Danger Den RBX Acclerator-nozzle 1vs5 Comparison

Cooling/Water Cooling by KeithSuppe @ 2004-02-12

An in-depth look at Danger Den?s latest water block. The standard nozzle of the RBX can be swapped for a more performant one; we set out to see what difference it actually makes!

  • prev
  • next

The results: Nozzle number 1

Prior to commencing I must comment on one confounding issue. The Asus P4C800E-Dlx Vcore is entirely inaccurate as any owner must know by now. There's even a fix for the issue in the hardware forums. Basically diode seems to read 0.07V above the actual Vcore. The 3.0C requires 1.550V; however the diode will read 1.62V every time. When setting the Vcore higher, it seems to read the voltage more accurately. When stressing the CPU, overclocking, and running programs such as Seti, or Pifast, the Vcore will then drop as low as 1.50V, and I feel this may be the reason behind the re-boots. An entire article could, and perhaps should be written on this subject, for such a beloved motherboard. For these tests it's simply important to note when the Vcore reads 1.62, it is actually set at 1.550V in the BIOS. When the Vcore is down around 1.54V, or 1.52V it's under Seti@Home, and/or Pifast. When overclocking, I do not raise the Vcore as it's not necessarily with this CPU, and the purist philosophy behind overclocking is running at a higher frequency without having to increase voltage, otherwise it might be given the moniker "Overvolting.".


Beginning with the accelerator nozzle number-1, which is the factory installed nozzle, and ships with the RBX. It has the widest inlet orifice which is an extended oval (similar to the nozzle on the top left in the photo above), and runs perpendicular to the heat-voids/channels at the base of the block, above the CPU. The system conditions below represent temps at idle. This involves running the 3.0C @ default (3014MHz), Vcore on default (1.550V) and running Corsair Twin-X4000PRO at 200FSB/1:1 Ratio on 2.65V I used Asus Probe, and CPUCool Ver.7.2.8 to measure temps/volts;

Madshrimps (c)



The next screenshot is representative of the system under LOAD conditions, while running at default speeds. All parameters remained the same as above, the variable involved taxing the system running Seti@Home ver. 3.8 for approximately 45-minutes, until the system reached a temp plateau;


Madshrimps (c)


There is an increase in the CPU temp of approximately +7C/10F between Idle and Load while running at default speeds. With the number-1 nozzle still installed, I took measurements Overclocked. I pushed the CPU to 3704 MHz (Vcore at default/1.550V) raising the FSB frequency to 250FSB at a 1:1 Ratio. I increased the VDIMM to 2.85V. Therefore screenshot below represents the system overclocked, yet running Idle;

Madshrimps (c)



For the final measurement with the number-1 nozzle installed, I stressed the system to full LOAD conditions, overclocked at 3704MHz (250FSB, 1:1 Ratio), concurrently running Seti@Home, this time I increased Vcore to 1.70V as I experienced several re-boots attempting to stress the system running it overclocked, and under LOAD conditions utilizing Seti@Home. In this circumstance the Vcore reads 1.62V per the P4C800E-Dlx thermistor/diode inaccuracies, which ironically, are at least consistent as to the margin of error in their inaccuracy;


Madshrimps (c)
  • prev
  • next