Most Reliable Fanless PSU?

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

jermaink
Posts: 265
Joined: Sat Mar 05, 2005 1:20 am
Location: Melbourne, Australia

Post by jermaink » Fri Jul 08, 2005 2:42 am

Maybe you should have gotten a Phantom 500.

I'm getting one next week, and it's rare because it's one of the few products where we aren't being ripped off in prices in Australia compared to the US. :)

perplex
Posts: 298
Joined: Wed Apr 13, 2005 1:29 pm
Location: Club Tropicana

Post by perplex » Fri Jul 08, 2005 3:23 am

if it's a new model i think you should be ok? :?

jermaink
Posts: 265
Joined: Sat Mar 05, 2005 1:20 am
Location: Melbourne, Australia

Post by jermaink » Fri Jul 08, 2005 8:46 am

I'm pretty certain it will be. I live in Australia, and the Phantom 500's only comming out next week. I also think that the Phantom 500's are all new.

I have heard of one of those 'not so reliable' Phantom 350's floating around in Australia though.

geminite
Posts: 12
Joined: Mon Apr 11, 2005 4:48 pm
Location: Toronto, Canada

Post by geminite » Wed Aug 31, 2005 6:51 pm

My SilverStone SST-ST30NF 300W Fanless Power Supply PSU just died.

Only took 2 months to fail on me.

So in total, I have gone through 2 Antec Phantoms and 1 Silverstone within 5 months.

I have a Sonata 2 case with a great heatsink and 2x 120mm fans, so I don't think cooling is the problem.

My personal opinion is fanless PSUs are just not ready for the market. This will be the last fanless PSU i will purchase.

jermaink
Posts: 265
Joined: Sat Mar 05, 2005 1:20 am
Location: Melbourne, Australia

Post by jermaink » Thu Sep 01, 2005 8:22 pm

Has anyone had a problem with a Phantom 500W? I hope no one has, because it would confirm that the 500W version uses the new Phantoms, and that they are stable.

vertigo
Posts: 647
Joined: Sun Apr 17, 2005 6:09 am
Location: UK

Post by vertigo » Fri Sep 02, 2005 12:05 am

I have a Sonata 2 case with a great heatsink and 2x 120mm fans, so I don't think cooling is the problem.
Cooling is the problem. If you have one fan blowing in and one blowing out, they match each other in pressure. There will be no pressure blowing air through the power supply.

Fanless power supplies are not meant to run with no air moving through them. They just allow you to move the fan elsewhere. In ATX cases you have a problem because you would want the case fan below the PSU to blow air in the same direction as the air moves through the PSU. If the PSU is an exhaust, you want the case fan to be an exhaust.

This is why I don't think fanless PSU's are useful in ATX cases (it's difficult to get good airflow). The P180 is the perfect case for a fanless PSU.

Lubb
Posts: 52
Joined: Wed Jan 19, 2005 11:11 am
Location: IL-USA

Post by Lubb » Sun Sep 04, 2005 11:34 am

....Fanless power supplies are not meant to run with no air moving through them. They just allow you to move the fan elsewhere. In ATX cases you have a problem because you would want the case fan below the PSU to blow air in the same direction as the air moves through the PSU. If the PSU is an exhaust, you want the case fan to be an exhaust.
-This question of airflow through a fanless PSU seems pretty obvious to me--that you MUST have a fan elsewhere forcing air through the PS. Many people here (and including the SPCR tester) seem to assume that a "fanless" PSU is meant to run with convection cooling only, when I (for one) never assumed that to be the purpose.
This is why I don't think fanless PSU's are useful in ATX cases (it's difficult to get good airflow). The P180 is the perfect case for a fanless PSU.
-Well I already got a P180, and planned on using the lower fan anyway... but you could make a fanless PSU work in a regular case; you would just use front-case-fans only, and then partially-block off the rear vents to make certain that a good portion of the exiting air was forced out through the PSU.
~

mongobilly
Posts: 58
Joined: Tue Mar 22, 2005 11:54 pm

Post by mongobilly » Sun Sep 04, 2005 10:51 pm

I have an Etasis (which is said to be mostly the same as the Silverstone but comes with a heatsink at the back).

It's been working for 3 months without any trouble. With room temperature of 25° C it gets slightly warm, with 20°C barely. Never showed me the red temp LED (reaching 55° it should do that IIRC).

But a somewhat sane concept to move the CPU heat out definitely helps it. Running the case fan at 400RPM makes it noticeably hotter.

Devonavar
SPCR Reviewer
Posts: 1850
Joined: Sun Sep 21, 2003 11:23 am
Location: Vancouver, BC, Canada

Post by Devonavar » Mon Sep 05, 2005 12:40 pm

Lubb wrote:-This question of airflow through a fanless PSU seems pretty obvious to me--that you MUST have a fan elsewhere forcing air through the PS. Many people here (and including the SPCR tester) seem to assume that a "fanless" PSU is meant to run with convection cooling only, when I (for one) never assumed that to be the purpose.
This is incorrect, the SPCR test rig takes intake air through the PSU tester and exausts is through / across the PSU. It's true that there is no fan blowing directly on the PSU, but there is enough airflow through the test rig that it doesn't rely on convection for cooling.

