[Forum Archive '07] More DDC Top & Pump Performance Testing

In March of 2007, I announced a minor revision to my DDCT-01 performance enhancing DDC pump top and with that revision, of course, came more testing. Total forum thread views: 11,926 (as of 4/29/2010). As a side note, all images have been re-sized via HTML, rather than my usual two image (preview & full image via link) approach, to speed along the archiving process. You can still right-click on the images and view them at full-size in a new tab/window. Also note that all vBulletin emoticons and their associated code have been removed from the archive entry.

!--Start of Archive--!

Hey everyone... I'm back with another round of DDC testing for your viewing enjoyment. However, this time, rather than doing somewhat crude comparative testing, I've gathered some more equipment and decided to go for some spiffy pressure vs. flow curves with power consumption. This will, hopefully, provide a clearer picture of overall pump performance for everyone.

Now, before I get started, let me just say that I've already run my data past my Laing contact and they indicated that my data is good (I say this now because I'm sure that a few of you won't be too happy with the data). Additionally, I'll have to wait 'till early next week to post photos of pumps and stuff because I have to take off early today.

So, here we go!

The Contenders:

18W DDC-2TPMP - This is the standard, red impellered DDC-2 that's available in the US. It'll be tested stock, with my DDCT-01 top, my new DDCT-01s top, and as Frankenpump (see below).

18W DDC-3.2VCTP - This is one of the new 18W, blue impellered pumps that's readily available in Europe but hasn't really hit the US market yet. In case you're wondering about the model number, this pump started life as a volume compensating DDC, but it has been modified by Laing to remove the volume compensator. Despite utilizing a v3.3 motor control board, the official name for the pump is the DDC-3.2 (more on this later). The blue impeller does have a smaller inlet (~9mm vs. the red impeller's 12mm inlet) and it is also slightly smaller in diameter (~36mm vs. ~36.75mm). Some other changes include the lengthening of the pump's startup procedure, a newly redesigned motor, RoHS compatibility, and the elimination of the DDC-2's nasty 4A overshoot at startup (which I happen to think may be, at least in part, responsible for the very early death suffered by several DDC-2's...but I still have some research and a lot of testing to do to see if this is the case). This pump will be tested stock and with my new DDCT-01s top.

The Frankenpump - Frankenpump is basically a DDC-2TPMP with a DDCT-01s top and a blue impeller, borrowed from the DDC-3.2VCTP. Please don't jump to try this at home as Frankenpump wasn't a very happy monster...

The DDCT-01s Top - This top is a very small revision to my original DDCT-01 design, which was mainly done to provide compatibility with the blue impellered DDC pumps. I was also able to eek a tiny bit of extra performance outta the old design in this revision. Make no mistake, this is not the DDCT-02 (that's something that's still in development). The DDCT-01s should be available by sometime next week as they're at the shop being made right now. However, if you're interested in making your current DDCT-01 top compatible with the blue impellered pumps, then all you have to do is remove the inlet ring with a Dremel (I'd advise using a flat-tipped tungsten-carbide bit and cleaning any burrs up with an X-Acto knife).

Testing Procedures:

Lacking an appropriate flow meter, I filled a container of known volume (verified very slowly with a graduated cylinder meeting ASTM Class B, E1272 spec.) from 0.5 Gal. to 4.5 Gal. and carefully measured the filling time at various head pressures (each data point collected the mean average of 3 trials, head pressure/flow was controlled with a gate valve just after the pump's outlet, and head pressure was monitored with one of my new pressure gauges [4" gauge face, glycerine filled to eliminate/reduce needle jitter, + or - 1% full scale accuracy, 0.2psig markings (pretty easily approximated to 0.1psig)]). All tests were carried out at 12.00V (measured at the pump's power connector, so it doesn't take into account the minimal resistance presented by the pump's 22ga. power wires) utilizing an HP 6264B variable DC power supply. Oh, yeah, and the charts may look a little cluttered but they're not nearly as bad as the chart with all of the data that I collected (and am still collecting) on it.

