SRX and Shuttle Makeover

The SRX and Shuttle Makeover

Date Log; 24 August, 2006

We have been using the water cooled Shuttle since the day that we completed it for one of our earlier projects here. Today we look at updating the SRX and Shuttle to the new Three Port Radiator©. The advent of our new Patent Pending Radiical Three Port Radiator© has increased the range of applications to which we can adapt the Three Port Radiator© special functions.

The original Shuttle featured one of the Laing DDC 12 volt pumps which had a miserly 350 l/h output. We corrected this by using one of our Turbo Heads which raised the output to almost 500l/h with a corresponding increase in pump head At the same time we took the opportunity to increase the cooling performance of the graphics card by water cooling it as well.

The most important aspect of this makeover was of course the addition of the Three Port Radiator©. With this fitted filling, bleeding and draining the system is a breeze. No more waiting for the system to remove the air to get started sort of plug and play really.

Prior to fitting the new radiator to the Shuttle I gave the radiator a thorough internal cleanup. Little point in my opinion in ending up having to take down what is a very difficult system to build because of the dreaded white tubing deposit. On this occasion I rinsed from both ends of the barbed inlet and outlet and also through the third port. At that point I would normally have been happy to proceed with the build except that I wanted to make real sure it was thoroughly cleaned. I decided to try the with vinegar cleaning process and here are the results. Remember the radiator has been rinsed in all possible directions before the vinegar treatment.

Clear unused vinegar on the left same vinegar after twelve hours of soaking in the radiator core on the right. The bottom of the glass used for the spent vinegar also contains a large amount of white powder similar to white sand destined I suppose for the interior of the tubing. Really a no brainier isn't it either clean your radiator first time round or put up with the white material shown here in your loop.

Alright with the radiator cleaned it's on with the project proper.

In the original Shuttle project the radiator placement involved two ports for filling and emptying of the system.

Three eights copper tubing was used originally to provide the necessary bend radii inside the case area in the original Shuttle project. For the makeover we will be taking advantage of the more powerful pump to drop the tubing size from 3/8" to a very manageable 5/16" ID tubing. To ensure clean lines we will be using black Tygon and compression fittings .

We will be removing the radiator and de facto these copper structures first to clear the space for the new design. Having cleared away the copper tubing and old radiator its time to fit the new radiator in place. We will be using our Radiical stand off kit again for this installation.

In this picture you can see the stand off kit and how the outlet has been placed into a PCI pass through and the inlet through a hole originally supplied for a rear case switch which has been removed.

We then mocked a solid Bleeder Wick©. With the Bleeder Wick© in place we get a much smoother top surface for the radiator mount and a much tidier external finish.

Next it's onto internal fitting. The pump is the first element so mounting this to provide the necessary clearances for tubing run out has to be carefully weighed up in the very small amount of space available.

We have mounted the Laing DDC pump with the Turbo Head in place on its side using one of our Laing pump stands. Clearance for mounting the pump on a parallel basis, the preferred option, was not available. A right angle discharge barb was fitted to the Turbo Head to direct flow to the GPU card first. Then onto the CPU block and back to the radiator to complete the loop. The picture illustrates the copper tubing used to fit up the GPU block so we better look at how we set that up next.

We are using a Silverprop Fusion SL GPU water block for this project based on the fact that it is the thinnest profile GPU block that we could get for the amount of space between graphics card and external case covering. No point having a GPU block fitted if the case sides will no longer fit on.

The problem with this block is it has molded barbs suitable for 1/2" tubing. Given that we will be using 5/16" tubing in this project we have to do some modification to the block to make it fit. With the view in mind that we will probably re use this block again in the future we have opted to retain the full 1/2" barbs and engineer the barbs to fit 5/16" tubing. To do this we have machined out the barbs internal diameter to accept copper 3/8" tubing.

Then we mock up the GPU block onto the graphics card to complete tube positioning prior to soldering the tubing in place.

In the completed block picture above the final adjustments involve very tedious measurements to remove any stress on the AGP slot and GPU so as to avoid any motherboard or video card damage. On each occasion something is changed you should be prepared to re measure fitting spaces for hard drives, optical devices like DVD and CD rom players and so on. With all the internal parts in place its time to seal the loop up add the fan and complete the wiring tidy.

We will be using our imbedded RTD to measure temperatures during the testing period so we better just mention the position for this on the SRX block.

The RTD is machined into the block base and sits slightly touching the IHS under surface. We have found this to provide very accurate temperature measurements when compared with say the Speedfan 4.29 core temperature reading facility. We will be using the same sort of cabinet reader as we used on the CM Stacker makeover project build so that we can readily access the incoming temperature readings.

With the system now completed it is onto the filling and bleeding part of our project.

If you use one of these Three Port Radiator©'s in a future project it is important to be aware of the filling process.

Filling

To fill the Three Port Radiator you will need a small length say 75 mm (3") of 1/2" tubing and a funnel set aside for your water cooling adventures. Add water by trickling it into the tubing rather than pouring at a rate which will fill the funnel. The filling process should be gradual so that an air lock does not form in the Bleeder Wick©. If you do form an air lock tilt the Shuttle case until the air lock is broken and the coolant flows into the header tank.

Care should be taken to ensure that the inlet tubing is full and that no air is present in the inlet tubing leading to the pump or within the pump. If any air is found tilt the air bubble so that it is encouraged to rise towards and into the radiator.

When you are satisfied that the radiator and pump are filled with water top up the funnel with a little water. Do not over fill the funnel. Switch on the pump and as soon as it starts switch it off again. The coolant will race into the unfilled portions of tubing. Refill the funnel and repeat until the whole of the tubing is filled and the radiator will accept no more fluid from the funnel. The pump can now be started and left on. At this stage what should be observed in the Bleeder Wick© is a mass migration of small bubbles rising out of the coolant in the Bleeder Wick© tubing. Within a matter of minutes the system will have reached a primary bleed state where very few bubbles are passing into the Bleeder Wick©. I prefer to leave the Bleeder Wick© in place for a few days to remove any rogue micro bubbles or air dissolved in the water after the pumps mixing action.

Bleeder Wick©'s come in a variety of shapes and designs. From fully rigid units to whatever flexible tubing colour's you want to use. We supply a small clear tubing version and top with our radiators like the one shown below.

Of course you could always just refit the sealing plug for an uncluttered appearance.

Mind you my favorite is the "Is it a water cooled Wi-Fi" version.

After filling its time for the the dress up shots and final views of the Shuttle makeover before it gets pressed back into action.

Conclusion;

With this makeover we have accomplished several objectives that we did not achieve in the first Shuttle build. We have reduced and simplified filling time substantially, provided a very valuable fast bleed function and made emptying the system a simple tilt and pour action. Add to this we have increased pump performance, added graphics card cooling and increased incase air flow with less tubing clutter. Overall a very good upgrade to our original Shuttle project and a very simple makeover.