Page 1 of 3 123 LastLast
Results 1 to 25 of 61

Thread: InTheFlow's Rig ~ Look at the purdy pictures...

  1. #1
    Xtreme Member
    Join Date
    Apr 2006
    Location
    Out West in the Rockies
    Posts
    270

    InTheFlow's Rig ~ Update 3-28-07

















    My loop is staying pretty cool so far. I just ran tests at each pump speed setting to see what the difference in temps would be. The difference between speed 1 and speed 5 was only 4C!

    My two 7900s at full load went from the 70's to running at only 42C and the overclocked opti went from a dual prime95 high of 57C on air down to a high of 47C at the speed setting of one on the pump! However, running the pump a bit faster knocks off a couple degrees from the temp so it keeps it at around 45C mostly. Idle temps on the chip are about the same as with air but the idle on the graphics cards dropped down to 38C!

    Tomorrow I'm going to run similar tests with different fan voltages to find the sweet spot between cooling and silence.

    I'm really thankful to everyone here and on other forums who have helped me learn this stuff...its making my first water cooling build quite fun!

    ************************************************** *
    Update 7-28-06

    As mentioned in one of the posts, I decided to pull the waterblock and proc to take off the metal spacers thinking it possibly needed to be tighter to get better temps.

    But check this out:



    I found out that even though my protector epoxy balls 'looked' flat, they were not. Not even close. The next pic shows where the proc was making good contact (right side) and the side that wasn't.



    Well, I got out a razorblade and shaved the baby's down. Then I found out that the core itself has a slight upward bow. That could have happened when my mainboard was bowed around the proc, not sure. Anyways, now I have re-attached everything and have seen only about 1* better temps. Maybe the AS5 will bring it down another couple of degrees.

    I also think my temps will go down when I get rid of the Fluid XP and replace it with deionized water and Pentosin.

    I'm also working on installing my UV lights. First step, take the custom made (by Mountain Mods) drive tray and install the wiring. This pic shows some JB Weld drying where the power connector is. (Under the clamp)



    Here is a close-up of the wiring: (The white connectors will go to the inverters.)



    There is a hidden benefit to using shrink tubing on the wires...if you re-heat them, they become flexible and then you can make the wires go almost anywhere you want. When they cool, it retains the shape you made.

    ***********************************************
    Update 7-30-06:

    I'm sort of close to finishing, no seriously! I finished placing the mirrors and lights and figured out how/where to run the wiring.

    Mirrors you ask? I had a local glass shop cut me strips of mirror the size of the UV light tubes. This makes the UV light that is normally absorbed into the black paint reflect back into the viewable area making it brighter.



    That is a pic of the mirror prepping for installation. I got some double sided heavy duty clear tape to mount the mirrors to the case. Also picked up a package of that dual locking Velcro type stuff to attach the lights to the mirrors.



    That shows you how it all will connect together.



    Those are the mirrors installed on the top panel ready for lights and the next pic is of two mirrors installed on the side.





    Here the lights are prepped and ready for placement in front of the mirrors and the next pic is a close-up of the dual locking Velcro stuff on the lights.





    This is the beginning of a power splitter I'm making that will go on the bottom of the drive cage. It will be for plugging in the pump as well as any future extras needed.



    And this is where I left off. I have a lot of sleeving and cable routing to do in addition to the power splitter project!

    The cool thing is that I have temporary cables hooked up for the fans and such which means I can still listen to the tunes that get me into my rhythm while doing the work! hehehehe

    As always, comments welcome.

    *********************************

    UPDATE 8-5-06

    This picture shows several fan connectors ready for painting.

    TIP: If you happen to buy a black powder coated case from Mountain Mods and need to paint additional items to match...use Krylon #1601 Glossy Black. It is very close to the black powder coat!



    Here are a couple of the finished fan extension cables with sleeving. The larger cable is my X-Fi ribbon cable that will go from the sound card to the drive bay.



    I got the power adapter finished that I showed a beginning picture of in the last update.

    The first pic shows how I soldered the different wire lengths together and the one following shows a close up of the joint itself.





