Page 4 of 8 FirstFirst 12345678 LastLast
Results 46 to 60 of 110
  1. #46
    Joined
    Feb 2003
    Location
    Boston area
    Posts
    319

    low noise water cooling

    What would make a good high performance water loop that is also low noise?
    I've been going the low noise w/c route for a couple of years now (my PCs, friends PCs).
    For instance, the system I'm typing at now is setup with
    Eheim 1046 (smallest/quietest inline pump they make)
    Innovatek "push on" reservoir (new style - topside inflow - compression fitting replaced with 1/2 inch hose barb)
    Swiftech old style flat plate water block
    Black Ice Pro (not Xtreme) rad
    Panaflo 120mm 'L' fan at 5V
    1/2 ID tubing throughout (except for "choke point" at push-in fittings on water block)

    With a 2600+ 'M' 1.V, 12.5x200MHz, I get temps of 32/42 (idle/"CPU Burn" load) and the pump/fan combo is inaudible. Oh - room temp is 20.5C or so.

    I'm about to upgrade this machine to a Swiftech MCW5002 (due in Monday - and a chevette heater core, waiting for a trip to the auto store). Mostly just because I'd like to see if it makes a difference.

    Why the MCW5002? Have a look at Overclockers.com Waterblock Test Results - note that the 5002 has a quite respectable C/W of .15 and a head drop of only 8.6 in/H20 - so it should mate with a low power pump just fine.
    (It looks like all the better C/W blocks want a "moderate" pump or better - have used a bunch of swiftech wbs over the last couple of years without a single leaker (and all had a mirror finish on the CPU-contact side - no lapping ) so an easy choice for me.

    I'm going to try making this upgrade in stages so I can tell if either component (or both) makes a difference (and how much).

    BTW, for a low noise system (and IMHO, of course) you don't really need push/pull fans on the radiator. I'd mount the radiator so I could pull room temp air into it and make sure there were no air restrictions (cut the PC case open to expose most/all the fins, make sure there are no inlet air restrictions - possibly put some holes into the sides of the case "front bezel").

    Oh - and the Eheim 1048 is the second quietest pump I've heard, so IMHO that's a reasonable "low noise PC" choice.
    Up to now, the 1046s have been quiet, reliable (all the ones I've set up in various PCs (5) have been going 24x7 since I set 'em up with no problems) and they're relatively cheap for what they are (a low of $39 currently - the 12V model seems to have pushed the price down).

    Of course, your idea of a quiet PC is "no more than 30dba max" and my idea of this is more like "no more than 17 or 20 dba" so we're very possibly on different wavelengths here...

  2. #47
    Joined
    Apr 2003
    Location
    USA MI.
    Posts
    346

    Re: [BAR] Silent Computing Thread

    bobkoure,

    Some good points you made, most of all the low resistence water block. A low resistence WB will no doubt help maintain a higher flow rate.

    BUT I'd suggest you might like to take a look at ProCooling where the resident guru on water cooling testing has some numbers on differant WBs with differant pumps used with them. While the flow rate is reduced by the White Water or Cascade the greater cooling, and thus higher O/C potential is worth it, at least to me. My choice of pump will most likly be the MCP600, for much the same reasons (if/when I can afford a second pump). With a Fusion GPU block combined with a high restriction block on the CPU like the Cascade, I'll need the higher head rate of the 600.

    Reason 30dba will seem quiet to me is I'm used to much louder O/C'd set ups. If I can maintain a strong O/C & lower the noise to 30dba I'll be well pleased. 50dba+ is what I'm used to, so 30dba or a bit less will seem like a huge improvement to me.

    As you make no mention of over clocking your systems much, I think you're correct when saying we are after somewhat differant goals. I doubt I'd even notice a 17-20dba system was running, which is why you like them I'm sure.

    I hope you will post your results here from your new cooling loop parts performance. And if I'm wrong, and you do O/C seriously, I'll be even more interested is seeing how high your O/C is.
    Last edited by Blackeagle; 04-17-2004 at 05:29 PM.