The Test Results:

Raw Max. Data:

- 18W DDC-2TPMP (stock) -
Max. Head Pressure: 8.7psig (~20' -- ~6.1m)
Max. Discharge (flow): 2.24 GPM

- 18W DDC-3.2VCTP (stock) -
Max. Head Pressure: 6psig (~13.84' -- ~4.22m)
Max. Discharge (flow): 2.2 GPM

- 18W DDC-2TPMP w/DDCT-01 top -
Max. Head Pressure: 8psig (~18.45' -- ~5.62m)
Max. Discharge (flow): 3.84 GPM

- 18W DDC-2TPMP w/DDCT-01s top -
Max. Head Pressure: 8.2psig (~18.91' -- ~5.76m)
Max. Discharge (flow): 4.11 GPM

- 18W DDC-3.2VCTP w/DDCT-01s top -
Max. Head Pressure: 5.6psig (~12.92' -- ~3.94m)
Max. Discharge (flow): 4.19 GPM

- 18W Frankenpump -
Max. Head Pressure: 8.4psig (~19.37' -- ~5.9m)
Max. Discharge (flow): 3.87 GPM

The Curves w/Power Consumption:

First up, a comparison of the DDCT-01 and DDCT-01s tops (both on the same DDC-2TPMP).


Slightly lower power consumption, slightly higher max. head, and slightly higher max. flow... yup, that's the DDCT-01s in a nutshell.

Moving on, here's a comparison of the DDC-2TPMP and the DDC-3.2VCTP in their stock form and with the DDCT-01s top. This one is a little messy but it's pretty clear that the newer DDC isn't doing so hot, with its substantially lower max. head pressure. Interesting to note the power consumption of the new pump, though--it starts a lot lower and ends higher than that of the DDC-2TPMP. Granted, with very low pressure drop waterblocks on the market now (like the D-Tek FuZion), the DDC-3.2 still stands as a good little pump for our purposes.


Since that's a rather confusing chart, here it is broken into two charts (one for the DDC-2 and one for the DDC-3.2):


...and now to Frankenpump.

Now, the reason I decided to do the Frankenpump test is because I wanted to drive home the point that I had made previously: the majority of the changes and the reason for the performance difference both lie in the new pump's motor, not the impeller. Drop the blue impeller into a DDC-2TPMP and you get higher head pressure and marginally better flow...and a lot of angry noises & vibrations. Interestingly, power consumption also went up considerably.

A lot of you are probably somewhat disappointed with the performance of the new 18W DDC-3.2 and, honestly, so am I. However, we likely won't be seeing them show up in the US through retail channels for about another 6 months or so. In the meantime, however, Laing is scrambling to make a version of the new DDC with performance equivalent to that of the DDC-2. Only time will tell whether they manage to get a production commitment, but, if it does go into production, it'll likely surface as the DDC-3.3.

I have more information, commentary, and photos to add to this but it'll have to wait until next week as I'm going to be out of town this weekend... and, speaking of which, I'm already an hour behind schedule (go figure).

-Later

Update!

Okay, I mentioned in another thread that I was going to post some data for the Iwaki RD-30 at 18.1V and, well, here it is!

Keep in mind that this is just a graph of the raw RD-30 data vs. the refined data for the DDC-2 w/top. The reason that I'm presenting the RD-30 data this way is due to the irregular shape of the curve and how difficult it would have been to create an accurate fitted curve for the data set (the data for the DDC's all came out to be pretty parabolic, so an accurate fitted curve was a snap). Additional tests only worked to confirm the linearity of the performance curve at higher head pressures, so this is how I'm presenting it to you. Also, power consumption in this chart is presented in Watts, instead of Amps, because the pumps were running at different voltages.


As you can see, even at 18.1V, it's substantially more powerful than the DDC-2... but its power consumption is also quite a bit higher than the DDC-2's. At 18.1V, the RD-30's head pressure topped out at 9.1 PSIG (~21 feet -- 6.4m) and the flow was pretty impressive. Honestly, though, comparing that to Iwaki's performance curve for the RD-20 at 24V (found HERE), the RD-20 at 24V may be a better choice in pump than an undervolted RD-30 (especially if you're running the pump above 6 psig). Granted, I don't have any power consumption information for the RD-20 but, if it proves to use less power than the undervolted RD-30... really, it depends on what you want to do with the pump, though.