    Here are all the wires soldered and ready for sleeving and assembly. The next pic shows the finished adapter. Installing it to the drive bay is a work in progress...

    TIP: If you ever go to do something like this, make sure you test for continuity before finalizing everything. I had accidentally connected one of the adapters in reverse. If I hadn't tested it, I would have found out when I plugged something in and it proceeded to cook my system.





    The next series of pics shows how I finished the cable that will go from the motherboard to the power adapter I made.

    First up shows the side I already finished and how I hope the other end will look when I'm done. (Except that it will be black.)



    To make it so you can't see different colored wires before the sleeving starts, the first step is to add short pieces of heat shrink tubing to the wires...



    Then I measured the amount of black sleeving I'd need and put it on. Then I figured out what size heatshrink I'd need for both the middle of the cable (where the color changes from blue to black.) and the part that will go near the connector. I put all that stuff on and then...



    ...installed the Molex connectors.



    Connecting the pins to the black adapter.



    Then it was time to move the small heatshrink down and shrink it.



    Once the sleeving was in place I heatshrinked the 2nd larger piece of tubing near the adapter. After that, zip ties were used to keep the cable orderly.



    So far so good...



    This is a close up of the place where the two sleeving colors meet. It's hard to see but there is a zip tie holding them together. After that was in place I heatshrinked the largest piece of tubing over it for a nice transition.



    Here is the finished cable.



    Then I ran out of sleeving and zip ties! Currently waiting on a shipment from Jab-Tech. (RANT: Jab-Tech is totally customer service oriented. They rock and are highly recommended!)

    Update: 3-28-07:

    I found a couple more pictures that may help out so I'm adding them in now. The job has been complete for a while now.

    This is a close-up of the 'round' Xfi cable. Notice the two different size pieces of heatshrink tubing on the end.



    Tip: Use clothes pins to keep the sleeving back from your working area. Like this:





    I had MANY problems attempting to get the lighting to work the way I had planned. Just to show you what happened...first in the light:



    And now in the dark...notice that only half the bulb is lit:



    I wound up having to re-do the entire wiring job on the switch box. I went into the details over in this thread. (Specifically post #7) So, now the inverters are hidden very close to the CCFLs.

    This is a (pretty bad) picture of when I was using Arctic Silver Thermal Adhesive to attach two small heatsinks to the three hot ICs on the rear of the card. It is hard to see but the heatsinks are being clamped with a removed IHS on top of them so they would stay perfectly flat during the curing process. This was REALLY hard and took about 6 tries to get right.



    Tip: If there is even a remote possibility that you may need to remove them in the future, do NOT use AS Thermal Adhesive by itself!! It is much better to use a mixture of AS 5 with the thermal adhesive instead. That way you can remove them without such a high risk of killing the board. See this thread for more info. I finally got them off the other day but it was a hair raising experience.

    Here is a pic of the heatsinks after they were permanently attached:



    And here is a close-up of how I got the RAM-sinks that were included in the Maze4 Low Profile SLI Kit to work. (I used a razor blade to bend them sideways.)



    Thanks for checking this out! Comments are welcome.
    Last edited by InTheFlow; 03-28-2007 at 09:34 PM. Reason: Update

  2. #2
    Registered User
    Join Date
    May 2006
    Location
    Redondo Beach, California
    Posts
    82
    pics and loop specs please.. ty

  3. #3
    Xtreme Member
    Join Date
    Apr 2005
    Posts
    306
    What's your MB brand, radiator type, gpu block types? So, you've got 3 waterblocks on with the numbers you gave above? I'm really interested in this because it sounds like your performace is pretty damn good considering you have 3 blocks and sounds like your mb's temp is possible one of the few accurate probes out. I've done tests on 3 different mbs and turned out your numbers are in line with the true temps I got when testing the accuracy of each mb's readings (of which all 3 were of course incorrect). I'm going to run 3 blocks soon since I've got sli and would soon like to compare temps with you once I get everything together. Try to get a baseline of your probe readings by down-volting your whole system, esp the cpu as low as it will allow (downclock as low too) and let the system stay unplugged for several hours. Get a good reading with a known accurate probe at the CPU area, assume this is the ambient temp of the liquid and immediately go into bios and check your temp upon boot-up to compare with your known good external temp probe.
    Last edited by scwam; 07-19-2006 at 12:25 AM.
    C300
    Abit IP-35 Pro