  3. #48
    Joined
    Feb 2003
    Location
    Boston area
    Posts
    319

    Re: [BAR] Silent Computing Thread

    Quote Originally Posted by Blackeagle
    bobkoure,
    ...And if I'm wrong, and you do O/C seriously, I'll be even more interested is seeing how high your O/C is.
    Thanks for the pointer to pro cooling - I'll go check that out.
    I don't know what you mean by "seriously" (same way we differ on what "quiet" means).
    I do tend to overclock a bit - the box I'm sitting at is a XP-mobile 2600+ running 12.5x200. Stock is 2GHz, so this is 25% over. My wife's PC is a XP-M 2400+ running 12x200. Stock is 1.8, so this is 33% over (much better "deal", too as the 2400s were $25 or $30 cheaper than the 2600s).
    Both can run "Cpu Burn" for a day without going over ambient + 25C. Neither blue screens in the process - and I can't hear either one of 'em - except late at night when the house is very quiet.
    Bob

  4. #49
    Joined
    Apr 2003
    Location
    USA MI.
    Posts
    346

    Re: [BAR] Silent Computing Thread

    Ha!

    A XP @ 2.5gig is a nice "serious" O/C in anyones book. Most of all when combined with noise levels like you are getting. Reading back through your original post I overlooked where you posted your O/C.

    Clear I need to look into the ultra quiet systems more. 30dba is still, to me, what seems like a pretty nice goal, but just as with O/Cing, more (or in this case less) is always better! And to much is never enough.......as the song goes!

    A really nice set up you have man!

    I'd figured that 2 fans in push=>pull @7v would be quieter than a single @12v while still giving near the same through rad air flow.

    And I'm surprised a BI pro with single Panflo L @ 5v can do that well. Perhaps this is due in part to the fact the BI pro is a thinner rad which allows enough air flow with pretty modest pressure rate from the fan?

    I'll be even more interested now in your results with the Chevette core and the 5002 block in place of what you are making use of now.

    And have you ever tried any of the Noiseblocker fans from germany? These also have become a interest to me due to ratio of air flow for dba levels. So far all I have are 2 of the Noiseblocker S2 80mm fans, but I'd like to try some of thier others as well when the chance offers. Next I want to get some of thier 120mm fans for rad use.

    Well, all for tonight, need........sleep.

  5. #50
    Joined
    Apr 2003
    Location
    sweden
    Posts
    24

    Re: [BAR] Silent Computing Thread

    Quote Originally Posted by Blackeagle
    I'd figured that 2 fans in push=>pull @7v would be quieter than a single @12v while still giving near the same through rad air flow.
    Howdy there,

    Hm, I never understood this thing about push/pull. I have a friend that told me that push/pull can sometimes actually restrict airflow, he read it somewhere, he says. Even my limited logics tells me that if the air throughput is not the same on the both ends you could end up with restricted airflow. Now that's my logic we're talking about. Do you have any links to some actual testing that shows benefits with push/pull.

    /q

  6. #51
    Joined
    Feb 2003
    Location
    Boston area
    Posts
    319

    Re: [BAR] Silent Computing Thread

    Thanks for the kind words!

    The original point of the BI Pro was to have as low air resistance.
    I'm hoping that a chevette core won't have any more resistance (core is a lot thicker, but there's greater frontal surface to counter-balance).
    The panaflo 120 'L' series are pretty good - a lot of the fans I've tried won't even start at 5V.
    I've been thinking about the silenx, but haven't tried 'em. Somebody (on one of the forums I wander through - can't remember who or which forum even though I PMed 'em a "thank you") mentioned that silenx weren't any better than other fans that had been undervolted) - so I've heard that they aren't any better - but haven't actually heard the fans in action.

    BTW, if you are using a fan for "pull" you really need to space the top surface of the fan away from any obstructions. I used some cheap/dead 120mm fans to make spacers (just cut off the struts that hold the fan motor in place).
    It looks like there's a premade item called a mini cool shroud that'd do even better for the BI rads.
    I've ordered the bigger cool shroud - and I'll be testing it for fit on chevette cores at the local auto parts store when it gets here...
    I've got an AA battery pack that outputs 6V - close enough to 5V that maybe I'll bring a fan/battery along with the shroud to check out airflow/noise while I'm looking at fit and build quality.

