project: Nearly Impossible

Hey all! I am currently starting my first, and very ambitious custom WC build. This is a winter project so don't expect final results overnight

Background: I am a drafter for a manufacturing company, so I am quite experienced with Solidworks (a popular 3D modelling software). I have access to a variety of materials, as well as a variety of CNC and manual machines which I can use at no charge. In addition, I have a fairly decent hands on skill set which will come in handy for the custom components of the build.

Purpose: This machine will be for recreational drafting and video editing as well as serving as a home file/media server, so the components will be more professional oriented rather than gaming oriented. There will be at least a mild overclock on it, but nothing that will degrade stability or reliability too much.

Case: I am currently designing a custom case which I plan on building out of 1/8" aluminum. I will post up some images soon, and as the design progresses.

Component parts list:
-Asus P9X79 pro MB -purchased-
-Intel i7 4820k
-16GB+ DDR3 1866mhz
-1x AMD/ATI Firepro V5800 -purchased-
-2x 128GB SSD in raid 0
-1x 2TB HDD
-850w+ PSU

Cooling parts list:
-1x D5 vario -purchased-
-2x 360mm rads hopefully Alphacool NexXxos XT45 or XSPC RX series -to be ordered very soon-
-PrimoChill rigid tubing
-PrimoChill rigid ghost compression fittings
-Custom acrylic "Tubing saver" panel over MB
-Custom acrylic res
-Custom CPU block
-Custom MB blocks (for show)
-Custom full cover GPU block
-Custom memory blocks (maybe, but only for show)
-Custom pump top (maybe)
-Fans... to be decided..

Goals:
-Custom waterblocks that work well!
-Delta Temp below 10c when loaded with fans running <1000rpm
-Near silent normal operation
-Fast
-Very clean and custom looking
-Enough cooling to support another V5800 in crossfire down the road without compromising my quietness goal

So, what are your reactions?

Looks like a nice project...What will you do with this rig?
Intrigued by the gpu choice. (and no hexa, and too few ram sticks )

snowman0040 said:

Goals:
-Custom waterblocks that work well!
-Delta Temp below 10c when loaded with fans running <1000rpm
-Near silent normal operation
-Fast
-Very clean and custom looking
-Enough cooling to support another V5800 in crossfire down the road without compromising my quietness goal

So, what are your reactions?

Click to expand...

Really interested in this Delta below 10c deal.

So at this point I'd like to ask what your going to chill the loop with?

Meh you could do this if you have enough Rad and an ambient room temp of probably around freezing or at the very least have liquid temps below 40F.

The to add another GPU and keep the same Delta? That's going to require more rads and fans at 10c or below.

Or perhaps the 10c was just a typo

10c delta means temp differential...not 10c temps.

The rig is for drafting and video editing, so that is why I have a professional series GPU. I decided against the 6 core CPUs for now. First of all because of price and secondly, the fact that I am experimenting with custom waterblocks so I dont really want to risk a $600+ CPU. 16gb of ram will be split into 4 sticks to fill all 4 channels.

Delta temp of 10 deg. is referring to the difference between water temp and ambient, which is usually around 21c / 70f. No chillers on this rig and with the future extra GPU, I know it will raise my delta, just not wanting to have to scream my fans to keep nice temps. These graphics cards are <75W so that doesn't concern me too much.

If you are going to do custom blocks, Make the CPU and MoBo chipset block all in one.

Otherwise I would just buy off the shelf blocks and custom make new tops for them.

Once you get your designs done, check with a machinist. I can't count the amount of times I have seen designs come in from drafting only to have to throw them back.

Soulcatcher668 said:

If you are going to do custom blocks, Make the CPU and MoBo chipset block all in one.

Otherwise I would just buy off the shelf blocks and custom make new tops for them.

Once you get your designs done, check with a machinist. I can't count the amount of times I have seen designs come in from drafting only to have to throw them back.

Click to expand...

hmm.. I like this all in one idea... the cpu and vrm could be one block, the chipset would have to be separate, but that's okay. I wasn't able to find any blocks available specifically for my board.

I don't exactly know a lot of machinists... I will be building these myself.. I can post up some drawings once I have the designs finalized if you or someone else here wants to look them over and give machining advice! My biggest concern lies with the poor machinability of copper.. ETP copper would give me the best thermal conductivity, but if I can sacrifice about 8% of that, I can use C145 copper which has much better machining properties. I wish i knew what alloy of copper is used for commercially available waterblocks.

For the cpu and gpu, I plan on using a copper "micro fin" design very similar to most of the current waterblock designs. These fins would be machined using a slitting saw in a cnc milling machine. I am considering a .020" thickness for fins and channels.

The 145 (tellurium) compound will be fine.
You are not going to see any gains going with the oxygen free stuff in this application. The temps are not high enough.

Soulcatcher668 said:

The 145 (tellurium) compound will be fine.
You are not going to see any gains going with the oxygen free stuff in this application. The temps are not high enough.

Click to expand...