UPDATE (again): After getting the RD-30 onto my testbed, I've noticed a couple oddities that I'll have to investigate further... the results of which may cause me to redo the above performance curve.

!--End of Archive--!

[Forum Archive '07] Design Project: Liquid Nitrogen/Dry Ice CPU Cooling "Pot"

From time to time, I start various R&D projects for myself with the intent being to keep my problem solving and information acquisition skills sharp. Sometimes these projects evolve into fully realized products set for production (like the DDC pump tops) but they mostly end up getting shelved for one reason or another, like the one shown here. For those not familiar with extreme overclocking and benchmarking, substances like liquid nitrogen and dry ice are frequently used to plunge a CPU's or GPU's operating temperature deep into the sub-zero range to allow for stable operation at very, very high frequencies. A CPU "pot" is a metal device (typically copper and/or aluminum) that is attached to the microprocessor to act as a heatsink, of sorts, that a dry ice/alcohol slurry or liquid nitrogen is poured into. Total forum thread views: 643 (as of 4/29/2010). As a side note, all images have been re-sized via HTML, rather than my usual two image (preview & full image via link) approach, to speed along the archiving process. You can still right-click on the images and view them at full-size in a new tab/window. Also note that all vBulletin emoticons and their associated code have been removed from the archive entry.

!--Start of Archive--!

Hey everyone... I've been thinking of making an LN2/DI pot for a while, so I decided to draw some stuff up in Solidworks this evening and I'd like to know what you guys think.





The C110 portion of the pot would start out as a 96mm x 100mm cylinder that, once machined, would have a 20mm thick base and 5mm thick upper walls. Fitted to the upper walls would be a 5mm thick, 200mm long T6061 extension. The internals would consist of a stepped design with a "star" column in the center and drilled dimples surrounding it.

Now, I'm somewhat unsure about that base thickness... thinking that 20mm may be a little too thick. Granted, it may buy some time if the pot weren't being very closely monitored but it may also cause problems if things get too cold, no? As for attaching the aluminum tube, I was kinda thinking of freezing the base, heating tube and fitting them together with a press or something.

Anyway, thoughts/suggestions/comments would be appreciated.

!--End of Archive--!

[Forum Archive '06] Early DDC Top & Pump Performance Comparison

September 2006 saw the debut of my custom, performance enhancing DDC pump tops. Yes, the testing was, by my current standards, quite crude... but it was the best that I could do with the resources I had at the time. Please keep in mind that "DDC+" actually refers to the old orange-impellered 18W DDC-2. Total forum thread views: 60,346 (as of 4/28/2010). As a side note, all images have been re-sized via HTML, rather than my usual two image (preview & full image via link) approach, to speed along the archiving process. You can still right-click on the images and view them at full-size in a new tab/window. Also note that all vBulletin emoticons and their associated code have been removed from the archive entry.

!--Start of Archive--!

Okay everyone, I finally got my prototypes back from the machine shop and I also managed to finish the pump testing that I was working on. Feel free to repost this test data where ever you wish as I'll only be posting it in XS and [H]. So, yeah...on with the show!

The Contenders:

I've included the AqX 50Z-DC12, the stock (18W) DDC+, and the Laing D5/MCP655 in the tests for comparison purposes and, as a result, I won't really talk about them in this section.

The Alphacool DDC Top (unmodded) - I didn't bother taking photos of this one because everyone is already pretty familiar with it. In short, it's somewhat large, utilizes 7mm inlets (tested with top inlet) and outlet, is tapped for use with 1/4" BSPP (G1/4) fittings, has a rather sharp outlet curve, and it has additional mounting points integrated into it. This design also retains the 'inlet ring' in the top of the impeller housing.

The Alphacool DDC Top (modded) - Again, no photos... but it's basically the same as the stock Alphacool top. Only differences being that the inlet channel has been enlarged from 7mm to approx. 10-10.5mm (I was going for something more representative of the average modder, as opposed to crazed perfectionist) and, as a result of the enlargement, the inlet ring has been removed.