  7. #52
    Joined
    Feb 2003
    Location
    Boston area
    Posts
    319

    Re: [BAR] Silent Computing Thread

    Quote Originally Posted by quadro
    Hm, I never understood this thing about push/pull. ...
    Your friend is right - you can restrict flow with push/pull. Not as much as with two water pumps, for instance as air is an elastic medium, but, yes, the fan moving slower than the airflow being pulled through it will take energy away, not put it in. There's a reason why pilots stop/feather their props if an engine has failed.
    The reason to have two fans is to increase pressure. You only get increased airflow if you are moving the air through a resistance. IMHO, you are much better off removing/reducing the resistance rather than adding more air pressure - but that's me.

    I don't have any links to offer, but if you are curious about how fans work (and about how wings in general work) google for "coanda effect". I'd also recommend the book Understanding Flight. You only really need to read the first one or two chapters to get all this stuff - the rest (as I remember) is about airframe dynamics and such - interesting stuff to me but possibly deadly-boring to you - and having nothing to do with fans...

  8. #53
    Joined
    Apr 2003
    Location
    USA MI.
    Posts
    346

    Re: [BAR] Silent Computing Thread

    In regards to fans in push pull.

    I agree with what Bob posted above, this is only of value when there is a resistence to air flow to be overcome. And the drag imposed on air flow by a rad is about the best place I can think of to use fans in push/pull, although many people do so to increase case air flow as well.

    I can offer you no links regarding this. I know that Bob Adams touched on this in a article of his I read at overclockers, but I can't find it for you at the moment. So without this link to offer you I only have my own experiance to offer.

    Without a significant resistence to air flow I can't imagine any reason to use fans in push/pull. 2 fans in push/pull in the absense of any resistence have no higher air flow than a single fan of the same type has. And the increased pressure push/pull offers has no purpose without some sort of resistence to the air flow, it's wasted. The second fan can, in this case, only add one thing, NOISE, which we don't want.

    Quadro,

    When a push/pull arrangement is set up it should always use 2 identical fans IMO. As you noted failure to match the air in to the air out will result in a less effective set up than one where the air flow is in balance.

    I'd suggest you try this for yourself on your own rad. Costs are limited to the 2nd fan and materials for a second shroud.

    Bob,

    I've seen others mention use of a fan frame as a standoff. Have you ever done a comparison between a fan frame and a made for the purpose shroud? While it's for sure a easy method, does it not limit flow of air to the corners of the rad, while focusing it in the center round area? I've made my shrouds out of clear plexi glued together in what ever shape an depth I can use. But then I've a cheap source of Lexan so that is just the cheapest and easyest way for me. From what I've seen I think sheet metal is the most populer material for shrouds.

  9. #54
    Joined
    Feb 2003
    Location
    Boston area
    Posts
    319

    Re: [BAR] Silent Computing Thread

    Quote Originally Posted by Blackeagle
    I've seen others mention use of a fan frame as a standoff. Have you ever done a comparison between a fan frame and a made for the purpose shroud? While it's for sure a easy method, does it not limit flow of air to the corners of the rad, while focusing it in the center round area? I've made my shrouds out of clear plexi glued together in what ever shape an depth I can use. But then I've a cheap source of Lexan so that is just the cheapest and easiest way for me. From what I've seen I think sheet metal is the most popular material for shrouds.
    I haven't done the empty frame / shroud comparison. If the radiator involved is a black ice, the difference is probably small - the corners are already covered so that a 120mm fan can mount directly to the radiator.
    I'm sure there's a big difference if the radiator in question is some sort of heater core as they are so much bigger than a 120mm diameter circle. IMHO, it'd be pretty dumb to only use a 120mm circle out of that.

    The attraction for the shroud I'm looking at is that it's set up for screws/rod/whatever to go around the outside of a chevette core (as opposed to straight down from the fan holes and through the radiator fins.
    I plan to use a single fan, so, with this shroud, I can just cut a square hole in my case that's just a little bit smaller than the finned area of the core (maybe even full height of fins, depending on the layout of the header tanks) and fasten the core right up against the case wall.
    Another way to do this would have been to use two shrouds and fasten to a hole for a 120mm fan - probably would be easier to do and the resistance compared to the open square hole would be minimal - but I've already cut out a square hole for the BI so it's easier for me to just cut a bigger hole - lower air resistance is just a bonus.
    Bob

  10. #55
    Joined
    Apr 2003
    Location
    USA MI.
    Posts
    346

    Re: [BAR] Silent Computing Thread

    Well I ordered some parts last night for a new system.

