Case airflow vs. Artic Cooling VGA coolers

kojak71 · Post by **kojak71** » Mon Jan 16, 2006 3:30 am

Following on from Mike's article "Quiet PC for Thailand" (http://www.silentpcreview.com/article293-page1.html) it got me thinking about the problems I have with my systems, and it sort of stems from the exhaust system of the graphics card (i.e. Artic Cooling VGA silencer kits). I love their design principle but in their stock form, the fan is too loud. This can be remedied by controlling the fan voltage with a fanmate, but this leads to another problem (on my system at least). Given the amount of negative pressure in the case, air is being sucked into the case via the VGA exhaust vent!

Both my mini-tower systems are housed in Antec BQE cases (plus acoustifoam), and they both have Seasonic S12 PSUs. The case exhaust port is manned by a 120mm Nexus fans. Even though the PSU never ramps up, and even though the Nexus fan is running at lowered voltages, they both cause enough negative pressure within the case to cause this sort of airflow "short-circuit" for want of a better term. I even tried to equalise the pressure differential by using a Nexus 120mm fan at the designated inlet port, but this has also increased the noise the system makes. In order to make the Artic Cooler work better I removed a PCI blanking plate a couple of slot below the graphics card, in effect creating an airflow "short-circuit" for the VGA card.

It's obvious that in free air, the Artic Cooling solution works well, but I'm starting to wonder how comprimised it is given that most PC cases use negative pressure to encourage airflow. No matter what the speed of the Artic Cooling fan it's always going to be fighting against air trying to get into the system via the exhaust vent. Even more horrific is the thought that there you could have a situation where the force of the air expelled is equal to the force of the air trying to get in, resulting in stalled air in the VGA cooler!

I was wondering if anyone else had thoughts on this issue, and if anyone could offer any advice in trying to improve things in this particular case.

datapappan · Post by **datapappan** » Mon Jan 16, 2006 8:08 am

Seems like you need to open up more inlet space. I've seen something similar trying an all exhaust setup - lot's of air was making its way into the case via the I/O backplate (seen with smoke pen). What you have is two large fans winning over the small one.

/datapappan

wundi · Post by **wundi** » Mon Jan 16, 2006 8:34 am

One possible solution would be to build an airflow duct for the VGA Silencer. Ie. isolate the space below the VGA card from the rest of the case and let the Silencer use fresh air from the removed PCI backplates' direction only. Not a very delicate solution perhaps, but it'd stop the bigger fans picking on the smaller one.

kojak71 · Post by **kojak71** » Sun Jan 22, 2006 2:31 pm

Thanks for the advice, I will definitely give the airflow duct a go, particularly in the case which has lots of space. Apart from solving my own particular problems, I justed wanted to raise this point for the SPCR crowd because I think that extra care needs to be taken with such VGA cooling solutions.

lloydo · Post by **lloydo** » Sun Jan 22, 2006 5:21 pm

I think what you have raised is very interesting; I will be building a similar setup soon and I'm curious as to how you could find out where your case airflow goes.

I'm thinking along the lines of taking off one side and replacing it with perspex; but what would you use as smoke? What's a "smoke pen?"

PS this is my first post at SPCR; I love this place and have been lurking for years

lloydo · Post by **lloydo** » Sun Jan 22, 2006 5:23 pm

double post, sorry

kojak71 · Post by **kojak71** » Wed Jan 25, 2006 7:05 am

Well I fashioned an air duct vent, and the temperature readings are the same as before (both the ambient & GPU temps). So it's no worse than what I had before, namely removing a PCI blanking plate. What I need to do is determine what effect this duct has on the rest of the system temperatures.

vertigo · Post by **vertigo** » Wed Jan 25, 2006 9:14 am

The problem is you have two 120mm exhaust fans (PSU and exhaust). Intuitively, the air being removed is the same as one of those fans spinning twice as fast. So we are talking about a Nexus 120mm at 1600-2000rpm, that's a lot of air.

Either leave out a back-panel or allow air to enter via the 5.25" bay. As you saw in the Thailand article, a PSU duct was used; you don't need a duct but you at least need a vent somewhere. I don't think those rotary fans have much pressure.