Funky Idea for Extra Cooling

PSUs: The source of DC power for all components in the PC & often a big noise source.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee, Devonavar

Post Reply
please
Posts: 22
Joined: Sun May 29, 2005 8:14 am

Funky Idea for Extra Cooling

Post by please » Fri Oct 14, 2005 10:30 am

Hey there everyone. I am a long-time lurker and recently started posting*.

I didn't see a 'crazy ideas' forum, so I posted here since it is mainly a PSU mod that this involves. I am probably not the first one with the idea, but I came up with this when thinking of how to cool the new Antec P150 as silently as possible. I am originally thinking of the NeoHE430 that is used with the single fan at the rear and vented front panel, but the concept may be applied in other cases(pun intended). I was contemplating the PSU fan speed ramping up due to excessive heat buildup in the case due to a hot, higher performance CPU/GPU. I was also thinking how the CPU HSF was right under the PSU bottom panel and if someone was using a passive Ninja, then some(lots?) of the heat would radiate upwards and heat up the PSU, also assisting in raising the fan speed and noise level.
I was thing of turning that negative attribute into a positive thing and embrace the passive heat exchange. Here's the idea...what if you made the bottom panel of the PSU from thick aluminum, possibly with fins projecting into the PSU interior(but obviously not thru the PCB). Then(and this is important) duct the PSU to receive air from the outside, thereby balancing the increased heat with cooler 'cooling air'. You would have to balance the passive heatsink size with the available cooling air, so as not to induce higher fan speeds, since the objective is to increase cooling without increasing fan speeds.

The benefit would be minimal, but maybe just enough to run a Ninja passively on a hot CPU, with 80% of the heat exiting thru the rear 120MM fan and 20% thru the passive PSU HS. I am guessing on the percentages.

A variation might be to do the bottom AND sides in aluminum to draw the heat up around the PSU PCB and have the fins on the side walls running parallel to the existing HSF(lengthwise)

Another variation (and maybe easier to attempt as a DIY, but not as effective IMHO) would be to construct/replace the current rear-bottom-front panels (which are constructed from one piece of steel) with aluminum. Using this model, the front grill gets cooled by incoming air and the rear panel enjoys some passive cooling as well as some residual airflow from the fan.

This is just a concept, so point out the good parts before you 'share' your constructive criticisms...ie flame me for the ignorant newbie spcr wannabe that I (happily)am :D .

* It seems that whenever someone starts their post with 'logtime lurker, new poster' it seems they are about to post something that is going to be inflammatory, and is maybe asking for leniency upfront.

(in a feeble attempt to deflect flaming, would this be a good time to point out that I have never been properly welcomed to spcr as is tradition?)

StarfishChris
Posts: 968
Joined: Fri Jan 07, 2005 7:13 pm
Location: Bristol, UK
Contact:

Post by StarfishChris » Fri Oct 14, 2005 4:37 pm

I guess you need a Welcome to SPCR!
I haven't seen one for ages! Time to start a new trend methinks?

TBH it sounds like that would be better for cooling the PSU rather than anything else. The PCB usually sits on the lower panel and I don't think heat affects it much with a tower heatsink and rear fan. Perhaps it does with a traditional HSF or Zalman flower, but I don't know - the only thermally controlled PSU I have is the original Sonata PSU and that heats up of its own accord!
Last edited by StarfishChris on Fri Oct 14, 2005 6:26 pm, edited 1 time in total.

stromgald
Posts: 887
Joined: Wed Jun 09, 2004 12:45 pm
Location: California, US

Post by stromgald » Fri Oct 14, 2005 5:42 pm

That certainly is an intersting idea, transferring the heat to the PSU and bringing in oustide air to combat that. That way you only have one area to cool.

I have two criticisms, the first is just a correction, and the second is more of a problem I see in your design. First, it'd be more efficient if you had a thin piece of aluminum on the bottom of the PSU, since heat will transfer more easily that way (less material to go through/hold heat). You would still want fins on the inside to radiate the the heat into the PSU, its always good to have more surface area where you want to radiate the heat.

And, second somehow you have to get the heat into the PSU when it probably doesn't want to. Heat generally travels to where it is coolest. It is basically energy transfer to wherever there is less energy. If the air next to the PSU is hotter than the metal of the bottom of the PSU, heat will go into the metal, and the reverse happens when the bottom of the PSU is hotter than the air around it (which is usually the situation).

The idea isn't so crazy, but it might be hard to implement. Collecting heat into one place so fewer fans are involved isn't anything really new. In fact the the Zalman TNN (totally no noise) does something like it, but they're crazy enough to move all the heat to the case walls/material. :P

please
Posts: 22
Joined: Sun May 29, 2005 8:14 am

Post by please » Mon Oct 17, 2005 8:15 pm

@ StarfishChris
Thanx for the welcome...I feel like I belong now :D . Thanks for the feedback as well. I would think there must be some heat that rises up to the bottom of the PSU, but it is certainly debateable as to how much, and certainly affected by the particular set-up that exists. I was thinking of the P150 and using a Ninja fanless, so that would be a scenario that would create a decent amount of heat up to the PSU bottom, even with the rear 120mm fan drawing air thru the Ninja.

@ stromgald

Thanks for the feedback. You make some valid points and observations. Once again I think the air enough between the bottom of the PSU and the Ninja would be warmer than the interior air of the fresh-air-ducted PSU to avoid any 'flowback' of heat into the P150 interior. I also think that the interior air of the PSU would be cooler due to the ducting in of room temp air. That would facilitate the flow of heat from the inner aluminum fins to the cooler air.

Constructing the piece could prove to be quite difficult. Your suggestion of a thinner sheet is helpful from the construction standpoint. I think it might be quite a project to not only build, but to test the differences. I think I would need to have 3 PSU's. Use one as the control, one with a front-bottom-back one-piece aluminum mod, and one with a side-bottom-side configured piece. I was thinking of a way to test without having to fully tear one apart, like an add-on piece. Maybe adding slots to the length of PSU on the bottom near the bottom/side crease, to be able to slide a U-shaped piece into the PSU

I guess I am getting a bit carried away as to the hassle versus the performance gain. I am going to put this one to bed...maybe someone else will be inspired to run with it.

Post Reply