they run very sub-optimally and just create a negative air pressure vortex (usually over your CPU). This is the most common area to have this issue due to exhaust fans or rad fans exhausting out the back along with top mount fans exhausting out the top.
The solution will vary greatly depending on your config, fans, case, what you want to achieve, etc., but I have found top mount fans to be unnecessary if you have both an open airflow at the top of your case (which you've already got if there's a fan there) AND something exhausting out the back of your case.

 

[Bearmann]

In addition to what has already been stated, in my opinion, more important than the positive pressure effect on dust build-up, is the total airflow through the case. Air contains a certain amount (concentration) of dust. Making up numbers, suppose that you are putting 100 cubic feet of air through the case per minute. Now you increase your airflow to 300 cfm. I would expect about three times as much dust deposited with the additional air flow. IMO, fewer fans at lower RPM's = less dust. Want to double the airflow to get 2 degrees lower temps? Expect twice as much dust. Of course, if you are running two GTX 1080ti's in SLI, you need that high air flow. I'm always amazed at how little dust is in at my Dell work computers which use one ducted fan only (for CPU and case.) My building was built in the 1960's and uses unfiltered air.

 

[Karadjgne]

In simple terms, you don't need gale force windage inside the case, you just need the air to move and remove heat with it.

 

[mrjyyyy]

Thank you for all great infos, learned a lot in this thread. Tried experimenting and it seems hard to achieve the positive preasure I want, Maybe there are so many factors like, my cpu cooler is a stock AMD Wraith spire which is blowing air downward the processor and sucking air facing the side panel, also, The Fans of my gpu on idle is on Fan Stop mode and maybe sucking air from back during gaming/on load.

My main purpose is to minimize the dust (it's hard living on a dry dusty place 😆) And It's almost summer in here so I need to keep my temps low during gaming. On a week or 2 the dust buildup is noticeable on my rig.

 

[Karadjgne]

I've found that using the most filtered intakes as possible, at a lower rpm draws less dust than fewer fans at higher rpm

 

[Bearmann]

View: https://www.youtube.com/watch?v=dLX54ounENY

 

[Karadjgne]

Hah, that's an FD Define R5 Window, the exact same case as mine, and it's supposedly not a good airflow case at all. Lol. Right!