Home
Ask a Question
Secure Orders
Latest News
Reviews
Projects
Customer Gallery
Guides
FAQ
Live Testing

Dander Den Torture Rack Three

I might start out by mentioning that there is a danger (sorry for the pun) in presenting results for this testing period using the Danger Den Torture Rack that the results will overshadow the product being reviewed. That is the Danger Den Torture Rack may be placed in a secondary position to the performance of the CPU and motherboard etcetera. To avoid this the results are presented in two parts. The Overview which deal with the Danger Den Torture Rack and the Performance section which looks at some of the large amount of data we collected over the last few weeks.

The Overview

Gotta love this Danger Den Case or should I say test bed. Igor still maintains that it should be referred to as the "Rack" but then again he has got issues.

From the outset this design shows that Danger Den have put a lot of consideration and thought into the Torture Rack construction. For example the placement of fan mounting points is in just the right position to cool VR's and Ram. The refractory air flow is carried over to the rear of the video card and on to heat pipe structures to assist with semi passive cooling. Cooling is this products major advantage over the traditional desk top tower case form. An open top case stops heat accumulating as would be the situation with a traditional tower case structure and allows the computer system to breath.

Danger Den also produce two closed versions of this case called the water box and air box respectively. Each is suitable for transport to LAN's and the like. Illustrated below is the Water Box version.

Fitted out with water cooling its looks like this;

and with a bit of bling it ends up looking something like this;

Pictures courtesy of Danger Den

Accessibility

The one major advantage of the Torture Rack is accessibility. Access to the parts contained inside this case is in my opinion peerless. I have never used anything before that even comes close. Want to change the water block or remount it for another test run...easy access. When you want to get at the pump to remove it and try another pump,guess what, there are convenient 120 mm fan hole spaces placed around the case to give direct access. As if that isn't enough.....there is more....the whole of the front of the case is open to fiddle with the bits. All this and security as well.

Security

You can locate all of the parts securely for transport should you need to move the case. Saves having that sinking feeling that you get after the new hard drive slips off the cardboard box you used to do a quick "out of case" build. You know the feeling as the Sata cable unplugs and the hard drive plunges Earthwards. The sound as it bounces off the floor. Well that's not likely to occur any more. Conveniently placed mounting points underneath the case permit the hard drive to be secured to the side wall out of harms way. Even water cooling pumps can be screwed to one of the many holes in the base or sides. Yep, its a big move forward for the regular tester of new products this Danger Den Torture Rack.

Alright so I have gushed I know it but I really am impressed. Not since the days of the desk top case has there been such a versatile case form. With careful case management it is conceivable that a quad,triple radiator based water cooled system could be constructed equaling and surpassing the UFO type cubes for a fraction of the cost. Not really but it is possible. Anyway the results coming out of this case indicate that even with a dual radiator based system the performance is quite acceptable..

Performance

Well we have reached the point where we try to make some sense out of what has been tested here at Radiical. It might be best at this time to review the parts used here and offer some comment on each as we go along.

Laing DDC 18W pump and Radiical Turbo Head top.

Lets first up look at the heart of the system the pump. We used one of our Radiical Turbo head tops with a Laing DDC 18W pump body. In addition we mounted the pump and top onto one of our pump stands to provide additional cooling through the aluminium stand material and to provide a mounting point to secure the pump.

Comments;

I might start out here by saying that the choice of pump in any system determines most of the efficiency outcomes for the system. During the testing period the pump was asked to provide sufficient flow to support a CPU block, GPU block, chipset block, hard drive block and of course the radiator. In this case the Radical Turbo Head modified Laing DDC pump worked flawlessly.

Radiical SRX water block.

Comments;

The SRX block was originally developed for Socket A based motherboards and has continued on as a multi fit block for practically all of the CPU variants available from AMD and Intel. We have even used the brass topped version as a water chiller pick up block in our development of water chilled water cooling systems. Combined with the use of more powerful pumps the SRX is now capable of coping with the significant heat loads of quad cored based CPU's. We anticipate that this design will become a leader in our water block range as the wattage values for CPU's continues to rise.

Maze 4 GPU block

Comments;

Danger Den have long been a favored GPU block for us because of the high build quality and reliable fitting kits supplied. Design wise Danger Den have recently changed models and now provide a Maze 5. We will present details of our testing of the Maze 4 against the Maze 5 on this system at a later date. Certainly we could find no issues with the performance of the Maze 4 in our primary baseline testing.

Radiical chipset block.

Comments;

The Radiical chipset block is becoming a very popular water block now that people are waking up to the potential overclocking available when additional cooling is provided for the Northbridge. There can be no doubt that replacing the heat pipe chipset cooler from this board with a water block greatly increased stability in our testing. Without the chipset water block the computer was prone to mysterious and un-accounted for reboots. Having previously found a lack of TIMM contact which we discussed in part two of this project there is no doubt that adding the chipset water block and a hard drive cooler was the correct decision.

Radiical Hard drive cooler.

Comments;

The Radiical Hard drive cooler was probably the best inclusion from a gaming point of view. Prior to its installation the hard drive would consistently run above 50 degrees Celsius. Failures in game were many and varied with our resident game jock getting increasingly vaporous on each occasion. Post hard drive cooler installation solid game play.