The Radiical DDC Top - A rather crude design overall, this DDC top consists of a block of acrylic (or Delrin) that has had a shallow cylinder machined into the bottom and, essentially, 1/2" holes drilled to intersect the cylinder, acting as inlet and outlet. The result is a large inlet (with no inlet ring) and an outlet that compromises a large portion of the impeller housing's ceiling. The photos illustrate this best...

Radiical top photos (in acrylic):





The Petra'sTech DDCT-01 (a.k.a. Petra's Top/The Petra Top/whatever) DDC Top - Designed by me and produced by the machine shop that's right next-door (in California, no less! Take that, Swiftech < /joke > )... This DDC top is machined from black Delrin and features a 9.5mm top inlet (with inlet ring), 7mm outlet, a gradual outlet curve, inlet and outlet tapped for use with 1/4" BSPP (G1/4) fittings, and a size/shape/profile which is meant to closely match the footprint of the DDC pump (granted, one end had to be made 4mm longer to allow the threaded portion of the outlet to clear one of the mounting screw holes...it doesn't protrude beyond the top portion of the front integrated mounting bracket). BTW, the 'DDCT-01' naming system will be used to denote revision numbers later in time (e.g. DDCT-01, DDCT-02, DDCT-03, etc.).

Photos:



Internals:




Testing:

Testing & Procedures:

A max head pressure test and several forms of flow testing were performed for each pump at both 12.00V and 13.80V. Each flow test consisted of three timed trials wherein a 5 US gallon Fort-Pak (with 0.5 gallons marked as a start point and 4.5 gallons as the end point) was filled--the results of these trials were averaged and the flowrate calculated. The Fort-Pak was tapped and a barbed fitting inserted to allow for easily repeatable filling conditions and the height of the Fort-Pak's inlet was adjusted so that it was approx. 4.5" above the water level in the reservoir that the pumps were drawing from (to negate any gravitational effects such as siphoning). Power for the pumps was controlled with a rather large, older HP model 6264B variable DC power supply and voltages were monitored using a digital multimeter (below).

Max Head Pressure Testing:

The head pressure tests were carried out using the same setup as I used for my previous round of head pressure tests (shown below).

Test Results:

AqX 50z: 12.00V -- 5.2psig (11.99 feet) | 13.80V -- 6.2psig (14.3 feet)

Laing D5: 12.00V -- 6psig (13.84 feet) | 13.80V -- 6psig (13.84 feet)

Laing DDC+: 12.00V -- 8.5psig (19.61 feet) | 13.80V -- 11.1psig (25.6 feet)

DDC+ w/ stock Alphacool: 12.00V -- 8.1psig (18.68 feet) | 13.80V -- 10.1psig (23.3 feet)

DDC+ w/ modded Alphacool: 12.00V -- 7.8psig (17.99 feet) | 13.80V -- 9.4psig (21.68 feet)

DDC+ w/ Radiical Top: 12.00V -- 5.6psig (12.92 feet) | 13.80V -- 6.8psig (15.68 feet)

DDC+ w/ Petra's Top: 12.00V -- 8.1psig (18.68 feet) | 13.80V -- 10.1psig (23.3 feet)

...and a more graphical representation:

Comments: Now, I'm still not quite sure how the D5 managed to sustain a max head pressure that high...but everything else seems to be about right (and I got that same result with multiple D5's). To be honest, I found the pressure loss due to the Radiical top rather surprising as I didn't think that it would do that poorly.

Max Flowrate Testing:

The max flowrate tests were carried out per the testing procedures described above. A 2.5" section of 1/2" ID tubing was used to link the pump's inlet to the reservoir inlet and a 1' 3.25" section of 1/2" ID tubing was used to join the pump's outlet with the Fort-Pak's inlet (below).

Test Results:

AqX 50z: 12.00V -- 2.83 Gal./min. | 13.80V -- 3.24 Gal./min.

Laing D5: 12.00V -- 3.89 Gal./min. | 13.80V -- 4.28 Gal./min.

Laing DDC+: 12.00V -- 2.25 Gal./min. | 13.80V -- 2.49 Gal./min.