Lubb
Posts: 52
Joined: Wed Jan 19, 2005 11:11 am
Location: IL-USA

Post by Lubb » Mon Sep 05, 2005 4:12 pm

...This is incorrect, the SPCR test rig takes intake air through the PSU tester and exausts is through / across the PSU....
-It would seem to me that, when testing a fanless PSU in a standard ATX case (with the PSU high in the rear) you would need to use front-intake fans (or the load tester) and then block off most of the rear vents, including the empty fan vents, and starting with the lower vents--to ensure that a good portion of the exiting air went through the fanless PSU.

I realize that I am new to the fanless-PSU end of thing here but it amazes me that so many manufacturers thought that fanless PSU's worked well enough to sell, and yet so many people can't keep them running. If you took out a fanned-PS from a standard desktop case and just dropped in a fanless, then it is very likely that the fanless is not getting enough air flow.
~

chylld
Posts: 1413
Joined: Fri Oct 10, 2003 4:45 pm
Location: Sydney, Australia
Contact:

Post by chylld » Mon Sep 05, 2005 4:20 pm

The problem with blocking off all of the rear exhaust holes and using positive pressure to force air through the psu is that not all psu's have exhaust holes on the back, so there isn't a path for the air to flow. for example the silverstone st-30nf has a closed back. but your general idea is sound, directing case air through or over the psu will do a great deal in keeping it cool.

i think the whole fanless psu instability thing is just a teething issue. antec is a big company with loads of experience and high quality standards, and even they had trouble getting it right. same goes with other companies that have tried fanless designs, and even some watercooled designs.

they'll get it right 'enough' one day, for now there are a few standout models that have better reliability than the others, namely the late-model phantom 350, phantom 500 and the silverstone st-30nf. at least i'm hoping that the reliability is high enough given that i'm typing this post using a st-30nf-powered computer :) it would be ironic if my computer crashed before this message made it to thread...

Devonavar
SPCR Reviewer
Posts: 1850
Joined: Sun Sep 21, 2003 11:23 am
Location: Vancouver, BC, Canada

Post by Devonavar » Mon Sep 05, 2005 4:27 pm

Lubb wrote:
...This is incorrect, the SPCR test rig takes intake air through the PSU tester and exausts is through / across the PSU....
-It would seem to me that, when testing a fanless PSU in a standard ATX case (with the PSU high in the rear) you would need to use front-intake fans (or the load tester) and then block off most of the rear vents, including the empty fan vents, and starting with the lower vents--to ensure that a good portion of the exiting air went through the fanless PSU.

I realize that I am new to the fanless-PSU end of thing here but it amazes me that so many manufacturers thought that fanless PSU's worked well enough to sell, and yet so many people can't keep them running. If you took out a fanned-PS from a standard desktop case and just dropped in a fanless, then it is very likely that the fanless is not getting enough air flow.
~
Our tester has 4 80mm intake fans and 1 slow moving 80mm exhaust fan. It is a positive pressure case, so there is airflow though the PSU.

Fanless PSUs do poorly with NO airflow, but they don't need a lot. "Some" is enough, and "some" is what is in our tester.

In my opinion, sealing off all exhausts except the PSU is a worse idea than leaving the case fan. Such a strategy would guarantee that the CPU exhaust would need to be routed through the PSU, heating it up more than it would otherwise.

wim
Posts: 777
Joined: Wed Apr 28, 2004 5:16 am
Location: canberra, australia

Post by wim » Mon Sep 05, 2005 8:03 pm

Lubb wrote:-This question of airflow through a fanless PSU seems pretty obvious to me--that you MUST have a fan elsewhere forcing air through the PS. Many people here (and including the SPCR tester) seem to assume that a "fanless" PSU is meant to run with convection cooling only, when I (for one) never assumed that to be the purpose.
my psu (st30nf) is running ok for some months now, with convection cooling only. but i didn't just assume that was ok, i asked the SST rep here. (s)he said that it is ok, the unit is able to cool itself, provided the led on the back has not turned red (which it hasn't..)
so here is at least one instance of a manufacturer saying, provisionally, that you don't need fans in your system. if someones st30nf fails within the warranty period without the temp warning showing up first, then i presume they would be entitled to a replacement or repair