    Antec 3700BQE case without a PSU.
    DFI Infinity ultra 400 motherboard.
    Barton XP3200
    Volcano 12 CPU cooler + Microcool NB cooler

    Will use the same optical, hard, zip & floppy drives + vid card that's in the system I'm typing on. Vid card is due for replacement soon, but not until I get a good look at the next generation of cards in 5-7 weeks (I say that long as I'll watch other peoples results for a bit).

    I've several sets of RAM sticks, will see what they give me. I also have some sound deadening foam on hand. Along with a Super Flower 550 PSU with the new 14cm fan.

    I'm going to give Bob's methods a try, see how close I can get to 20dba. I should at least get below 30dba with what I'm going to try.

    If need be I may buy another pump and some blocks and go with a water rig, have one water rig, but don't want to remove it from my main gaming, movie & music rig. I really do prefer water rigs, but don't want to spend the money on another full loop right now.

  11. #56
    Joined
    Feb 2003
    Location
    Boston area
    Posts
    319

    Re: [BAR] Silent Computing Thread

    If you want quiet - but don't want to go to the expense of water cooling, try the Thermalright AX-7. IMHO, this is one of the best heatsinks for low airflow. It's got a big solid hunk of aluminum under the fan motor's "dead spot", and semi-radial fins coming out from that. The fins are widely spaced, which means you don't need lots of pressure to get air to flow all the way down to the root of each fin. Best thing is that they're cheap - like around US$20, not counting the 80mm fan it'll need.

    On a different note - I've just ordered one of the new Eneim/Innovatek "HPPS" 12VDC pumps. Looks like it'll outperform a 1048 and be quieter than a 1046. I'll report back when I get it in and swap out one of the 1046s I'm using.
    Contra-wise, if you're going to go water, but want quiet/cheap/reliable, I'm now seeing the Eheim 1046 pumps for under US$40 (1046 for $36.90 at reef fanatic).

  12. #57
    Joined
    Apr 2003
    Location
    USA MI.
    Posts
    346

    Re: [BAR] Silent Computing Thread

    That Innovatek pump sounds very interesting. What is it's draw from the PSU?

    Quieter than a 1046 yet performs better than a 1048 will be a winning combo. Plus the ease of just plugging it into the PSU.

    Your review of this pump is one I'll look forward to.

  13. #58
    Joined
    Feb 2003
    Location
    Boston area
    Posts
    319

    Re: [BAR] Silent Computing Thread

    According to Scott at HighSpeed PC, it's approximately 2.5W.
    I think all this info is on the Innovatek page for this pump - but I can't find it.

  14. #59
    Joined
    Apr 2003
    Location
    sweden
    Posts
    24

    Re: [BAR] Silent Computing Thread

    Hey folks,

    Thought this maybe interesting since nobody seems to be talking about it. Take a look at the pdf's below. It's in Swedish and German but I'm sure you'll find your way around. As you can see this particular solution aims to block the noise coming out through the back of the case and you'd be surprised to see how much noise that is. Just put your hand for a moment over the PSU's exhaust hole and you'll know what I'm talking about. I even managed to build a crippled version of this solution, looks a bit ghetto, using plain cardboard and the results are quite pleasing. I'm sure it would make wonders if I'd used proper materials since the cardboard is all but noise insulator. Btw, take a peek at www.whynoi.se. They might have some solutions for silent computing, that's their specialty, that could inspire you guys to make your own silent solutions. And no, I'm not their employee.

    http://www.whynoi.se/support/pdf/wb.pdf
    http://www.whynoi.se/support/pdf/magic_vent_screen.pdf

    Cheers,

    /q
    Last edited by quadro; 04-25-2004 at 10:19 AM.

  15. #60
    Joined
    Feb 2003
    Location
    Boston area
    Posts
    319

    Re: [BAR] Silent Computing Thread

    Try lining your "muffler box" with this stuff.
    ...and you might also want to read this thread
    ... and it's worth browsing Silent PC Review and the forums there.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •