ducting Mod

Ex22 · Post by **Ex22** » Fri Jun 27, 2003 9:44 am

Anyone know if this will make any diff with a SLK900 and a bit of duct tape to seal it on? Just curious.
http://www.thermaltake.com/products/hea ... ingMod.htm

GamingGod · Post by **GamingGod** » Fri Jun 27, 2003 11:14 am

offtopic but i was always curious as to how a slk900 would perform in suck mode, and ducted to the back.

IceFire · Post by **IceFire** » Fri Jun 27, 2003 11:27 am

Ex22 wrote:Anyone know if this will make any diff with a SLK900 and a bit of duct tape to seal it on? Just curious.
http://www.thermaltake.com/products/hea ... ingMod.htm

You'll need a way to mount it on the SLK900, but it should help a little 1 or 2 C.

As for tring to set it up for sucking, as is, it looks like it'll do worst by a good 3 to 5 C. You might be able to just seal around the fan and one side of the sink and be about the same as blowing on it. I might give that a try on my slk800 this weekend.

Ex22 · Post by **Ex22** » Sat Jun 28, 2003 9:42 am

Let me know how it turns out. They also claim that it will make it more quiet but I don't think that's possible with a Panaflo

IceFire · Post by **IceFire** » Mon Jun 30, 2003 6:16 pm

Well, I'm screw. I have to RMA my motherboard back to NewEgg so I will not be able to do this for at least another week.

Ex22 · Post by **Ex22** » Tue Jul 01, 2003 7:53 am

outch! I wish all companies followed Maxtor RMA's way. They send you the new part and you have 3 weeks to send back the broken HD.

lucienrau · Post by **lucienrau** » Tue Jul 01, 2003 9:18 am

I have your pain IceFire. I just had to RMA my mb as well and I'm without a pc at home till I get it back.

I should have read better though. If you plug the aux power supply (the 4 pin one for intel chips) then it cause an unfixable coil whine when the hd is accessed. argh.

Roger Beckett · Post by **Roger Beckett** » Mon Jul 07, 2003 7:55 am

I doubt very much if the actual airflow would increase appreciably by simply backing it away from the sink a few inches. Yes, the fan would see less resistance, but not appreciably less.

Now, if you could mount the fan on what is usually reserved for the exhaust fan and then run the duct all the way to the sink, you would be blowing ambient room air on the heatsink and pressurizing the case at the same time, a technique that would improve cooling.

I am, however, very curious as to how one of our intrepid cohorts can say with such athority that it would give you 1 or 2 degrees improvment, or that it would be worse by 5 degrees or so in "suck" mode. Sounds too much like numbers drawn from thin air to me.

The variables one would have to know to predict a 1 or 2 degree improvment are numerous. Number one being the increase in total airflow.

Any duct decreases airflow due to friction with the walls of the duct. A curved duct decreases airflow more than a straight one. It is concievable that you would actually decrease total airflow across the heatsink.

Furthermore, heat transfer between a solid, like the heatsink, and air is better with turbulent flow. If this duct decreases turbulence it could actually be less efficient even if the total volume of air were identical.

Let's review, less total airflow, perhaps less turbulence at air to heatsink interface = fewer BTU's removed. However, the duct may bring air from a cooler area. The greater the temperature difference between the air and the HS the greater the efficiency.

Personally, looks like a gimmick designed to remove money from the owner rather than BTU's from the CPU.

IceFire · Post by **IceFire** » Mon Jul 07, 2003 10:25 am

Roger Beckett wrote:I am, however, very curious as to how one of our intrepid cohorts can say with such athority that it would give you 1 or 2 degrees improvment, or that it would be worse by 5 degrees or so in "suck" mode. Sounds too much like numbers drawn from thin air to me.

These numbers are not out of thin air but out of trial and error. These numbers come from trying it w/ a volcano 7 and slk-800. Also in the distance past, I tried it on an alpha. These were roughly the numbers I was getting.

On my Volcano 7, I put a 80 onto to suck air verse blowing on it. The temp went up by 5C at partially to full load. I next used a duct/adaptor (60mm to 80mm) and turn it back to blowing on it. The temp went down 2 degree C from just having the 80 mm on.

On the slk-800, I did the same test except no 60 to 80 duct/adaptor. The results was similar.

Because of the design of the Alpha, sucking air was better.