Finally, the Radiical Three Port© dual fan single pass radiator to form the loop.

To get it all connected we used some Excellon 2.9 metre's of 7/16" tubing.

Post testing results

A continuing note on the Radiical Turbo Head top and Laing DDC 18 W pump would indicate that it continues to perform flawlessly in this system proving quite conclusively that with these components there is absolutely zero need for a pump head greater than 4.5 M and a flow rate greater than 15 L/min. In answer to the most common question on pumping power "will such and such a pump work with my system" in this case the Laing DDC 18W fitted with a Radiical Turbo head top the answer is a resounding yes.

We were able to gain the following temperature reductions using the components mentioned above. At an ambient of between 28and 30 degrees the following reductions were noted.

Hard Drive Cooler

The pre water-cooled hard drive temperature at an ambient of 28 degrees Celsius was 50 degrees C. After water cooling was applied this became 34 degrees C.

Note; Game play stability was greatly enhanced with the addition of the water-cooling.

Chipset Cooler

Chipset temperature using heat pipes 56.2 degrees C when water cooled 34 degrees C.

Note; Chipset cooling increased memory performance and stabilized the higher overclock's that were reached.

GPU Cooler

GPU core temperature was 36 degrees on idle at 28 degrees ambient down from 52 degrees Celsius on air cooling.

Note; The Nvidia 7900GS GPU block was left at stock settings during these trials. Overclocking trials will be carried out later.

And after a short round of game playing the temperature was 38 degrees Celsius.

We placed the data coming from SpeedFan and the Web Temp program onto our live testing page for several days during early testing. We have selected here the data from a 1 hour gaming session and present this here to represent the temperature information relating to CPU, system and Hard drive temperatures all of which are now considered to be in a stable relationship with the ambient at a given time. Simply put the TIMM material has cured and we are now happy to consider these temperatures as useful data.

The ambient for this test run was 28.0 degrees Celsius.

Overclock speed and voltages are shown in the CPUZ graphic as a comparison.

As illustrated above the processor voltage is still within the range of Intel's specification and below what is the upper nominal amount of 1.325 volts usually attributed to this class of CPU. Stability at this level was also demonstrated with SuperPi in single processor testing for many hours before any stability level was accepted.

TAT readings were recorded for idle temps at this ambient (30 degrees Celsius) and are shown below.

TAT was used in conjunction with SuperPi to monitor upper ranges for temperature fluctuations and the session was logged with TAT.

It is interesting that the range of temperature variation was significantly less than would have been expected. Varying from 65 to 67 degrees across the active CPU.The idle CPU leveled out at 66 degrees Celsius and remained stable. Ambient temperature was 30 degrees Celsius during the course of this testing. Of course TAT and similar forms of temperature measurement relate to a single spot on the CPU core and do not aggregate the CPU surface temperature. As a result the temperature reading is taken from the hottest part of the CPU. We could not record anywhere near these sorts of temperatures when we used our RTD machined into the base base of the CPU water block. Neither was the water temperature anywhere near this upper level of 67 degrees Celsius. Water entry temperatures constantly remained less than half of this figure. With no figures even remotely near the upper operating temperatures being reported by TAT we were left to conclude that temperatures in the areas being measured by the Intel's TAT cannot be reached effectively by any form of cooling save water chilling and/or phase change. We have, therefore, some reservations as to the significance of this type of onboard TAT measurement in assessing cooling effectiveness.

In a further development it appears that the mysteriously high temperatures reffered to above are being presented by TAT incorrectly. The following URL discusses this issue at greater depth.
http://forumz.tomshardware.com/hardware/Core-Duo-Temperature-Guide-ftopict221745.htmlThe major finding is that the TAT values presented above for CPU temperature should be reduced by 15 degrees. This is more inline with the observations we made about water temperatures and highlights the need for a more accurate and reliable method of temperature evaluation than is currently possible with these software based solutions.

Game Playing;

The whole idea of building a gaming machine is to be competitive with aunty Grace on the Sundays when you visit for lunch correct........................................................Nah. The idea of Internet game play is to kick some remote booty and show your the "One" right.

So how did we go. Well we took a basic system which would stay connected to Battle Field 2 internet servers for about 15 minutes and then we made it an all day every day machine. The resident game jock is still bleary eyed and lacking any coherence after all night sessions.

The transformation has been quite remarkable and along the way we learned quite a lot about EA's BF2 game. One such revelation was when we used the Web Temp program to measure Ram usage during game play. During play the ram usage would rise intermittently to levels just exceeding 1Gig. We removed 1 Gig of Ram from the machine to see what would happen and sure enough erratic game instability. Typically the game would freeze and have to be restarted. The freezing problem was at best irregular but the Punk Buster denials were very very frustrating. We would imagine this sort of issue would have the average player pulling their hair out in frustration as it usually occurred during heavy fire fights. So for all you out there having game stability issues when using 1Gig of ram the quick cure is to just add another gig and "go get em".

Needless to say there are gaming sites out there that can provide a more in depth guides to this game so we will leave the other game configuration details to them.

Conclusion

Overall we are very impressed with the outcome of our Danger Den Torture Rack and the gaming machine we have created. We look forward to much higher scores in the future in what is now a very competitive gaming rig.

Copyright Radiical Pty Ltd 2007