  4. #4
    Xtreme Member
    Join Date
    Apr 2006
    Location
    Out West in the Rockies
    Posts
    270
    I'll see about adding some pics later today. But for now the loop specs are:

    Swiftech MCP655 Pump>Storm Rev2>DD Maze 4>DD Maze 4>Thermochill PA120.3>Wye Fitting (for fill ups)>Pump

    (Ambient Temps were between 23-24*C throughout the tests yesterday.)

    I know that motherboard's temp probes are very inaccurate (I have the Asus A8N SLI Premium) I decided to test based on before and after watercooling temp readings by the motherboard. My thinking was that if the board was consistant in it's readings (even if they are inaccurate), I'd at least be able to know the differences in temperatures with reasonalble certainty. (If I'm wrong, enlighten me please.) So wether it is running at a real 45*C is impossible for me to know for sure. But I'd say the delta between what it was with air and now with water should be accurate.

    So what I did was figure out what thermometers were accurate in my home. I did this by putting them in the same area. Three of the four matched the temperature of one another. The fourth was 2*C higher than the others so whenever reading from that one, I'd subtract two degrees from it.

    I have no way of getting a temp probe on top of my naked opti in order to even try to veryify if the mobo's temps are semi-correct. That's why I just went with the 'relative' comparison between the board's before and after readings.

  5. #5
    Xtreme Member
    Join Date
    Apr 2006
    Location
    Out West in the Rockies
    Posts
    270
    Pics added and comments welcome.

  6. #6
    Git-R-Done
    Join Date
    May 2006
    Location
    Toronto, Canada
    Posts
    1,305
    You did that etching yourself? Or laser etched? What material is that on - smoked plexi?

    Either way, very nice looking.

  7. #7
    Xtreme Member
    Join Date
    Apr 2006
    Location
    Out West in the Rockies
    Posts
    270
    Quote Originally Posted by Shpoon
    You did that etching yourself? Or laser etched? What material is that on - smoked plexi?

    Either way, very nice looking.
    Thanks!

    Nope, Mountain Mods did the pet etching for me and it was on black acrylic. The StarWars etching was done by HyperKore and I'm not sure what type of acrylic it is.

  8. #8
    Git-R-Done
    Join Date
    May 2006
    Location
    Toronto, Canada
    Posts
    1,305
    Quote Originally Posted by InTheFlow
    Thanks!

    Nope, Mountain Mods did the pet etching for me and it was on black acrylic. The StarWars etching was done by HyperKore and I'm not sure what type of acrylic it is.
    YGPM.

    Very nice looking case though.

  9. #9
    Xtreme Enthusiast
    Join Date
    Feb 2004
    Location
    Thousand Oaks, CA
    Posts
    574
    Nice job on the case. When you had Hypercore do the Star Wars window did you send them a pre-cut piece of acrylic? I want to use them for a custom image I have and it would be awesome if they could send me back a finished piece with the holes drilled ready to go.
    Intel Xeon 3520 | eVGA Classified E759, GTX-285 SSC Tri-SLI | Corsair Dominator GT 2000Mhz 7-8-7-20 | Auzentech X-Fi Prelude | Intel 160GB X25-MG2 x 2, WDC 1TB Green Storage | Watercool HK 3.0, GPU-X² GTX285, MO-RA 2 Pro | Mips Chipset Kühler | Dual Laing 18w DDC+ | Corsair HX1000 | murderMod TJ07 #007 | Dell 3008WFP | Windows 7 Ultimate x64 |

  10. #10
    Hamster Powered
    Join Date
    Apr 2006
    Location
    Virginia, USA [Krunching since 2001]
    Posts
    7,623
    Very nice!
    XSWCG Disclaimer:
    We are not responsible for the large sums of money that you WILL want to spend to upgrade and add additional equipment. This is an addiction and the forum takes no responsibility morally or financially for the equipment and therapy cost. Thank you and have a great day.