DDC+ w/ stock Alphacool: 12.00V -- 3.19 Gal./min. | 13.80V -- 3.48 Gal./min.

DDC+ w/ modded Alphacool: 12.00V -- 3.61 Gal./min. | 13.80V -- 3.94 Gal./min.

DDC+ w/ Radiical Top: 12.00V -- 3.49 Gal./min. | 13.80V -- 3.82 Gal./min.

DDC+ w/ Petra's Top: 12.00V -- 3.5 Gal./min. | 13.80V -- 3.99 Gal./min.

...and a more graphical representation:

Comments: It's interesting, but somewhat perplexing, that increasing the voltage to the D5 yielded an increase in flowrate but no measurable increase in head pressure.

Experimental Cooling Loop 1 (average restriction):

This was the first of two "real world" flowrate tests--mainly done out of curiosity. This particular test loop consisted of the following:
Reservoir --> 2.5" section of 1/2" ID Tygon --> Pump --> 1' section of 1/2" ID Tygon --> Swiftech MCW60 --> 1' 8.25" section of 7/16" ID Masterkleer --> HW Labs Black Ice Pro Radiator --> 1' 3" section of 7/16" ID Masterkleer --> Swiftech Apogee --> 1' 3.25" section of 1/2" ID Tygon --> Fort-Pak
Photo of setup (with Fort-Pak being drained):

Test Results:

AqX 50z: 12.00V -- 1.71 Gal./min. | 13.80V -- 1.96 Gal./min.

Laing D5: 12.00V -- 2.03 Gal./min. | 13.80V -- 2.12 Gal./min.

Laing DDC+: 12.00V -- 1.8 Gal./min. | 13.80V -- 1.99 Gal./min.

DDC+ w/ stock Alphacool: 12.00V -- 2.2 Gal./min. | 13.80V -- 2.44 Gal./min.

DDC+ w/ modded Alphacool: 12.00V -- 2.34 Gal./min. | 13.80V -- 2.58 Gal./min.

DDC+ w/ Radiical Top: 12.00V -- 1.93 Gal./min. | 13.80V -- 2.13 Gal./min.

DDC+ w/ Petra's Top: 12.00V -- 2.32 Gal./min. | 13.80V -- 2.58 Gal./min.

...and a more graphical representation:

Comments: As the max head pressure data suggested would happen, here you can see the Radiical top losing ground to the Laing D5 (which is kind of sad, considering that the top costs about $48 total to the US...plus the cost of a DDC+). Here you can also see that my top is sitting right about where I thought it would in this sort of loop--between the stock and modded Alphacool tops. Though, really, it ended up performing a lot closer, in terms of flowrate, to the modded Alphacool top than I expected.

Experimental Cooling Loop 2 (high restriction):

This was the second of two "real world" flowrate tests--mainly done out of curiosity. This particular test loop consisted of the following:
Reservoir --> 2.5" section of 1/2" ID Tygon --> Pump --> 1' section of 1/2" ID Tygon --> AqX MP-1 GPU --> 1' 8.25" section of 7/16" ID Masterkleer --> HW Labs Black Ice Pro Radiator --> 1' 3" section of 7/16" ID Masterkleer --> AqX MP-05 SP LE --> 1' 3.25" section of 1/2" ID Tygon --> Fort-Pak
I don't have a photo for this one... but it was set up in the same way that the previous test loop was.

Test Results:

AqX 50z: 12.00V -- 1.08 Gal./min. | 13.80V -- 1.22 Gal./min.

Laing D5: 12.00V -- 1.18 Gal./min. | 13.80V -- No Data.

Laing DDC+: 12.00V -- 1.29 Gal./min. | 13.80V -- 1.48 Gal./min.

DDC+ w/ stock Alphacool: 12.00V -- 1.36 Gal./min. | 13.80V -- 1.57 Gal./min.

DDC+ w/ modded Alphacool: 12.00V -- 1.38 Gal./min. | 13.80V -- 1.59 Gal./min.

DDC+ w/ Radiical Top: 12.00V -- 1.12 Gal./min. | 13.80V -- 1.29 Gal./min.