That is great to hear. my assumption would be that the full hard temper would be best for a project like this? I am much more familiar with different steels and aluminum than I am with copper.

Welcome to OCFs!

Thanks for sharing this adventure with us. Can't wait to see what's in store.

Subbed!

+1^

Thanks for the warm welcome guys! I will do my best to keep you all up to date as my build progresses!

Hey guys, here is a couple case pictures to give you an idea of what's in store.
The design isn't even close to done, especially all of the WC components, but its a start!

pump will be located on the other side of the case, which wont have a clear window. res is located behind the rectangular cut out on the front panel. you can see the connection block at the bottom front corner of the removable MoBo tray, this will alow quick disconnect of the loop for removal of the tray. yet to be drawn is the tubing saver panel which will also act as the window on the MoBo side of the case.

Update: 01

Hey guys! sorry for the long silence... I have been quite busy with work and stuff lately. I do finally have a bit of an update though.

A couple weeks ago, I picked up a bunch of WC parts from a guy at work for $150 (some of which is pictured below). included two 12fpi 240mm rads, a DDC pump with top, some assorted bitspower fittings, and a waterblock that I will probably never use lol. It was all run in a loop, but its in good condition so I figured the price was right. I will most likely save most of it for a future htpc project.

On a more relevant note, I got two new XSPC EX360's for this build. The shipment got screwed up and I ended up with one dual pass and one crossflow, but after looking at how it would work in the case, I decided to keep both as this configuration will make my tube routing easier.



In other news, I sent my last case design to the circular file. I am now planning on building the slightly more traditionally styled revision 5.1 which is pictured below. As you can see, this design still has both rads exhausting out the top. I cannot decide where to put my air intake though... I suppose on the bottom would be best as it looks the cleanest. I don't really want fans in the front of the case because I will hopefully be building an LCD into there for system monitoring software.

Where do you guys think I should put my intakes?

Yup, I like the bottom as intake and top as exhaust. A lot of the cases nowadays have that design and it works great.

In my 900D I do Front, top and bottom as intake and back as exhaust. (Positive Pressure)

I purchased some slitting saws on eBay today which I plan on using to machine the fins in my custom copper water blocks. The blades are .014", .015", and .020" thick. The thinner grooves and fins created by the .014 blade would be nice because I could get a higher FPI in the blocks, but I have a feeling that .020 will be difficult enough to machine. I will do some tests once I get my hands on some copper stock, and build an arbor to run these saws in the milling machine.



I decided to use C110 copper instead of 145 as it seems to be the standard for waterblocks, plus it's more readily available.
Does anyone here have experience with slitting saws or machining C110 copper? Any insight/advice would be much appreciated as I don't have experience with this tool, or copper.

The reason that C110 is the standard is because it is the cheapest. Works fine though.

If those slitting saws are used, throw them away. You need to use new tools or tools that have only been used on Bronze Brass or copper.

Some of the things to keep in mind.
Copper likes to heat up while being cut. If you are not careful that will close your cut on the tool and break it. So, lots of coolant.
Thin saws are not a problem, thin walls are. The fins will want to bend away from the saw. You may want to support the finish fins with shim stock.
Always climb cut with slitting saws.

Be careful with cam software.
I can tell Mastercam to use a .010" end mill that is 10 inches long and it will simulate that cutter working perfectly.

Soulcatcher668 said:

The reason that C110 is the standard is because it is the cheapest. Works fine though.

If those slitting saws are used, throw them away. You need to use new tools or tools that have only been used on Bronze Brass or copper.

Some of the things to keep in mind.
Copper likes to heat up while being cut. If you are not careful that will close your cut on the tool and break it. So, lots of coolant.
Thin saws are not a problem, thin walls are. The fins will want to bend away from the saw. You may want to support the finish fins with shim stock.
Always climb cut with slitting saws.

Be careful with cam software.
I can tell Mastercam to use a .010" end mill that is 10 inches long and it will simulate that cutter working perfectly.

Click to expand...

Thanks for the great info!

The saws are new, just looks like they got a bit of staining on them. I can clean them up if it makes a difference on the cut quality.

I was planning on using regular flood coolant for the milling operations, would this be sufficient? or do I need to look at a different cooling/lubrication method.
Fins bending was my biggest concern with this part of the machining process. I will see what I can set up for supporting the fins.

I would agree with the mastercam thing, I will probably end up doing my testing in manual mode to figure out what feeds, speeds, and depths work best, then make programs to cut the blocks.

Flood coolant is fine.

I would start somewhere around 80 SFM and .001 FPT. The feed rate is really going to depend on how good your arbor and saw is.
If it is running out, only a few teeth will actually be cutting. Unfortunately with slitting saws that is the norm.

Soulcatcher668 said:

Flood coolant is fine.

I would start somewhere around 80 SFM and .001 FPT. The feed rate is really going to depend on how good your arbor and saw is.
If it is running out, only a few teeth will actually be cutting. Unfortunately with slitting saws that is the norm.

Click to expand...

Thanks! I will keep this in mind for my starting point.