    Sigmund Freud said... "Failure to CRUNCH is a sign of Sexual Inadequacies".

  11. #11
    Xtreme Member
    Join Date
    Apr 2006
    Location
    Out West in the Rockies
    Posts
    270
    Thanks Sig!

    Quote Originally Posted by ShoNuff
    Nice job on the case. When you had Hypercore do the Star Wars window did you send them a pre-cut piece of acrylic? I want to use them for a custom image I have and it would be awesome if they could send me back a finished piece with the holes drilled ready to go.
    I did not send Hyperkore anything. They made the image on acrylic that they stock and then shipped it to Mountain Mods for me. MM put the holes in the window and attached it to the custom cut panel.

    Hyperkore might put the holes in for you if you ask though.

  12. #12
    Xtreme Addict
    Join Date
    Jul 2005
    Location
    ATX
    Posts
    1,004
    thats pretty nice. i really like the blue/black... it uh... looks nice!

  13. #13
    Xtreme Member
    Join Date
    Apr 2005
    Posts
    306
    Quote Originally Posted by InTheFlow
    I know that motherboard's temp probes are very inaccurate (I have the Asus A8N SLI Premium) I decided to test based on before and after watercooling temp readings by the motherboard. My thinking was that if the board was consistant in it's readings (even if they are inaccurate), I'd at least be able to know the differences in temperatures with reasonalble certainty. (If I'm wrong, enlighten me please.) So wether it is running at a real 45*C is impossible for me to know for sure. But I'd say the delta between what it was with air and now with water should be accurate.

    So what I did was figure out what thermometers were accurate in my home. I did this by putting them in the same area. Three of the four matched the temperature of one another. The fourth was 2*C higher than the others so whenever reading from that one, I'd subtract two degrees from it.

    I have no way of getting a temp probe on top of my naked opti in order to even try to veryify if the mobo's temps are semi-correct. That's why I just went with the 'relative' comparison between the board's before and after readings.
    I feel your method is good. I'm doing it the same way in regards to consistency with the mb readouts. however, I've got an extrememly accurate lab probe that is +/-0.2c accuracy. It was only $30 and it's helped me even further confirm the consistency your speaking with on the mb. If you ever need a very reliable thermistor then check this out. It's $10 and can attach to the side of even a naked core, or go between the pins (not suggested as your alignment of the cpu would not be exactly straight). You can hook this straight to a fan controller with temp readout or to the mb temp input. I've tested on both my Epox 9npa+SLI and superflower fan controller and readings were 1c within each other:
    Last edited by scwam; 07-21-2006 at 03:52 AM.
    C300
    Abit IP-35 Pro

  14. #14
    Xtreme Member
    Join Date
    Apr 2006
    Location
    Out West in the Rockies
    Posts
    270
    Quote Originally Posted by scwam
    If you ever need a very reliable thermistor then check this out. It's $10 and can attach to the side of even a naked core,
    Awesome find...what would you use for attaching it to the side of the core? Artic Silver Adhesive?

  15. #15
    XS_THE_MACHINE
    Join Date
    Jan 2006
    Location
    USA/California/Sacramento
    Posts
    1,619
    I hate that case it looks like sh*t

    ~RS

    PS:im wondering when to order my MM UFO
    Asus P5B-Deluxe 0614
    Conroe E6400 @ 2.56ghz
    Team Xtreem DDR2 2x1GB @ 900
    ATI X1900XTX // MC14 Ramsinks -DEAD- Using spare 7900gt now
    Hitachi 80gb | Hitachi 80gb | Seagate 320gb
    Mountainmods U2-UFO Original Black \\ BenQ DVD-RAM
    Corsair Hx620
    Swiftech Storm r2 \\ Swiftech MCW60 \\ Swiftech MCW30 \\ Thermochill PA120.3 \\ Swiftech MCRES Micro \\ DDC+ Plexi Top

    Heatware | Guide To Current WaterCooling Components | Guide To WaterCooling and Leak Testing

  16. #16
    Xtreme Member
    Join Date
    Apr 2006
    Location
    Out West in the Rockies
    Posts
    270
    Quote Originally Posted by ReD.SkY
    I hate that case it looks like sh*t
    Dude, that was really funny! I'm still chuckling...