DDC+ w/ Petra's Top: 12.00V -- 1.36 Gal./min. | 13.80V -- 1.55 Gal./min.

...and a more graphical representation:

Comments: Now, this test I don't really consider that accurate, as far as my own top is concerned, because it was the first test done with the prototype and it was carried out late in the evening, several days after all of the other testing (i.e. human error). Though, really, the results between both Alphacool tops and my own are so close that the differences could easily fall under 'margin of error.' At higher restriction, I would expect my top to perform a tiny bit better than the stock Alphacool top and the modded Alphacool top due to it matching the max head of the stock Alphacool top and exceeding the max flow. Granted, the differences I'm talking about here are so small that they would likely never be noticed in an actual cooling loop.

Anyway, you can see the Radiical top is, yet again, bested by the D5... However, this time the stock DDC+ has also joined the D5 in out-flowing the Radiical top. Speaking of the D5, you're likely wondering why there's no 13.80V data for that test. Well, as we were finishing up the D5's 12V testing, it had the misfortune of ingesting a small rock which, even after removal and cleaning, resulted in the pump being no longer able to produce consistent results and occasionally failing.

Random Photos & Comments:

Take what you will from all of that... but I think that things went pretty well. Also, thanks to Margaret for helping out with refining/helping with the tests and providing that spiffy power supply. I'll work on getting pricing and such figured out for our DDC tops this week as I need to speak with the guys at the machine shop to see what can be done. Also, I likely won't be able to address any questions until Monday afternoon since I'll be spending most of Monday morning driving back from SLO.

Whee! Random photos!

Margaret prepping between trials...while I play with the camera

Our newly re-built water cooled Cardboard Box Computer (this time around, I went for the USPS Edition Tech Station look):


Extended Testing Results:

Series Testing:

Anyway, as I mentioned, I don't have enough of the other tops to do dual pump testing (only have one stock Alphacool, one modded Alphacool, and one Radiical) so I ran the tests with my two prototypes and compared them to the results I obtained previously for my top. There was no point in attempting to do max head pressure testing with the dual pump setup as it would likely exceed the max pressure my gauges can measure (15psig). As such, I stuck to loop testing only. The only change made to the loops for these tests was the addition of 1 foot of tubing to link the pumps.

The Twins:

Now, I realize that I didn't join the pumps in the most ideal way... I just didn't think of it until I was finished. Anyway, it shouldn't make a detectable difference, as far as this method of testing goes.

Experimental Cooling Loop 1 (average restriction):

Test Results:

DDC+ w/ Petra's Top: 12.00V -- 2.32 Gal./min. | 13.80V -- 2.58 Gal./min.

2x DDC+ w/ Petra's Top: 12.00V -- 2.78 Gal./min. | 13.80V -- 3.08 Gal./min.

...and a more graphical representation:

Experimental Cooling Loop 2 (high restriction):

Test Results:

DDC+ w/ Petra's Top: 12.00V -- 1.36 Gal./min. | 13.80V -- 1.55 Gal./min.

2x DDC+ w/ Petra's Top: 12.00V -- 1.84 Gal./min. | 13.80V -- 2.05 Gal./min.

...and a more graphical representation:


Parallel Testing:

Experimental Cooling Loop 1 (average restriction):

Test Results:

DDC+ w/ Petra's Top: 12.00V -- 2.32 Gal./min. | 13.80V -- 2.58 Gal./min.

2x DDC+ w/ Petra's Top (Series): 12.00V -- 2.78 Gal./min. | 13.80V -- 3.08 Gal./min.

2x DDC+ w/ Petra's Top (Parallel): 12.00V -- 2.43 Gal./min. | 13.80V -- 2.78 Gal./min.

...and a more graphical representation:

...more evidence that, even in a somewhat low restriction loop, parallel pumps don't do much.

Additionally, I fiddled around and determined that the DDC+'s will start at up to about 13.19V--to achieve anything above that, you have to start the pumps at that point (or below) and then increase the voltage to the desired amount (13.8V, in this case).