Remember, this is w/ my setup. I am also using a range (3 to 5C) and not stating that it will be worst by 5C.

Roger Beckett · Post by **Roger Beckett** » Mon Jul 07, 2003 1:33 pm

I'm sorry, but there are just too many variables involved to assign an effect of 1 or 2 degrees to a specific change. You could see a 5 degree change in temp from one test to another without changing anything.

For that matter, if 5 degrees is important to your cooling setup you need a complete redesign. At any reasonable operating temperature 5 degrees is virtually irrelavent. If you take a look at the algorithms that graph failure rate to temperature you will see that the curve is pretty flat until you approach operating max at which point it goes virtually exponential.

If you are operating within 5 or even 10 degrees of max you're going to crash and burn sooner rather than later anyway.

I say all this not because I wish to challenge your methodology, or intent, but to remind you that there is a big picture involved.

Bluefront · Post by **Bluefront** » Mon Jul 07, 2003 1:57 pm

Well I have to mostly agree with Roger Beckett......this is one of the mods you'll just have to try out to know for sure. It's like putting a heatsink on a cpu. Do it several times in a row, and you'll get different heat transfer results every time.

For me, it's better to blow ambient air right on a CPU through a duct, Alpha's suggestions not-withstanding. The cpu always runs cooler. The other temps in the case might go the other way. It depends on the other variables, airflow, fan performance, etc.

The thing to keep in mind is....ducts work. How you implement the setup, is up to you. Do it right and you'll benefit.

Td_nw · Post by **Td_nw** » Mon Jul 07, 2003 5:22 pm

OK

Totally agree with the ambient air from outside the case being brought to the CPU, but I have a question...

I want to use an Alpha (suck mode) and that Thermaltake duct mod as an exhaust out to a blowhole to take any possible load off the PSU fan, but how do I do a duct to get the outside air to the base of the Alpha?

Is it good enough to get an open duct from the rear fan (two intake 7v fans, one front - one rear) to about an inch from the Alpha?

Or is this just a not a good idea for some reason that I am missing?

Bluefront · Post by **Bluefront** » Tue Jul 08, 2003 2:50 am

How to get ambient air to the base of an Alpha. then to be blown outside by a duct? Good question, but probably no easy answer, or solution.

Tell me again what is your objection to blowing ambient air right onto the alpha using a duct setup easily made. Alpha's suck/blow suggestions were never made with ducting in mind. This setup works perfectly for me.

IceFire · Post by **IceFire** » Tue Jul 08, 2003 5:08 am

Roger Beckett wrote:I'm sorry, but there are just too many variables involved to assign an effect of 1 or 2 degrees to a specific change. You could see a 5 degree change in temp from one test to another without changing anything.

For that matter, if 5 degrees is important to your cooling setup you need a complete redesign. At any reasonable operating temperature 5 degrees is virtually irrelavent. If you take a look at the algorithms that graph failure rate to temperature you will see that the curve is pretty flat until you approach operating max at which point it goes virtually exponential.

If you are operating within 5 or even 10 degrees of max you're going to crash and burn sooner rather than later anyway.

I say all this not because I wish to challenge your methodology, or intent, but to remind you that there is a big picture involved.

Roger,

I agree w/ you for the most part. I should have stated evironmental variables, hardward setup, .... However as you know, a home office is not a testing lab. Is 3.13%(1/32C) to 6.25%(2/32C) from one test to another within margin of error? To many variables to know for sure in my case. Even 3 to 4 degrees is questionable. Of coarse, newer processors are running hotter in the 50C range so is around 2% (1C) and 4% (2C).

For quiet computer, I would think 5C COULD (depending on PS, the case, the room temp, ...) be the difference of the PS fan (CPU fan and what ever else that is affect by it) running partial to full speed.

God, I feel like I was in the EE and Chem lab in college again w/ all this if, or, but, etc. So bad memories there!

Anyway, in conclusion. Roger, you are mostly right. But if that is the case, what ever we advise him/her other than try it out is of no use.

So TRY IT OUT YOUSELF and let us know what the result is.

Td_nw · Post by **Td_nw** » Tue Jul 08, 2003 5:17 pm

Bluefront wrote:Tell me again what is your objection to blowing ambient air right onto the alpha using a duct setup easily made.