  17. #17
    Hamster Powered
    Join Date
    Apr 2006
    Location
    Virginia, USA [Krunching since 2001]
    Posts
    7,623
    Quote Originally Posted by ReD.SkY
    I hate that case it looks like sh*t

    ~RS

    PS:im wondering when to order my MM UFO
    Yea, he has the bug REALLLLLLL BAAAAAAAAAAD!
    XSWCG Disclaimer:
    We are not responsible for the large sums of money that you WILL want to spend to upgrade and add additional equipment. This is an addiction and the forum takes no responsibility morally or financially for the equipment and therapy cost. Thank you and have a great day.

    Sigmund Freud said... "Failure to CRUNCH is a sign of Sexual Inadequacies".

  18. #18
    Xtreme Member
    Join Date
    Apr 2006
    Location
    Out West in the Rockies
    Posts
    270

    Testing Results

    I'm attaching my testing results in the hopes that some of you more experienced folks can tell me what is up.

    I'm confused by the results I got on the two seperate days testing.

    The first day I just tested the pump speed temp differences and ambient was pretty consistant. The second day I kept the speed at 5 and was changing the yate loon's speed using my rehobus. (I measured the volts with a multimeter) The ambient temp was all over the place that day though.

    Did I jack something up or was my testing somehow incorrect? The results just don't seem to corolate properly so comments are welcome.
    Attached Files Attached Files

  19. #19
    Xtreme Member
    Join Date
    Apr 2006
    Location
    Out West in the Rockies
    Posts
    270
    Quote Originally Posted by scwam
    If you ever need a very reliable thermistor then check this out.
    Could you attach that thermistor to this or this temp reader to make it accurate or is the accuracy also based on the display equipment?

  20. #20
    Xtreme Member
    Join Date
    Apr 2005
    Posts
    306
    Well, I can say if that is truely accurate to 1c then it's a really good deal for under $10. But IF, you really want to calibrate, get the thermistor I mentioned. It's so damn accurate. I've tested it in water baths from 20c to 70c, against rebound of hot to cold and it still bouced back within <1c accuracy. If you have that thermistor you can compare it against the one you want to make sure they are right. Those guages are ntc probe thermistors but are the standard model that could go off by a few degrees c when subject to 70c and suddenly to 20-25c. They will ueually record under about 5-7c after being exposed to extreme temps. The problem with those is the tip is so big that you really can't make a complete mount to the core's side like you can with a small thermistor. I simply used sekisui double sided thermal tape from "pctechnician" on ebay to mount it to the core's side. Those are nice indeed, if they are truely accurate. I'd just waste the $10 for fun to see if they are and compare it against the Mcshanes thermistor. Most like it is a 10k thermistor just like mcshane's and operates on the fenwell curve. I got the same results using that type thermistor and mcshane's on both my EPox 9NPA+SLI temp imput spdif connect as I did on my Superflower fan controller/temp input. They both used 10k ohm inputs.
    C300
    Abit IP-35 Pro

  21. #21
    Xtreme Member
    Join Date
    Apr 2006
    Location
    Out West in the Rockies
    Posts
    270
    Quote Originally Posted by KaptCrunch
    yes its how the storm is mounted, turn it 180 an have the outlet on the top while the center is the inlet. also make sure the all the air is out of the rads, with water system sealed.
    flip twist all water hardware to burp out air to fill spot area.
    i see 2 rads?
    if so pump>rad>CPU>Y>GPU's>Y>rad>.

    ithinks u'll get better results doing that order

    Thanks for the reply Kaptn'. However, I am confused as to why changing the mounting of the storm 180 degrees would make a difference since the center is already the inlet? How would moving the outlet higher than the inlet change the temp performance?

    There are zero air bubbles in the loop, I'm very confident of that. It took about two days of moving it around and coaxing the bubbles all out, but there are none now.

    The current flow is from the pump>storm>GPU1>GPU2>RAD>Wye>&Back to Pump. So there is only one RAD.

  22. #22
    Xtreme Member
    Join Date
    Apr 2006
    Location
    Out West in the Rockies
    Posts
    270
    Quote Originally Posted by scwam
    But IF, you really want to calibrate, get the thermistor I mentioned.
    I understand that but what I don't understand is how to use it once I get it. As in what I would connect it to since I don't have a digital temp prob other than the backlit blue one I put in the link earlier.

    It is off by two *C so I figured if I attached the better thermistor to it, it would become accurate. Is that not correct?

    Thanks for the help!

  23. #23
    Xtreme Member
    Join Date
    Apr 2005
    Posts
    306
    Normally connect to an spdif connector and to the mb temp input, if any, or a fan controller w/temp input. The way you want to do it could work but also there may be a chance that the instrument your using uses a 15k or 8k input. In that case you would get wrong readings using the thermistor I suggested. Look up the part number if it's listed on the probe you have, google it and you'll eventually find some chinese sight listing it's specs. It should show it's impedence there or on the instruction manual of the instrument. For the thermistor I suggest you need 10k and I believe it has to be on the fenwell curve, but I wouldn't worry to much about what curve it is.

    I have no idea of the instrument input, I just know that my experience with a $29 fan controller and an epox mb resulted the same temps using the same thermistor. So chances are that accuracy of the instruments is pretty close, unless they go by the different curve readings.
    Last edited by scwam; 07-23-2006 at 01:30 PM.
    C300
    Abit IP-35 Pro

  24. #24
    Xtreme Guru
    Join Date
    Jan 2006
    Location
    Milan, Italy
    Posts
    4,177
    GPU blocks in parallel is a terrible idea. If you think about what happends when you split wateflow accross 2 channels and the reason why we try to get the highest flowrate in loop, youll realize why.

    There is NO PERFORMANCE REASON for turning the storm in any one direction. Convection heat transfer can only happen in still or quasi-still liquids. Ultra-turblulent flow @ 1+gpm through the tiny channels in a storm makes that whole argument risible, I'm afraid.

    There is an advantage if you have a Rev. 1 stomr, in that bubbles tend to get trapped in the top part if outlet below inlet. This does not happen in Rev. 2 Storms.

    There is also 0 measurable performance difference between different loop orders for end-users, because trying to make your loop comply to some idealized loop order will always screw you over.

    The 0.1c you could gain by putting the rad after the pump to shed those 9W of heat before the coolant hits the CPU gets overshadows by the 0.5c you gain from the extra tubing and bends. Not that you can measure anything like that with onboard sensors, and any temp difference less than 2-3c measured with onboard sensors can be safely attributed to ambient variations and whatnot, unless youre measure ambeint with a caled probe and run multiple samples and average the deltas.

    Bottom line: Always use series, always go for the shortest loop with the least sharp bends and no kinks. Everything else is meaningless complication with no real effect on performance.
    Last edited by creidiki; 07-23-2006 at 04:07 PM.
    Quote Originally Posted by Bloody_Sorcerer
    flowrate is for losers!
    Quote Originally Posted by MaxxxRacer
    Thermaltake is kind of like AIDS; it won't go away just by ignoring it.

  25. #25
    Registered User
    Join Date
    Apr 2006
    Location
    Plant City, Florida
    Posts
    20
    Damn dude! VERY VERY nice!
    ASUS P5B-Deluxe
    e6400 @3.1Ghz/Scythe Infinity
    512x4 G.Skill HZ (
    eVGA 7800GT/vf-900cu (515/1230)
    80x2 WD in RAID0 + SeaGate 120 backup
    Thermaltake Xaser III
    Thermaltake NEXUS fan controller
    Sunbeam NUUO 550w
    7 case fans

Page 1 of 3 123 LastLast

Bookmarks

Bookmarks

Posting Permissions

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