Lubb
Posts: 52
Joined: Wed Jan 19, 2005 11:11 am
Location: IL-USA

Post by Lubb » Mon Sep 05, 2005 9:33 pm

...Fanless PSUs do poorly with NO airflow, but they don't need a lot. "Some" is enough, and "some" is what is in our tester....
-Perhaps it would be useful to contact all the "label manufacturers", point out the higher failure rate of fanless PSU's and ask for diagrams of what testing setups they used. And I would at the same time also suggest they embed a temperature probe in the PSU that connects to external monitoring panels. As it is there's no way to tell if a fanless PSU is staying cool enough or not, other than it is still working. The one with the rear-LED indicator is something, but a temp probe and a manufacturer's temperature limit is really what is needed here.
....In my opinion, sealing off all exhausts except the PSU is a worse idea than leaving the case fan. Such a strategy would guarantee that the CPU exhaust would need to be routed through the PSU, heating it up more than it would otherwise....
The PSU in a normal ATX system sucks out all the hottest air anyway, doesn't it? --being in the top of the case. The only difference is that it has a fan that ensures a certain amount of airflow passes through the PSU itself. On a fanless setup, the only way you can force more air through the PS is to use front-mounted fans on the case, and block off some of the rear lower vents.
~

mongobilly
Posts: 58
Joined: Tue Mar 22, 2005 11:54 pm

Post by mongobilly » Tue Sep 06, 2005 12:42 am

The Etasis has (AFAIK similar to the Silverstone) holes at the bottom and the backside (the on in the case). I was wondering if it would make sense to somehow duct it to the case exhaust (single fan in my machine) but as long as the unit gets barely warm, I don't think I'll bother. I also never seen the red led on, either.

Hezz
Posts: 18
Joined: Sun Sep 11, 2005 5:39 pm

Post by Hezz » Sun Sep 11, 2005 6:14 pm

Guys,

I'm new here and am in the process of building a custom PC case that uses water and natural convention cooling. It is my second attempt to build a quiet PC case but this one is a more ambitious project. THe first case was eventually a good success using a Zalman quiet power supply.

I would like to use a fanless design if possible. I believe the problems with fanless designs are more due to ineffective PC case design. The fact is that normal convective air moves vertically for the most part and that means that the top and bottom of the case need losts of openings for natural convection to work.

In nearly all PC cases this is not how things are done but the way the case is designed it actually traps hot air inside the top of the case and thus requires fans to exhuast the system. I'm hoping to be succesful by making a special pedestal for the case that allowes nearly the whole bottom of the case to be open except for a mesh filter.

THe top will have three levels of horizontal slits around the top and a top plate to keep anything from falling into the case. The natural convention should be orders of magnitude better than any commercial case I have seen. And I'm hoping that it will be sufficient to eliminate all fans.

Hezz

BrianE
Posts: 667
Joined: Tue Mar 29, 2005 7:39 pm
Location: Vancouver, BC, Canada

Post by BrianE » Mon Sep 12, 2005 1:18 am

Sounds interesting for sure Hezz, good luck with your project and be sure to take some pictures if you can. Keep us updated! :)

cotdt
Posts: 295
Joined: Tue May 10, 2005 3:30 pm

Post by cotdt » Mon Sep 12, 2005 2:02 am

Hezz wrote:Guys,

I'm new here and am in the process of building a custom PC case that uses water and natural convention cooling. It is my second attempt to build a quiet PC case but this one is a more ambitious project. THe first case was eventually a good success using a Zalman quiet power supply.

I would like to use a fanless design if possible. I believe the problems with fanless designs are more due to ineffective PC case design. The fact is that normal convective air moves vertically for the most part and that means that the top and bottom of the case need losts of openings for natural convection to work.

In nearly all PC cases this is not how things are done but the way the case is designed it actually traps hot air inside the top of the case and thus requires fans to exhuast the system. I'm hoping to be succesful by making a special pedestal for the case that allowes nearly the whole bottom of the case to be open except for a mesh filter.

THe top will have three levels of horizontal slits around the top and a top plate to keep anything from falling into the case. The natural convention should be orders of magnitude better than any commercial case I have seen. And I'm hoping that it will be sufficient to eliminate all fans.

Hezz
I definately agree with all your points. But if you're using watercooling...

1. The case runs so cool that a fanless PSU won't have any heating problems. I have an Antec Phantom 350 (new model) and never had any problems despite there being no fans in my computer, which is passive watercooled.

2. If you want, you can watercool the fanless PSU! Everything needs to be cooled including the coils, but it will be really easy to just add a waterblock to a premade fanless PSU like a Phantom 350 or Silverstone. Won't work as well for the Fortron Zen though.

3. w/o watercooling, modding the case for better passive ventilation or simply mounting the PSU externally works too.

Hezz
Posts: 18
Joined: Sun Sep 11, 2005 5:39 pm

Post by Hezz » Tue Sep 13, 2005 8:32 pm

Guys,

Thanks for the feedback. I will post some pictures when I get to that point but since I don't have a digital camera it is kind of a pain to get the pictures developed on CD.

I think my design will be overkill with the water cooling but I am trying to make the case as quiet as possible as it will be a HTPC. I also have a tendency to over built so that there is high potential for using high end or performance parts and moderate over clocking.

I can maybe post some pictures of the solid models I have worked on but I am not sure since there is an outside possiblity that I may start up a small business on the side manufacturing the cases in small quantities and sell via the internet. There is no real new technology just a well executed design approach. And the cases would be expensive specialty items.

Hezz

Post Reply