I might just be taking this to far but...... I have seen most of your duct work. That is actually what made me think up this current project.

My loudest component is going to be the PSU.

I want to take the hot air from the Alpha and take it out the top blowhole. That way the heat from the CPU is being taken out instead of getting into the PSU. With the case over pressure, and no added heat from the CPU, I can get the PSU as quiet as I can.

Also a duct from the CPU all the way to a blowhole will allow me some turns in the duct to decrease the noise from the PSU fan.

Bluefront · Post by **Bluefront** » Wed Jul 09, 2003 1:45 am

Actually, with the duct setup I'm using now, I thought the PSU would suffer heat-wise since it was exhausting most of the heat from the CPU. It hasn't happened. The PSU output temps (I monitor that temp all the time with a DD5) barely increase at 100% CPU usage, and it's fan speed sounds the same.

If you're worried about PSU noise generated by ducting you might consider this.....move the PSU out of the case completely. Construct a small case the computer sits on. Put the PSU in that case, route the wires into the computer through a small hole, and let the PSU cool itself with vent holes underneath using it's own fan for exhaust. I'm sure it would work....just a small project.

Tigr · Post by **Tigr** » Wed Jul 09, 2003 8:24 am

I was very curious with the DuctingMod so I went out and found one. Iam very disappointed. I used it with my CoolerMaster and the temps jumped up by 3 degrees (full load).

Roger Beckett · Post by **Roger Beckett** » Wed Jul 09, 2003 8:44 am

My loudest component is going to be the PSU. ... the heat from the CPU is being taken out instead of getting into the PSU. With the case over pressure, and no added heat from the CPU, I can get the PSU as quiet as I can.

Also a duct from the CPU all the way to a blowhole will allow me some turns in the duct to decrease the noise from the PSU fan.

Why decide what you loudest component is going to be if the goal is to have the quietest overall system? This is a common design mistake, approach with an open mind, it may well be that the PSU will not be your loudest component.

Do you have a logical well thought out reason for not letting the CPU air going through the PSU? The heatsinks and other components in the PSU will do a great deal to baffle sound. Furthermore, air movement, even if that air is warm, will make a huge difference in PSU cooling. In fact, if you could duct air through the CPU heatsink and then straight through the PSU you might only have to move 1/2 as much air.

For that matter, if the CPU air is exhausted straight into the PSU the PSU fan may never kick on, and the CPU fan, not the PSU would be the loudest component

Td_nw · Post by **Td_nw** » Thu Jul 10, 2003 2:42 pm

Yeah, all good points.

I have the parts on the way. Guess I will start with an rear intake duct to the CPU and see what response I get from the PSU. No way to know what really is going to happen until I try it.

What I really want (probably in my next system) is a heatpipe that works good in a horizontal orintation. (The original Heatlane Zen was kinda a stinker in most reviews I read, but maybe in the future something similar will get better.) That way a duct could run intake from outside the case and then across the cooler and exhaust duct to outside the case. The air would never even mix with the case air.

Bah, I will probably just go to water cooling.

Tigr · Post by **Tigr** » Sun Jul 13, 2003 1:52 am

I am not working with a SLK but I think it may be interesting still. I use the CoolerMaster XDream with a Papst NGL on top of it. Here are the results of using DuctingMod. The temps are given after 15 minutes of CPUburn. The temps are in degrees Celsius

Day 1:
Fan blows onto the heatsink, without DuctingMod:
CPU 63 mobo 33
Fan blows onto the heatsink, with DuctingMod:
CPU 64 mobo 33
Fan sucks the air out of heatsink, without DuctingMod:
CPU 63 mobo 32

Day 2: (ambient changed!)
Fan sucks the air out of heatsink, without DuctingMod:
CPU 58 mobo 29
Fan sucks the air out of heatsink, with DuctingMod:
CPU 61 mobo 30

So, at its best, DuctingMod is useless in my setup. It interferes slightly with the airflow, more when the air sucked out than when it is blown on.

I have a feeling that DuctingMod is really destined for those fans that pump a lot of air, like more than 50 cfm or something like that. For the amount of air Papst NGL pumps thorugh it makes no difference whatsoever. I think, thus, it is more for overclockers than for us

ducting Mod

ducting Mod

Re: ducting Mod

Airfow

Re: Airfow

Testing is good...but...

Re: Testing is good...but...

Couple of questions