Oh, yeah, take a look at this.... (the sample barbs that Eddy was nice enough to send to me arrived yesterday)

...I'm such a tease

*edit...again*Here are some updated photos:

The machine shop decided to put a MUCH nicer finish (internally and externally) on the production version than they did on the prototypes. The tops are so smooth that they're shiny and a little reflective!!! (though, tooling marks can still be seen... the photos really don't do the finish justice)







!--End of Archive--!

[Forum Archive '06] Scrap Acrylic Shrouds

Here's a little tidbit from early 2006... a pair of radiator fan shrouds that I made out of some scrap acrylic that I had sitting around the warehouse. Total forum thread views: 1,086 (as of 4/28/2010). As a side note, all images have been re-sized via HTML, rather than my usual two image (preview & full image via link) approach, to speed along the archiving process. You can still right-click on the images and view them at full-size in a new tab/window. Also note that all vBulletin emoticons and their associated code have been removed from the archive entry.

!--Start of Archive--!

So.... What happens when I get my hands on some scrap acrylic and a bonneville heater core? Well, this, of course!


Yup, just a couple quick and dirty fan shrouds made from 1/4" acrylic, put together with acrylic solvent, and sealed with some clear RTV silicone. Okay, so there was a little sanding involved, too...


The edges of the shrouds (including the center divider) were sealed with this spiffy, soft natural rubber sealing cord stuff. It's really cool and works great...and I'm kinda thinking about adding it to my store but, at the moment, I can only obtain maybe 50' of the stuff...


Bolt everything together, add 4 120x38mm Sanyo Denki fans @ 12v in push/pull, and what you get is a crazy amount of air moving through that heater core.


The whole assembly is about 8 1/4" thick and should serve my test rig well (especially since the ambient temp in here is usually between 15°c and 17°c during this time of the year).

Just thought I'd share.

!--End of Archive--!

[Forum Archive '05] T3h uber l337 Water Cooled Cardboard Box!

I bring you a little treat from 2005... T3h uber l337 Water Cooled Cardboard Box! The build was used as a display system in the front office of Petra's Tech Shop and went through several rebuild phases as time passed (with this, of course, being its first form). Total forum thread views: 6,685 (as of 4/28/2010). As a side note, all images have been re-sized via HTML, rather than my usual two image (preview & full image via link) approach, to speed along the archiving process. You can still right-click on the images and view them at full-size in a new tab/window. Also note that all vBulletin emoticons and their associated code have been removed from the archive entry.

!--Start of Archive--!

You guys'll probably find this one amusing... First, a bit of background, though...

I've been planning on doing a nice bit of modding to a Sonata2 to use as a display machine; however, due to the amount of modding and painting involved, it would likely take me about two weeks from start to finish for the paint alone. Unfortunately, I didn't have that kind of time available before a little 'opening party' that was thrown a couple weeks ago... so... I decided to use a cardboard box instead.

It was actually surprisingly difficult to get things mounted in there...then again, I don't normally build computers into cardboard boxes. Also, all of the tubing and stuff was cut for the Sonata, so it's kinda crazy in there (with a little extra wire mess thrown in for added ghettoness).

Specs:
A64 3000+ Venice (haven't had time to overclock yet...)
1GB Crucial Ballistix PC-3200 (borrowed from personal gaming rig)
DFI LP NF4 Ultra-D
MSI GF6600GT
Water Cooling: Silverprop Cyclone EvolutionSE (CPU), Silverprop Cyclone FusionHL (GPU), BIP2, Swiftech MCP655, and 7/16" ID Tygon tubing.






Fun with the fish-eye lens...

The wall prints are on loan from my parent's photography studio, in case you were wondering. ...and the default WinXP background is there because I had just installed the OS the night before. Yay for procrastination... ;)

My kid brother playing NFS:MW

Anyway, I can't wait to find the time to overclock it (especially since my warehouse/office gets really cold). At stock speeds (which really isn't saying much) the CPU is idling at 22°C...which is a mere 4°C above the ambient room temp!

That's 4°C assuming that the thermostat on the wall a few feet away is somewhat accurate and assuming that the temps reported by nTune aren't just FUD, of course....

!--End of Archive--!